Skip to main content

Full text of "1957-1958 Year Book - Carnegie Institution of Washington"

See other formats

Digitized  by  the  Internet  Archive 

in  2012  with  funding  from 

LYRASIS  Members  and  Sloan  Foundation 

YEAR    BOOK     57 

July    1,    1957— June   30,    195 



Library  of  Congress  Catalog  Card  No.  3-16716 






Mount  Wilson  and  Palomar  Observatories  49 

Joint  Committee  on  Image  Tubes  for  Telescopes  89 

Department  of  Terrestrial  Magnetism  93 

Geophysical  Laboratory  167 

Department  of  Plant  Biology  261 

Department  of  Embryology  307 

Department  of  Genetics  373 

Department  of  Archaeology  433 



Report  of  the  Executive  Committee  461 

Report  of  Auditors  463 

Abstract  of  Minutes  of  the  Sixtieth  Meeting  of  the  Board  of  Trustees  477 



INDEX  489 


Caryl  P.  Haskins 

Walter  S.  Gififord,  Chairman 

Barklie  McKee  Henry,  Vice-Chairman 

Robert  Woods  Bliss,  Secretary 

James  F.  Bell 
Robert  Woods  Bliss 
Lindsay  Bradford* 
Omar  N.  Bradley 
Vannevar  Bush 
Walter  S.  GifTord 
Crawford  H.  Greene  wait 
Caryl  P.  Haskins 
Barklie  McKee  Henry 
Ernest  O.  Lawrence  { 
Alfred  L.  Loomis 
Robert  A.  Lovett 
Keith  S.  McHugh 
Margaret  Carnegie  Miller 
Henry  S.  Morgan 
Seeley  G.  Mudd 
William  I.  Myers 
Henning  W.  Prentis,  Jr. 
Elihu  Root,  Jr. 
Henry  R.  Shepley 
Charles  P.  Taft 
Juan  T.  Trippe 
James  N.  White 
Robert  E.  Wilson 

*  Resigned    1958. 

%  Died  August  27,  1958. 

TRUSTEES  Continued 


Barklie  McKee  Henry,  Chairman 
Robert  Woods  Bliss 
Walter  S.  Gitford 
Caryl  P.  Haskins 

Robert  A.  Lovett 
Henry  S.  Morgan 
Henning  W.  Prentis, 
Henry  R.  Shepley 



James  N.  White,  Chairman 
Walter  S.  Giflord 
Alfred  L.  Loomis 
Henry  S.  Morgan 
Henning  W.  Prentis,  Jr. 


Elihu  Root,  Jr.,  Chairman 
Robert  Woods  Bliss 
Walter  S.  Giflord 
William  I.  Myers 



Keith  S.  McHugh 
Alfred  L.  Loomis 
Juan  T.  Trippe 


Barklie  McKee  Henry 
Henry  S.  Morgan 


Seeley  G.  Mudd,  Chairman 
Crawford  H.  Greenewalt 
Elihu  Root,  Jr. 


Alfred  L.  Loomis,  Chairman 
Margaret  Carnegie  Miller 
William  I.  Myers 
Charles  P.  Taft 


Ernest  O.  Lawrence,  Chairman 
Barklie  McKee  Henry 
Henning  W.  Prentis,  Jr. 
Robert  E.  Wilson 


Henry  R.  Shepley,  Chairman 
James  F.  Bell 
Robert  Woods  Bliss 
Juan  T.  Trippe 

Note:  Membership  of  Committees  as  of  June  30,  1958. 




Daniel  Coit  Gilman,  1902-1904  Robert  Simpson  Woodward,  1904-1920 

John  Campbell  Merriam,  President  1921-1938;  President  Emeritus  1939-1945 
Vannevar  Bush,  1939-1955 


Alexander  Agassiz 


Wayne  MacVeagh 


George  J.  Baldwin 


Andrew  W.  Mellon 


Thomas  Barbour 


Roswell  Miller 


John  S.  Billings 


Darius  O.  Mills 


Robert  S.  Brookings 


S.  Weir  Mitchell 


John  L.  Cadwalader 


Andrew  J.  Montague 


William  W.  Campbell 


William  W.  Morrow 


John  J.  Carty 


William  Church  Osborn 


Whitefoord  R.  Cole 


James  Parmelee 


Frederic  A.  Delano 


Wm.  Barclay  Parsons 


Cleveland  H.  Dodge 


Stewart  Paton 


William  E.  Dodge 


George  W.  Pepper 


Charles  P.  Fenner 


John  J.  Pershing 


Homer  L.  Ferguson 


Henry  S.  Pritchett 


Simon  Flexner 


Gordon  S.  Rentschler 


W.  Cameron  Forbes 


David  Rockefeller 


James  Forrestal 


Elihu  Root 


William  N.  Frew 


Julius  Rosenwald 


Lyman  J.  Gage 


Martin  A.  Ryerson 


Cass  Gilbert 


Theobald  Smith 


Frederick  H.  Gillett 


John  C.  Spooner 


Daniel  C.  Gilman 


William  Benson  Storey 


John  Hay 


Richard  P.  Strong 


Myron  T.  Herrick 


William  H.  Taft 


Abram  S.  Hewitt 


William  S.  Thayer 


Henry  L.  Higginson 


James  W.  Wadsworth 


Ethan  A.  Hitchcock 


Charles  D.  Walcott 


Henry  Hitchcock 


Frederic  C.  Walcott 


Herbert  Hoover 


Henry  P.  Walcott 


William  Wirt  Howe 


Lewis  H.  Weed 


Charles  L.  Hutchinson 


William  H.  Welch 


Walter  A.  Jessup 


Andrew  D.  White 


Frank  B.  Jewett 


Edward  D.  White 


Samuel  P.  Langley 


Henry  White 


Charles  A.  Lindbergh 


George  W.  Wickersham 


William  Lindsay 


Robert  S.  Woodward 


Henry  Cabot  Lodge 


Carroll  D.  Wright 


Seth  Low 


Under  the  original  charter,  from  the  date  of  organization  until  April  28,  1904,  the  following  were 
ex  officio  members  of  the  Board  of  Trustees:  the  President  of  the  United  States,  the  President  of  the  Senate, 
the  Speaker  of  the  House  of  Representatives,  the  Secretary  of  the  Smithsonian  Institution,  and  the  President 
of  the  National  Academy  of  Sciences. 




813  Santa  Barbara  Street,  Pasadena  4,  California 

Ira  S.  Bo  wen,  Director;  Horace  W.  Babcock,  Assistant  Director 

Halton  C.  Arp 
Walter  Baade  * 
William  A.  Baum 
Arthur  D.  Code  % 
Armin  J.  Deutsch 

Jesse  L.  Greenstein 
Fred  Hoyle 

Rudolph  L.  Minkowski 
Guido  Munch 
J.  Beverly  Oke 

Donald  E.  Osterbrock  § 
Robert  S.  Richardson  § 
Allan  R.  Sandage 
Olin  C.  Wilson 
Fritz  Zwicky 

5241  Broad  Branch  Road,  N.  W .,  Washington  15,  D.  C. 

L.  Thomas  Aldrich 
Ellis  T.  Bolton 
Roy  }.  Britten 
Bernard  F.  Burke 
Dean  B.  Cowie 

Merle  A.  Tuve,  Director 

John  W.  Firor 
Scott  E.  Forbush 
W.  Kent  Ford,  Jr. 
John  W.Graham  1 1 
Norman  P.  Heydenburg 

Richard  B.  Roberts 
Howard  E.  Tatel  ff 
Georges  M.  Temmer 
Harry  W.  Wells 
George  W.  Wetherill 


2801  Upton  Street,  N.  W .,  Washington  8,  D.  C. 

Philip  H.  Abelson,  Director 

Francis  R.  Boyd,  Jr. 
Felix  Chayes 
Sydney  P.  Clark,  Jr. 
Gordon  L.  Davis 

*  Retired  June  30,  1958. 
X  Resigned  August  31,  1958. 
§  Resigned  June  30,  1958. 
||  Resigned  February  8,  1958. 
f  Died  November  15,  1957. 

Gabrielle  Donnay 
Joseph  L.  England 
Hans  P.  Eugster 
Joseph  W.  Greig 

Gunnar  Kullerud 
J.  Frank  Schairer 
George  R.  Tilton 
Hatten  S.  Yoder,  Jr. 


STAFF  Continued 


Stanford,  California 

C.  Stacy  French,  Director 

William  M.  Hiesey  Harold  W.  Milner  James  H.  C.  Smith 

Malcolm  A.  Nobs 


Wolfe  and  Madison  Streets,  Baltimore  5,  Maryland 

James  D.  Ebert,  Director 

David  W.  Bishop  Elizabeth  M.  Ramsey 

Bent  G.  Boving  Mary  E.  Rawles 

Robert  K.  Burns  Royal  F.  Ruth 
Robert  L.  DeHaan 


Cold  Spring  Harbor,  Long  Island,  New  Yor\ 

Milislav  Demerec,  Director 

Alfred  D.  Hershey  Margaret  R.  McDonald 

Berwind  P.  Kaufmann  George  Streisinger  * 

Barbara  McClintock 


10  Frisbie  Place,  Cambridge  38,  Massachusetts 

Harry  E.  D.  Pollock,  Director 

Tatiana  ProskouriakofiE  A.  Ledyard  Smith 

Anna  O.  Shepard  Robert  E.  Smith 

Edwin  M.  Shook  *  J.  Eric  S.  Thompson 

*  On  leave  of  absence. 


STAFF  Continued 


1530  P  Street,  N.  W.,  Washington  5,  D.  C. 

Caryl  P.  Haskins 

Paul  A.  Scherer 

Executive  Officer 

Samuel  Callaway 

Assistant  to  the  President 

Ailene  J.  Bauer 

Director  of  Publications 

Lucile  B.  Stryker 

Earle  B.  Biesecker 

Bursar;  Secretary-Treasurer,  Retirement  Trust 

James  W.  Boise 

Assistant  Bursar;  Assistant  Treasurer,  Retirement  Trust 

James  F.  Sullivan 

Assistant  to  the  Bursar 

Richard  F.  F.  Nichols 

Executive  Secretary  to  the  Finance  Committee 

STAFF  Continued 


of  Carnegie  Institution  of  Washington 

William  A.  Arnold,  Oak  Ridge  National  Laboratory 

D.  G.  Catcheside,  University  of  Birmingham 

Hessel  de  Vries,  University  of  Groningen 

Louis  B.  Flexner,  University  of  Pennsylvania 

Willard  F.  Libby,  University  of  Chicago 

Paul  W.  Merrill,  Mount  Wilson  Observatory 

Jan  Hendrick  Oort,  Leiden  Observatory 

Harry  E.  D.  Pollock,  Carnegie  Institution  of  Washington 

Hans  Ramberg,  University  of  Chicago 

C.  E.  Tilley,  Cambridge  University 

M.  Westergaard,  Universitetets  Genetiske  Institut,  Copenhagen 

Evelyn  M.  Witkin,  State  University  of  New  York 

R.  v.  d.  Woolley,  Royal  Greenwich  Observatory 






The  stage  of  romance  is  the  stage  of  first  apprehension.  The  subject-matter  has  the  vivid- 
ness of  novelty;  it  holds  within  itself  unexplored  connexions  with  possibilities  half -disclosed 
by  glimpses  and  half-concealed  by  the  wealth  of  material.  In  this  stage  knowledge  is  not 
dominated  by  systematic  procedure.  Such  system  as  there  is  must  be  created  piecemeal  ad  hoc. 
We  are  in  the  presence  of  immediate  cognisance  of  fact,  only  intermittently  subjecting  fact  to 
systematic  dissection.  ...  In  our  conception  of  education  we  tend  to  confine  it  to  .  .  .  the 
stage  of  precision.  But  we  cannot  so  limit  our  task  without  misconceiving  the  whole  problem. 
We  are  concerned  alike  with  the  ferment,  with  the  acquirement  of  precision,  and  with  the 
subsequent  fruition. — Alfred  North  Whitehead  in  The  Rhythm  of  Education. 

The  greatest  of  man's  advances  are  made,  not  in  the  harshness  of  old  neces- 
sity, but  in  the  richness  of  new  opportunity.  With  mankind,  as  in  all  life, 
grinding  hardship  to  the  limit  of  endurance  may  be  met  by  sinewy  resistance, 
may  lead  to  extraordinary  heightening  of  old  skills,  to  a  marvelous  sharpening 
and  extension  of  well  tried  modes  of  existence.  But,  in  the  larger  sense,  it  may 
evoke  little  that  is  really  new.  Innumerable  generations  of  some  plant  or  animal 
species  may  live  successfully  and  die  peacefully  in  a  stable  and  long-occupied 
environment  and  in  the  end  show  little  evolutionary  change  beyond  a  host  of 
minor  adaptations,  albeit  exquisitely  specific  and  precise.  But  just  as  the  open- 
ing of  some  untried  ecological  realm  to  that  plant  or  animal  is  characteristically 
met  by  a  burst  of  large-scale  evolutionary  change,  swift  and  often  comprehen- 
sive, so  it  is  against  the  great  and  novel  challenge,  in  fresh  and  unknown 
gardens  of  the  intellect  and  spirit,  that  the  mind  of  man  has  always  found  un- 
recognized powers,  has  always  gleaned  strength  and  courage  and  capacity  to 
reach  new  worlds. 

On  our  well  worn  and  crowded  planet,  the  time  has  long  since  passed  when 
new  opportunities  of  this  order  can  be  grasped  unless  not  only  their  exploitation 
but  their  very  making  are  in  large  part  the  work  of  human  hand  and  mind. 
In  our  world,  the  whole  exploratory  process  has  become  essentially  self-stimu- 
lating and  self-creative,  the  spirit  soaring  on  wings  which,  by  necessity,  are 
largely  of  its  own  fashioning.  So  every  innovation  of  significance  brings  new 
opportunity  in  its  train,  and  that  new  opportunity,  in  its  turn,  breeds  innova- 
tion still  wider  and  more  encompassing. 

It  is  clear  that  this  situation,  in  essence,  has  been  characteristic  of  the  history 
of  man  at  least  since  the  tremendous  discovery  that  crops  could  be  reaped  from 
the  sown  seed.  But  by  its  very  nature  the  cycle  from  discovery  to  opportunity 
to  new  discovery  is  self-accelerating,  in  its  later  phases  reaching  explosive  di- 
mensions. So  we  could  perhaps  have  predicted  what  is  nonetheless  a  continuing 
miracle — the  golden  quality  of  our  age,  the  incomparable  richness  of  innova- 
tional  opportunity  in  our  time. 


No  vision  of  the  scale  of  the  heavens  and  earth's  place  within  them  of  a 
Galileo  or  a  Kepler  or  a  Tycho  Brahe,  no  revelation  with  the  crude  microscope 
of  a  Leeuwenhoek  of  a  whole  living  world  in  microcosm  which  for  an  earlier 
generation  was  simply  nonexistent,  can  have  enlarged  man's  view  of  the  uni- 
verse more  radically  than  some  of  the  modern  scientific  advances  of  the  last 
few  years.  That  change  of  scale  in  knowledge  and  in  viewpoint,  so  great  in 
magnitude  as  in  effect  to  differ  in  kind,  is  only  partly  represented  by  the  new 
conquests  of  the  physical  universe  by  which  today  we  often  characterize  it — 
the  building  of  transuranic  elements,  the  vast  and  varied  potentialities  inherent 
in  nuclear  fission  and  fusion,  the  advent  of  man-made  satellites.  It  is  only  partly 
represented  by  the  tremendous  advances  that  mark  our  time  in  knowledge  of 
the  nature  and  evolution  of  the  universe,  of  the  structure  of  matter,  of  the 
quality  of  that  immediate  shell  of  space  that  lies  hundreds  or  thousands  of 
miles  beyond  our  earth.  It  is  only  partly  represented  by  the  rate  of  growth  of 
the  scientific  effort  itself,  which  in  England — and  we  can  probably  apply  the 
relationship  to  the  recent  history  of  our  own  country  without  serious  error — 
has  been  shown  almost  certainly  to  have  doubled  two  or  three  times  in  a  gen- 
eration ever  since  the  days  of  Newton.  It  is  only  partly  represented  by  the 
present  magnitude  of  the  scientific  and  technical  effort,  merely  suggested  in 
our  own  nation  in  an  estimate  recently  made  by  the  Department  of  Economics 
of  McGraw-Hill  Publishing  Company  that  in  1958  the  expenditure  of  Ameri- 
can industry  alone  for  science  and  technology  reached  eight  billion  dollars. 
These  are  massive  and  powerful  factors  shaping  and  scaling  our  future  oppor- 
tunities for  innovation.  Yet  even  they  may  not  be  ultimately  the  most  signifi- 
cant of  our  time. 

The  major  revolutions  in  human  thought — the  silent  revolutions  of  deepest 
import,  when  the  basic  nature  of  man's  concept  of  the  universe,  of  his  rela- 
tion to  it  and  his  purpose  and  goal  within  it,  undergoes  the  greatest  changes — 
are  not  only  difficult  to  document  long  after  they  have  occurred.  They  are 
largely  invisible  to  those  who  pass  through  them,  who  have  the  high  fortune 
to  be  their  contemporaries.  Even  the  great  tangible  events  that  may  accompany 
such  changes — the  scientific  conquests,  the  social  upheavals,  the  political  unrest 
and  change — appear  at  the  time  largely  as  a  series  of  striking  but  disconnected 
and  disrupting  events.  In  the  perspective  of  three  hundred  years  it  is  not  very 
hard  to  see  how  interrelated  were  the  turbulent  material  changes  of  the 
sixteenth  and  seventeenth  centuries  with  the  deeper  revolutions  of  concept 
which  accompanied  them,  or  to  realize  how  often  that  inner  revolution  was 
reflected  in  the  more  material,  more  evident,  external  one.  Perhaps  we  too  are 
privileged  to  live  in  a  time  of  remolding  of  basic  concepts  that  will  ultimately 
be  as  fraught  with  opportunities  for  new  exploration  as  was  the  scientific  revo- 
lution of  the  sixteenth  century — that  may,  indeed,  form  the  most  powerful  of 
all  the  stimuli  for  innovation  provided  by  our  age. 


Consider,  for  instance,  the  great  changes  upon  which  we  may  just  now  be 
embarking  in  our  ideas  of  the  very  business  of  science,  defined  as  the  search 
for  truth — the  revolution  which  may  come  in  our  notions  of  the  nature  of 
scientific  truth  itself.  When  science  is  in  its  earliest  observational  stages  of  de- 
velopment, truth  may  mean  largely  the  quantified,  accurate  description  of 
phenomena,  so  made  that  any  investigator  who  repeats  the  same  observation 
under  the  same  conditions  will  emerge  with  the  same  set  of  sensory  impressions, 
and  will  be  likely  to  draw  similar  conclusions  from  them.  Thus  it  is  perhaps 
fair  to  say  that  the  earliest  and  most  basic  scientific  "truth"  was  the  sensory 
impression,  verifiable  by  many  men,  of  some  feature  of  that  world  of  phe- 
nomena which  was  conceived  to  lie  outside  of,  away  from,  and  unaffected  by 
the  observer.  It  was,  perhaps,  "knowledge"  as  Locke  and  his  contemporary 
empiricists  would  have  used  the  term. 

This  stage  had  already  been  outgrown  by  seventeenth-century  science.  Al- 
ready the  notions  of  truth  and  falsity  had  been  extended  beyond  observations 
alone  to  the  theories  constructed  to  contain  them.  A  prime  criterion  of  merit 
was  that  a  structure  of  cause  and  effect  should  be  objectively  defined,  should 
be,  as  Isaac  Newton  emphasized,  kept  "free  of  occult  influences."  That  concept 
of  cause  and  effect,  of  course,  derived  its  immense  influence  in  large  part  from 
the  spectacular  success  of  the  new  ideas  of  gravitation  in  providing  theoretically 
calculated  orbits  to  replace  the  descriptive  treatment  of  the  motions  of  the 
planets  which  had  been  current  since  the  days  of  the  Medievalists  and  their 
brilliant  Arab  co-workers — accurately  predicting  orbits  which  could  be  experi- 
mentally verified.  Through  all  the  following  years  to  our  own  century,  to  the 
very  beginning  of  the  great  revolution  of  thought  upon  which  we  are  now 
well  embarked,  the  accepted  basis  of  scientific  truth  which  experiment  should 
test  was,  in  large  measure,  whether  other  realms  of  matter  would  conform  to  the 
same  laws.  Such  realms  need  not  be  perceptible  to  the  senses.  Much  of  the  work 
of  the  later  Newtonian  physics,  of  course,  was  concerned  with  confirming  the 
validity  of  those  concepts  of  causality  and  truth  in  terms  of  molecular  phe- 
nomena, in  the  theories  of  gases  and  of  heat.  It  was  here  indeed  that  theory 
achieved  some  of  its  greatest  successes,  and  experiment  some  of  its  major 
triumphs  of  technique  and  ingenuity.  Through  the  greater  part  of  three  cen- 
turies, science  advanced  principally  through  a  succession  of  physical  observa- 
tions, the  building  of  hypotheses  to  connect  them,  the  elaborating  of  results 
predicted  from  those  hypotheses,  the  experimental  testing  of  these  results.  These 
processes  are  still  the  essential  building  stones  of  scientific  inquiry. 

But  in  our  own  immediate  day  the  revolutionary  developments  in  nuclear 
physics  and  in  relativity  have  made  it  clear  that  these  straightforward  pro- 
cedures and  the  older  logic  that  has  so  long  connected  them  cannot,  unaided, 
comprehend  the  whole  of  nature,  or  be  adequate  to  all  the  needs  of  science.  A 
door  has  been  opened  on  vistas  so  wide  and  unfamiliar  that  the  question  seri- 


ously  arises  as  to  how  far  a  revolutionary  science  of  the  future  can  be,  in  the 
older  conventional  sense,  strictly  empirical  or  rational.  Quite  possibly  the  ulti- 
mate task  of  science,  which  we  conceive  in  the  context  of  our  day  simply  as  the 
search  for  truth,  must  be  redefined  in  more  comprehensive  ways,  not  yet  clear. 
Perhaps,  following  a  suggestion  of  Martin  Johnson  compelling  most  sober  re- 
flection, we  will  come  to  think  of  scientific  truth  primarily  as  a  measure  of  the 
communicability  and  coherence  of  systems  of  thought,  as  a  property  of  inter- 
changeability  between  observers  on  the  one  hand,  and  of  scientific  propositions 
between  differing  situations  on  the  other.  If  the  old  definition  of  the  primary 
task  of  science  as  the  search  for  truth  is  to  retain  its  essential  meaning,  we  could 
then  well  say  that  the  highest  aim  of  scientific  inquiry  is  to  make  knowledge 
communicable  to  all  possible  situations,  and  that  knowledge  shall  be  judged  as 
"true"  by  the  degree  of  "coherence"  among  such  situations.  It  is  particularly 
noteworthy  that  an  essential  core  of  such  thinking,  the  concept  of  communica- 
tion and  communicability,  may  well  be  destined  to  become  one  of  the  central 
ideas  of  science  in  the  years  to  come,  coloring  all  its  concerns,  from  particle 
physics  to  the  sciences  of  life  and  mind. 

It  may  be  long  before  we  achieve  new  ground  with  the  certitude  with  which 
for  three  centuries  we  have  occupied  the  old.  But  as  we  struggle  to  do  so,  op- 
portunities— and  indeed  a  fierce  demand — for  wide  vision  and  penetrating 
originality  will  press  upon  us  to  a  degree  perhaps  unexcelled  in  our  experience. 
It  is  quite  conceivable  that  when  we  have  successfully  met  these  new  challenges, 
and  embraced  the  opportunities  that  they  will  surely  bring,  we  shall  have  com- 
pleted an  intellectual  revolution  no  less  profound  than  that  which  fashioned 
the  modern  from  the  medieval  mind.  Surely  there  has  never  been  a  time  more 
rich  in  scientific  opportunity. 

At  first  glance,  it  seems  curious  that  in  an  age  which  thus  offers  unparalleled 
challenges  to  break  new  ground,  in  an  age  which  probably  presents  the  richest 
fabric  of  novel  enterprise  in  all  human  history,  truly  great  innovation  should 
still  be  a  comparatively  rare  event.  Even  more  striking  is  the  rare  and  precious 
quality  of  the  innovator  himself,  relative  to  the  comparatively  huge  and  ever- 
growing number  of  fine  minds  devoted  to  the  scientific  way.  It  is  hard  to  escape 
the  impression  that  the  opportunities  for  original  advance  have  not  been  as 
fully  grasped  as  they  well  might  be. 

This  rarity  of  innovation,  and  innovators,  of  course,  is  as  old  as  man's  experi- 
ence. We  are  so  accustomed  to  the  notion  of  continuous  progress  in  our  modern 
world  that  we  find  it  hard  to  comprehend  that  real  progress  has  been  the 
shining  event,  not  the  rule,  of  human  history.  The  astonishing  uniformity  of 
some  primitive  artifacts  used  by  early  peoples  over  wide  regions  of  the  globe 
suggests  how  far  each  single  original  improvement  of  technique  may  have 
diffused  by  copying,  and  how  influential  it  may  have  been  in  the  advance  of 


all  mankind.  There  is  suggestive  evidence  that  many  elements  which  formed 
the  material  basis  of  the  early  civilizations  of  the  Old  World — and  so  provided 
the  very  foundation  upon  which  our  own  culture  was  erected — trace  their 
origins  from  a  limited  region  of  the  Middle  East.  How  wide  the  influence  of 
a  single  cultural  innovation,  once  made,  can  be,  and  how  rapidly  it  can  make 
itself  felt,  are  vividly  suggested  in  historic  times  by  the  speed  of  the  spread  of  the 
culture  of  the  horse  among  the  plains  Indians  of  the  Southwest,  or  by  the  even 
more  explosive  propagation  of  the  culture  and  use  of  the  domestic  fowl  along 
the  basin  of  the  Amazon.  It  is  strikingly  attested  in  our  own  day  in  the  rapidity 
and  facility  with  which  technological,  as  opposed  to  cultural  and  conceptual, 
innovations  may  be  accepted  and  take  root  and  flourish  among  societies  other- 
wise of  the  most  divergent  background. 

One  reason  for  the  historical  rarity  of  great  conceptual  innovation  in  more 
recent  times  undoubtedly  lies  in  the  difficulty  of  its  communication — not  so 
much  in  its  failure  to  be  recognized  as  in  the  misconstruction  of  its  message. 
The  followers  of  Darwin,  after  his  death,  introduced  all  sorts  of  implications 
extraneous  to  his  theory.  They  applied  it  to  areas  where  it  was  quite  inappro- 
priate, which  he  was  especially  careful  to  avoid,  generating  misconceptions 
that  have  required  two  generations  of  intensive  work  to  clear  away.  The  dis- 
coveries of  Gregor  Mendel,  the  monk  of  Altbriinn,  whose  epoch-making  ex- 
periments in  the  heredity  of  garden  peas  and  the  conclusions  he  drew  from 
them  were  models  of  precision  in  measurement  and  clarity  of  reasoning,  set 
the  whole  stage  for  modern  genetics.  But  they  were  of  necessity  published  in 
a  local  journal,  obscure  and  little  read,  and  so  were  lost  for  almost  thirty-five 
years.  When  they  were  finally  unearthed  in  1900,  Mendel  was  already  dead, 
and  those  who  found  his  paper  rediscovered  only  a  part  of  his  conception.  They 
saw  in  his  work,  not  his  vision  of  a  wide  and  grand  design,  but  rather  much 
narrower  and  more  specific  findings,  to  be  tested  by  their  own  kinds  of  experi- 
ments. For  this  failure  of  communication,  the  whole  science  of  genetics  suf- 
fered for  many  years. 

The  intrinsic  difficulty  of  great  innovation  and  the  numerous  impediments 
to  its  communication  are  surely  but  two  of  the  reasons  for  its  comparative  in- 
frequency.  An  important  and  perhaps  an  ultimately  limiting  factor,  of  course, 
is  the  inherent  rarity  of  the  deeply  innovating  mind.  Perhaps  we  must  accept 
as  given  the  fact  that  in  any  human  population  there  will  be  a  low  proportion 
of  truly  original  figures.  But  we  are  far  from  certain  just  how  low  that  propor- 
tion needs  to  be  under  the  best  conditions.  The  nature  of  those  conditions 
demands  the  most  serious  consideration. 

It  is  a  striking  paradox  that,  though  our  contemporary  culture  affords  op- 
portunities for  creativeness  probably  unequaled  in  human  history,  it  also  poses 
hazards  that  may  likewise  be  unequaled.  One  important  kind  of  hazard  is  in- 
herent in  the  very  structure  of  thought  and  philosophy  which  may  crystallize 


about  a  field  of  inquiry.  Such  well  formed  systems,  when  rigidly  and  tena- 
ciously held,  may  constitute  more  powerful  and  dangerous  barriers  to  the  ad- 
vance of  concept  and  experiment  alike  than  we  ordinarily  recognize.  The 
more  elaborate  the  pattern,  the  more  adequate,  the  more  satisfactory,  the  more 
self-consistent  it  seems,  the  more  difficult  it  is  to  transcend  and  the  greater  may 
be  the  resistance  to  truly  original  departures  from  it.  This  paradoxical  balance 
between  the  older  thinking  and  the  new  is  the  more  delicate  because  it  seems 
inevitable  that,  although  all  new  scientific  concepts  must  transcend  the  older 
matrix  from  which  they  arise,  if  they  are  to  become  established  and  permanent 
they  must  also  find  firm  links  with  it. 

This  dual  relationship  between  the  new  advance  and  the  older  structure  of 
knowledge  and  reasoning  from  which  it  springs  may  be  significant  in  shaping 
the  history  of  human  thought  so  characteristically  to  long  interludes  of  refine- 
ment and  consolidation,  punctuated  by  shorter  periods  of  explosive  revolution. 
The  record  is  replete  with  examples  of  the  power  and  persistence  of  a  great 
central  idea  and  the  elaborate  ways  in  which  new  and  potentially  revolutionary 
thought,  for  better  or  worse,  may  long  be  organized  around  it  or  accommodated 
within  it.  Through  a  whole  age,  truth  in  observation  for  the  medieval  mind 
was  measured  by  the  success  with  which  impressions  of  the  world  could  be 
selected  and  organized  about  current  religious  images.  What  we  would  call 
symbolic  meaning  was  for  that  age,  in  large  measure,  its  fundamental  reality. 
Research  in  its  modern  sense  was  retarded  for  at  least  a  hundred  years  by  the 
belief,  still  powerful  in  the  days  of  colonial  New  England,  that  all  knowledge 
about  the  world  is  given  in  advance  of  an  investigation,  is  hidden  within  it  like 
the  imaginary  statue  in  a  block  of  marble,  awaiting  only  the  master's  chisel; 
that  the  most  that  right  reasoning  can  possibly  do  is  to  make  it  stand  forth 
more  clearly  and  brilliantly  revealed;  that  there  is  therefore  nothing  essentially 
creative  in  the  whole  process  of  investigation  of  nature.  It  required  a  major 
revolution  to  reach  the  view  that  the  investigation  itself  might  determine  the 
concepts  which  would  follow  it,  and  that  reality  might  be  more  closely  oriented 
toward  the  primary  evidence  of  our  senses.  It  was  a  revolution  vividly  symbol- 
ized, as  Lewis  Mumford  has  pointed  out,  by  the  general  introduction  into  men's 
homes  of  clear  glass  windows  letting  in  a  wide  and  literal  view  to  replace  the 
great  stained  glasses  of  the  medieval  world,  with  their  brilliant  slanting  shafts  of 
colored  light,  mirroring  symbolic  images  already  fixed  and  finally  conceived. 

The  revolution  that  replaced  the  medieval  by  the  essentially  modern  out- 
look was  comparatively  swift  and  violent,  once  the  old  containing  vessels  had 
been  broken.  But  the  wonder,  perhaps,  is  not  so  much  in  their  fragility  as  in 
their  elastic  and  confining  durability.  Other,  similar  vessels  of  concept, 
fashioned  at  almost  as  remote  a  time,  only  now  are  beginning  to  show  the  fine 
cracks  that  in  our  own  generation  may  presage  another  revolution  as  great. 

There  can  be  no  more  vivid  examples  of  the  tremendous  power  of  a  central 


idea  in  advancing  or  retarding  innovation  than  those  offered  by  the  sciences  of 
life  and  of  mind,  in  our  generation  in  ferment  such  as  they  have  not  known 
since  the  days  of  Galen  or  of  Leeuwenhoek.  The  distinction  between  the  "liv- 
ing" and  the  "not-alive"  as  two  quite  different  and  opposite  categories  is  far 
older  than  even  medieval  science.  From  the  days  of  Harvey  it  was  refined  even 
further,  until,  for  almost-modern  biology,  it  was  not  considered  difficult  in 
practice  to  characterize  a  newly  discovered  object  as  living  or  nonliving.  Into 
our  own  time  the  life  sciences  almost  unconsciously  retained  as  a  central  con- 
cept a  notion  that  actually  had  its  greatest  currency  in  sixteenth-century  think- 
ing. Implicitly,  if  not  explicitly,  the  organism  was  commonly  endowed  with 
the  "property"  of  life  as  a  unique,  inhabiting  quality  much  as  the  alchemist  en- 
dowed the  burning  body  with  phlogiston,  or  the  early  physicist  the  bar  with 
weight  and  color,  length  and  hardness,  or  the  physician  the  bitter  pill  with 

Perhaps  what  did  as  much  as  anything  to  shake  the  central  idea  of  life  as  a 
unique  "inhabiting"  property  was  the  preparation  of  the  "crystallizable"  plant 
viruses  a  quarter  of  a  century  ago  and  the  discovery  of  their  long  durability  upon 
the  shelf,  so  like  a  "lifeless"  chemical,  combined  with  the  power  to  resume  the 
lifelike  properties  of  growth  and  reproduction  after  long  periods  when  re- 
inoculated  to  a  suitable  host.  With  that  step  taken,  with  the  recognition  that 
the  "living"  and  the  "nonliving"  may  in  some  contexts  at  least  differ  in  degree 
rather  than  in  kind,  the  advances  of  concept  and  experiment  that  now  are  in 
full  tide  crowded  upon  one  another.  Recent  studies  of  the  dynamics  of  me- 
tabolism, especially  in  microorganisms,  have  helped  to  replace  the  older  notions 
of  the  structures  of  life  as  gross  and  essentially  static  morphologies,  like  cell 
walls  and  semipermeable  membranes,  with  those  far  more  exciting  concepts 
of  precise  structural  lattices  and  highly  specific  sites  of  adsorption  which  in  the 
last  years  have  become  dominant.  Radical  experiments  in  the  synthesis  of 
amino  acids  under  conditions  approximating  those  believed  to  have  obtained  on 
the  lifeless  earth,  and  illuminating  researches  in  paleobiochemistry,  have  made 
common  currency  of  the  notion  that  living  things  may  have  indeed  originated 
terrestrially,  and  have  further  obscured  the  division  between  the  living  and  the 
nonliving.  Quantitative  biochemical  genetics,  the  fine  analysis  of  the  structure 
and  function  of  the  chromosome  that  is  proceeding  apace  today,  with  its  ac- 
companying new  insights  into  the  mechanisms  of  replication  and  its  demon- 
strations of  transduction  and  transformation,  further  reinforce  the  new  ideas 
of  biological  systems  as  wonderfully  complex  dynamic  structures  of  exquisitely 
detailed  precision  in  both  space  and  time.  Some  modern  investigators  of  the 
structure  of  the  chromosome  see  in  it  a  precisely  designed  mechanism  for  the 
transfer  of  information,  as  truly  an  instrument  of  communication  as  any  de- 
signed by  man,  and  far  more  intricate  in  function.  Indeed,  it  seems  quite  possible 
that  in  our  time  some  of  the  new  central  ideas  about  living  matter,  whether  as 


individual  organism,  as  structured  population,  or  as  the  changing  product  of 
evolution,  will  describe  it  primarily  in  terms  of  self-regulating,  goal-directed 
systems  to  which  modern  concepts  of  information  theory  and  information 
flow  can  profitably  be  applied.  It  is  hard  to  realize  the  swift  and  tremendous 
strides  of  thirty  years  which  the  escape  from  too  exclusive  a  preoccupation  with 
an  old  idea  has  brought. 

A  similar  revolution  is  beginning  in  the  science  of  the  mind,  again  consequent 
upon  the  abandoning  of  an  older  concept,  this  time  even  more  confining. 
Here,  to  the  notion  of  life  as  a  "property"  of  living  matter  was  added  a  further 
barrier — the  central  idea  of  the  mind  as  fundamentally  different  in  kind  even 
from  the  living  body,  the  res  cogitans  as  distinct  from  the  res  extensa — inherited 
from  the  brilliant  eighteenth-century  advocacy  of  Descartes.  It  is  only  in  the 
last  few  years  that  mind,  too,  has  widely  come  to  be  regarded  as  a  self-regulated, 
goal-directed  dynamic  system,  structured,  like  life  itself,  with  exquisite  precision 
and  almost  unimaginable  complexity  in  both  space  and  time,  but  not  necessarily 
differing  in  essence,  not  necessarily  wholly  beyond  the  modes  of  investigation 
already  available  to  us.  It  is  only  now  that  the  central  idea  of  communication 
which  has  become  so  prominent  in  the  life  sciences — and  indeed  may  be  so 
important  to  the  whole  structure  of  science  itself — has  been  extended  to  the 
subject  of  mind  as  well;  that  mental  processes  in  some  of  their  aspects  are  be- 
ginning to  be  translated,  like  the  processes  of  heredity,  into  terms  of  communi- 
cation theory  and  information  flow.  We  can  only  speculate  today  what  the 
consequences  may  be.  As  with  the  life  sciences,  experimental  evidence  is  ac- 
cumulating along  the  novel  paths  of  thought  that  escape  from  the  older  pivotal 
notions  has  made  possible,  evidence  provided  from  such  diverse  sources  as 
studies  of  therapeutic  neurochemicals  and  of  electronic  computing  systems. 

No  more  vivid  historical  examples,  perhaps,  could  be  cited  of  the  long  delays 
to  basically  new  departures  that  too  static  and  finished,  too  self-contained  and 
self-consistent  systems  of  thought  may  impose  than  these  sciences  of  life  and 

There  are  other  barriers  of  a  more  practical  sort.  One  of  the  profound  ways 
in  which  the  modern  world  may  differ  from  other  centuries  is  in  its  conscious 
recognition  of,  and  its  high  respect  for,  the  fact — but  not  always  the  process — of 
innovation.  Recognizing  its  power,  our  own  society  would  like  to  hurry  it 
along,  to  enlarge  its  scale,  to  organize  it  more  efficiently.  Impressed  as  we  are 
by  the  extent  to  which  the  processes  of  development  built  upon  innovation 
can  be  accelerated  by  organization,  we  are  greatly  tempted  to  extend  this 
thinking  to  creative  matters,  to  believe  that  we  can  organize  for  great  innova- 
tion too.  Such  misconceptions  are  peculiarly  liable  to  occur  in  massive  and 
highly  organized  environments  devoted  to  research  and  development.  They 
are  inherent  in  the  all-too-prevalent  confusion  between  the  processes  of  scientific 


research  itself  and  of  development,  in  the  frequently  uncritical  transfer  of  ideas 
and  techniques  from  development  and  production  into  the  field  of  scientific 
innovation.  Among  them,  too  often,  is  the  destructive  concept  that  the  efTective- 
ness  of  innovation,  like  that  of  production,  can  be  doubled  simply  by  doubling 
the  size  of  the  accompanying  effort.  Some  departments  of  research  and  de- 
velopment in  industry  and  in  government  are  particularly  vulnerable,  and  the 
academic  world  is  by  no  means  immune. 

But  we  should  be  well  warned.  It  is  deeply  significant  that,  in  our  own  time 
of  unparalleled  technical  advance,  in  a  day  when  engineering  achievements 
built  about  new  principles  can  accomplish  unbelievable  things,  the  great  in- 
novations are  still  highly  individual,  are  still  associated,  at  their  inception,  with 
a  mere  handful  of  names. 

The  comparative  rarity  of  the  environments  most  favorable  to  truly  original 
investigation,  their  unusual  and  special  character,  and  the  frequent  practical 
impediments  to  their  best  development  may  impose  sharper  limitations  on  the 
creative  potential  of  our  nation  than  we  are  inclined  to  recognize.  Climates  of 
research  must  comprehend  many  and  varied  elements,  some  of  which  are  rarely 
joined  in  more  ordinary  situations.  Important  among  them  are  understanding 
and  protection.  Since  innovation  is  and  must  always  remain  a  uniquely  indi- 
vidual experience,  it  poses  unusual  social  dangers  for  the  investigator  from 
which  he  must  be  adequately  shielded — the  animosity  which  the  new,  with  its 
threat  to  current  things,  must  often  provoke.  Such  perils  may  not  be  fatal  to 
the  advance  once  it  is  fairly  made.  For  if  the  advance  is  truly  novel  and  im- 
portant, and  if  the  society  in  which  it  occurs  is  self-confident,  it  is  likely  to  be 
eventually  accepted  and  incorporated,  and  may  even  become  symbolic.  The 
greater  risk  is  that  the  prospect  of  such  hardships  may  inhibit  the  individual 
capable  of  great  innovation  from  undertaking  it  or  from  fully  developing  its 
results.  Such  considerations  of  the  hazards  opposing  the  unexampled  richness 
of  opportunity  for  scientific  creativeness  in  our  society  bring  home  anew  the 
gravity  of  our  responsibility  to  analyze  most  carefully  this  climate  in  which  true 
scientific  originality  best  flourishes  and  to  see  that  environments  embodying 
it  are  adequately  provided  and  conserved. 

It  is  not  hard  to  set  down  the  major  tasks  that  institutions  assuming  these 
great  responsibilities  must  perform.  The  most  important,  of  course,  must  be 
the  discovery,  the  encouragement,  the  adequate  training  of  the  outstandingly 
original  investigator,  and  the  fostering  of  his  work.  Such  institutions  must 
be  unusually  sensitive  to  the  rare,  natively  innovating  mind,  and  unusually 
adept  in  its  discovery.  They  must  have  in  high  degree  the  ability  to  promote 
and  develop  the  peculiar  drive,  the  special  verve  and  flare,  the  fine  sense  of  style, 
which  great  innovation  inherently  requires.   They  must  be  able  to  train  the 


innovator  of  exceptional  promise  in  the  company  of  other  gifted  investigators 
already  mature  and  at  the  height  of  their  powers. 

Above  all,  the  environments  that  such  institutions  create  must  be  unconven- 
tional and  malleable.  Their  responsibility  is  not  only  to  provide  for  the  very 
different  requirements  of  different  workers,  based  in  divergences  of  tempera- 
ment, of  method,  of  timing.  It  is  also  to  meet  the  very  different  requirements 
of  the  same  investigator  at  various  periods  of  his  working  life.  There  must 
be  ready,  yet  penetrating,  astringent  tolerance — and  indeed  welcome  and  en- 
couragement— for  the  long  preliminary  phase  of  apparently  rambling  and 
aimless  effort  that  so  often  and  so  characteristically  precedes  the  great  individual 
discovery  or  the  brilliant  innovational  career,  while  the  ground  is  being  can- 
vassed and  a  wealth  of  divergent  but  relevant  material  is  being  mastered  and 
consolidated.  There  must  be  provision  for  the  long  intervals  of  seclusion  and 
introversion  that  may  be  essential  during  periods  of  highest  creativity,  when 
white-hot  effort  will  not  brook  interruption.  There  must  be  ample  understand- 
ing and  ample  capacity  to  shield  the  worker  in  this  his  most  vulnerable  time 
from  the  insistent  demands  that  will  surely  come  to  prematurely  formulate 
results  and  conclusions  only  half-attained.  Conversely,  there  must  also  be  ade- 
quate provision  for  compensating  periods  of  extroversion,  when  the  immediate 
barrier  is  surmounted  and  the  conclusions  are  consolidated  and  the  extensive 
communication  that  must  accompany  or  follow  creativity  is  encouraged  and 

Institutions  for  creative  research  must  be  able  to  set  and  apply  subtle  standards 
of  excellence.  They  must  be  able  to  judge  and  reward  high  standards  by 
criteria  more  searching  and  more  true  than  those  too  often  and  too  super- 
ficially accepted.  They  must  provide  teachers  of  research  of  the  very  highest 
quality,  and  in  high  ratio  to  those  they  train — indeed,  those  primarily  in  train- 
ing as  investigators  should,  ideally,  be  in  the  minority.  And  they  have  one 
final  and  supremely  important  responsibility — the  task  of  communicating  the 
results  of  research  rapidly  and  effectively  to  workers  in  fields  that  may  seem  far 
remote,  yet  may  actually  be  contiguous  to  those  in  which  the  discoveries  have 
been  made.  Such  institutions  will  benefit  greatly  by  being  relatively  small,  and 
will  almost  certainly  benefit,  too,  if  they  are  somewhat  decentralized,  both 
physically  and  in  the  subject  matter  with  which  they  deal. 

The  discovery  and  training  of  new  generations  of  investigators  in  science 
are,  of  course,  basically  functions  of  the  universities.  Theirs  must  be  a  grave 
portion  of  the  responsibility.  But  they  are  inevitably  hindered  in  this  special 
task  by  many  of  the  current  pressures  that  bear  especially  heavily  upon  them: 
the  pressures  to  large  size,  to  an  ever-lessening  ratio  of  teachers  to  students,  to  a 
lowered  quality  in  the  teaching  staff  itself,  valiantly  as  this  is — and  must  always 
be — resisted.  Moreover,  the  whole  mission  of  the  universities  is  of  course  far 
broader.  The  training  of  great  innovators,  and,  even  more  specifically,  of  great 


innovators  in  the  scientific  way,  must  always  remain  one  of  their  most  important 
challenges.  But  it  is,  of  necessity,  a  somewhat  narrow  sector  in  the  total  con- 
text of  their  work.  They  need  assistance. 

Such  circumstances  place  renewed  emphasis  in  our  day  on  the  great  and 
continuing  importance  of  the  research  institute  in  our  national  life.  Its  task  is 
not  only  that  of  scientific  innovation  and  discovery.  This  indeed  is  important 
enough.  But,  far  more  important,  to  it  must  fall  the  essential  function  of  sym- 
bolizing in  the  best  and  purest  form  the  way  of  scientific  innovation  with  all 
that  it  means  for  our  nation  and  our  culture — undismayed  and  undeterred 
by  pressure  or  by  hazard.  To  it  belongs  the  great  responsibility  of  discovering 
and  fostering  new  minds  of  high  promise,  of  selecting  them  with  discernment, 
of  affording  them  a  rare  training  as  investigators,  and  then  of  seeing  to  it  that 
their  careers  can  be  fulfilled  where  their  talents  are  most  effective.  Far  beyond 
the  task  of  research  itself  lie  the  challenge  and  the  responsibility  to  cultivate 
new  minds  through  all  the  possible  and  relevant  means — symposia  and  ex- 
change fellowships,  the  flow  to  university  and  industry  and  back  again,  all  the 
particular  media  of  publication. 

More  and  more,  in  our  time,  it  is  recognized  that  the  true  wealth  of  a  nation 
resides  not  so  much  in  the  volume  and  the  variety  of  goods  that  it  can  produce, 
not  even  so  much  in  its  total  resources  of  production,  but  rather  in  two  human 
factors  more  subtle,  more  powerful,  more  determining:  in  the  quality  and  excel- 
lence of  its  people  as  a  whole,  and  in  their  capacity  for  innovation.  In  the  guard- 
ing, the  fostering,  the  building  of  this  last  element  of  national  wealth  lies  the 
ultimate  challenge  to  such  a  research  organization  as  the  Carnegie  Institution 
of  Washington.  Never  in  the  history  of  science  in  our  nation  has  that  challenge 
been  more  exciting,  that  responsibility  more  keenly  privileged. 



The  past  year  has  been  an  unusually  full  and  active  one  in  many  facets  of 
the  work  of  the  Carnegie  Institution.  It  has  been  notable  for  the  number  and 
variety  of  investigations  going  forward,  both  those  newly  initiated  and  those 
continuing  from  other  years.  In  consequence,  it  is  even  more  difficult  than  usual 
to  select  individual  projects  for  this  review  on  other  than  a  largely  arbitrary 
basis.  The  ones  described  are  intended  only  as  representative  examples  of  the 
kind  of  work  in  progress  through  the  year.  They  may  not  be  more  or  less  im- 
portant, more  or  less  striking,  than  others  that  might  equally  well  have  been 
included.  The  reader  interested  in  following  the  programs  of  the  various  de- 
partments in  greater  detail  should  turn  to  their  individual  and  more  complete 
accounts,  which  follow. 

This  year  is  noteworthy  in  other  respects.  It  marks  the  tenth  anniversary  of 
the  dedication  of  the  200-inch  Hale  telescope  on  Palomar  Mountain  and  the 
tenth  anniversary  of  the  agreement  for  the  joint  operation  of  the  Mount  Wilson 
and  Palomar  Observatories  by  the  Carnegie  Institution  of  Washington  and  the 
California  Institute  of  Technology.  It  also  marks  the  termination  of  the  work 
of  the  Department  of  Archaeology  (though  not  of  all  the  work  of  the  Institu- 
tion in  that  field),  bringing  to  a  close  a  record  of  distinguished  pioneering 
achievement  in  the  history  of  the  Maya  in  Middle  America.  It  has  been  a 
chapter  of  exploration  not  only  extremely  productive  in  its  own  right,  but  one 
that  largely  shaped  the  standards  and  set  the  stage  for  all  the  following  in- 
vestigations in  that  field.  It  seems  particularly  appropriate  in  this  report,  there- 
fore, to  present  a  brief  review  of  the  past  ten  years  of  the  Mount  Wilson  and 
Palomar  Observatories,  and  of  the  entire  program  of  the  Department  of 

Ten  Years  of  Mount  Wilson  and  Palomar  Observatories 

The  ten  years  since  the  Hale  telescope  came  into  operation  comprise  an 
exceedingly  notable  decade  in  astronomy,  marked  by  very  important  advances 
in  knowledge.  Two  such  advances  demand  special  notice  because  of  their  gen- 
eral and  fundamental  character.  The  first  relates  to  the  changes  that  have  come 
in  our  ideas  of  the  size  of  the  observable  universe,  the  magnitudes  of  the 
distances  within  it,  and  its  age.  The  second  concerns  the  equally  striking  gains 
in  our  knowledge  of  the  ways  in  which  stellar  evolution  occurred  in  the  past 
and  is  currently  taking  place. 

Between  the  two  world  wars  Edwin  Hubble  made  an  extensive  series  of 
measurements  of  the  distances,  diameters,  and  luminosities  of  the  galaxies  with 
the  100-inch  telescope  on  Mount  Wilson.   His  findings  provided  definite  evi- 


dence  that  the  galaxies  are  huge  systems  of  stars,  similar  to  our  Milky  Way, 
and  that  they  form  the  major  units  in  which  the  mass  of  the  universe  is  dis- 
tributed. When  the  Hale  telescope  became  available,  a  thorough  restudy  of 
the  near-by  galaxies  was  a  major  project.  It  was  planned  in  particular  that  the 
procedures  used  in  fixing  galactic  distances  be  re-examined,  to  reduce  or 
eliminate  the  large  uncertainties  known  to  exist  in  the  initial  measurements. 
In  general,  such  procedures  involve  the  comparison  of  the  apparent  brightness 
of  some  specific  distance  indicator  in  the  galaxy,  such  as  a  star,  with  its  bright- 
ness or  luminosity  as  determined  from  near-by  examples  taken  from  our  own 
Galaxy.  The  required  answer  is  obtained  by  applying  the  inverse  square  law 
for  the  falling  off  of  apparent  brightness  with  distance.  To  be  useful  as  a 
distance  indicator,  the  star  or  other  object  must  be  unusually  bright  and  must 
have  some  special  characteristic,  to  permit  its  identification  as  a  particular  type 
with  the  same  brightness  as  examples  from  our  own  system.  Cepheid  variables 
were  the  classic  distance  indicators  used  by  Hubble.  They  are  particularly  con- 
venient because  of  the  ready  identification  made  possible  by  their  light-fluctua- 

In  1952  the  conclusion  was  reached  that  the  classic  cepheid  variables  are 
about  1.5  magnitudes  brighter  than  had  been  thought  from  earlier  work.  In- 
dependent observations  by  Walter  Baade  with  the  Hale  telescope,  and  by  A.  D. 
Thackeray  and  A.  J.  Wesselink  on  the  Magellanic  Clouds  at  Pretoria,  together 
with  theoretical  studies  by  Henri  Mineur  and  by  Adriaan  Blaauw  and  H.  R. 
Morgan  of  proper  motions  of  cepheids  in  our  own  Galaxy,  all  pointed  in  this 
direction.  A  careful  remeasurement  was  therefore  undertaken  of  the  apparent 
magnitudes  of  a  substantial  number  of  cepheids  in  several  of  the  nearer  galaxies. 

The  absolute  luminosity  of  a  cepheid  variable  is  a  function  of  its  period. 
Several  dozen  plates  must  therefore  be  made  of  each  field  of  view,  distributed 
over  a  sufficient  time  so  that  a  reliable  light-curve  can  be  obtained  of  each 
variable.  Sets  of  plates  were  obtained  for  this  purpose  by  Baade  for  four  fields 
in  the  Andromeda  galaxy  and  for  several  other  galaxies  in  the  local  group. 

While  this  program  of  remeasurement  of  apparent  magnitude  was  in  prog- 
ress, another  approach  to  the  same  question  was  being  explored  simultaneously 
— the  investigation  of  properties  of  other  unusually  bright  objects  that  might 
be  useful  as  distance  indicators.  Hubble  and  Allan  R.  Sandage  measured  the 
brightest  stars  in  the  Andromeda  galaxy.  Sandage  investigated  the  integrated 
magnitude  of  the  gaseous  nebula  or  H  II  regions  surrounding  several  of  these 
very  luminous  stars.  Arthur  D.  Code  and  T.  E.  Houck  compared  the  excep- 
tionally bright  blue  stars  in  NGC  6822,  M  33,  the  Large  Magellanic  Cloud,  and 
our  own  Galaxy.  Halton  C.  Arp  made  a  very  extensive  study  of  the  magni- 
tudes, light-curves,  and  frequency  of  occurrence  of  ordinary  novae  in  the 
Andromeda  galaxy,  photographing  fields  in  that  galaxy  with  the  60-inch  tele- 


scope  on  every  clear,  moonless  night  for  two  observing  seasons.  Thirty  novae 
were  found  during  this  period.  Measurements  of  the  distribution  in  magnitude 
of  the  globular  clusters  in  the  Andromeda  galaxy  were  made  by  William  A. 
Baum.  Milton  L.  Humason  and  Sandage  made  observations  of  the  red  super- 
giants  in  M  33.  Hubble,  Sandage,  and  Baade  observed  various  distance  indica- 
tors in  the  more  distant  galaxies  outside  the  local  group. 

Lastly,  it  was  necessary  to  recalibrate  the  magnitudes  of  the  stars  in  specific 
Selected  Areas  that  had  been  used  as  standards  of  magnitude  for  the  very  faint 
stars  in  these  galaxies.  Such  a  recalibration,  using  photoelectric  techniques, 
was  started  by  Dr.  Joel  Stebbins  and  Dr.  A.  E.  Whitford  of  the  Washburn  Ob- 
servatory, using  the  100-inch  telescope;  it  was  completed  by  Baum  with  the  Hale 
instrument.  It  developed  that  the  earlier  recorded  magnitudes  of  very  faint  stars, 
photographically  determined,  were  too  bright  by  as  much  as  one  magnitude. 

When  various  corrections  were  made,  Baade  calculated  in  1952  that  the 
Andromeda  galaxy  is  2  million  light-years  distant — nearly  three  times  as  far 
away  as  the  previously  accepted  value.  By  the  same  token,  its  diameter  and 
mass  must  be  increased  by  a  similar  factor,  and  its  luminosity  by  a  factor  of 
about  8.  The  application  of  the  new  values  of  the  magnitudes  of  various  distance 
indicators  to  the  preliminary  observations  of  more  distant  galaxies  points  to  an 
even  larger  correction  to  the  distances  and  dimensions  of  these  objects.  Thus 
Sandage,  in  a  preliminary  study,  during  the  current  year,  of  the  distances  of  the 
farthest  galaxies  in  which  individual  distance  indicators  can  be  observed,  esti- 
mates that  the  old  distances  of  the  Virgo  cluster  and  all  more  remote  galaxies 
must  be  increased  by  a  factor  of  between  5  and  10  over  the  values  obtained 
in  the  1930's.  If  confirmed  by  further  measurements,  this  leads  to  the  striking 
conclusion  that  estimates  of  the  diameter  of  the  observable  universe  must  also 
be  increased  by  a  factor  of  5  to  10.  Finally,  if  the  length  of  time  since  the  begin- 
ning of  the  expansion  of  the  universe  as  indicated  by  the  redshift  is  taken  as 
the  age  of  the  universe,  this  age  is  increased  by  approximately  the  same  factor, 
that  is  to  between  7  and  13  billion  years — a  revision  that  brings  the  age  into 
better  agreement  with  other  age  determinations,  such  as  that  of  the  solar  system. 

During  World  War  II  Baade  took  advantage  of  the  unusually  good  observ- 
ing conditions  afforded  by  the  widespread  black-out  in  the  Los  Angeles  area  to 
reinvestigate  the  types  of  stars  in  the  nucleus  and  in  the  spiral  arms  of  the 
Andromeda  galaxy  and  in  the  elliptical  galaxies,  using  the  Mount  Wilson  tele- 
scope. He  found  that  the  most  conspicuous  stars  in  the  galactic  nuclei  were  red 
giants,  while  in  the  spiral  arms  of  Andromeda  and  in  the  neighborhood  of  the 
sun  in  our  own  Galaxy  the  brightest  stars  were  blue.  From  these  observations 
the  concept  of  differing  population  types  of  stars  arose,  the  stars  in  our  own 
neighborhood  being  designated  as  Population  I  and  those  in  the  galactic 
nuclei  and  in  the  elliptical  galaxies  as  Population  II. 

An  extensive  study  of  the  properties  of  the  stars  in  the  two  populations  was 


instituted.  Measurements  of  the  color-magnitude  relations  were  made  for 
various  groups  of  stars,  including  several  globular  clusters  and  galactic  clusters, 
observed  by  Arp,  Baum,  Donald  E.  Osterbrock,  Sandage,  Maarten  Schmidt,  and 
Merle  F.  Walker,  and  including  a  few  members  of  the  local  group  of  galaxies 
investigated  by  Baade  and  Sandage.  Studies  were  also  undertaken  of  the 
period-luminosity  function  of  cepheid  and  cluster-type  variables  in  these  objects. 

From  these  studies  the  picture  soon  emerged  that  Population  I  is  char- 
acterized by  very  young  stars,  only  recently  condensed  from  clouds  of  dust 
and  gas.  Population  II,  on  the  other  hand,  contains  only  old  stars.  Parallel 
theoretical  and  laboratory  developments,  showing  that  the  energy  radiated  by 
the  stars  comes  from  the  transformation  of  hydrogen  into  helium  and  heavier 
elements  in  the  hot  stellar  core,  confirmed  this  picture.  Detailed  analyses  of 
the  nuclear  processes  occurring  in  these  stars  were  developed  by  Fred  Hoyle 
and  G.  R.  Burbidge  in  collaboration  with  Dr.  W.  A.  Fowler  and  Dr.  E.  M. 
Burbidge  of  the  Kellogg  Radiation  Laboratory.  The  effect  of  the  depletion  of 
the  hydrogen  fuel  on  the  characteristics  of  a  star  was  investigated  theoretically 
by  Hoyle  and  by  Dr.  Martin  Schwarzschild  and  others. 

The  picture  of  stellar  evolution  that  emerges  is,  in  general,  as  follows.  When 
a  mass  of  gas  consisting  predominantly  of  hydrogen  condenses  into  a  star  on 
what  is  called  the  main  sequence,  its  luminosity  and  surface  temperature  on 
this  sequence  are  fixed  by  the  mass  condensed,  the  luminosity  varying  approxi- 
mately as  the  cube  of  the  mass,  the  surface  temperature  increasing  slowly  but 
continuously  with  the  mass.  Rapid  expenditure  of  their  hydrogen  fuel  by  the 
very  massive  stars  soon  exhausts  it.  As  the  fuel  becomes  depleted,  the  star  ex- 
pands. Its  surface  cools  while  its  total  radiation  is  increased.  At  this  critical 
stage  the  star  leaves  the  main  sequence.  As  the  fuel  is  finally  exhausted  the 
radiation  rapidly  falls  off  and  the  star  probably  ends  its  career  as  a  faint  white 

On  this  picture,  a  young  group  of  stars  is  characterized  by  a  continuous  pro- 
gression of  properties  from  massive,  extremely  luminous  blue  stars  to  the 
reddish  faint  dwarfs.  As  the  group  ages,  the  very  luminous  blue  stars  exhaust 
their  fuel  and  drop  out.  The  brightest  stars  in  such  a  group  are  those  of  moder- 
ate initial  luminosity  which  with  approaching  fuel  depletion  have  expanded 
and  cooled  off,  but  have  temporarily  become  more  luminous.  Such  a  stellar 
group,  a  few  billion  years  old,  should  have  the  distribution  in  luminosity  and 
color  found  in  fact  to  exist  among  the  globular  clusters  and  the  elliptical 

The  discovery  of  the  role  played  by  nuclear  transformations  in  the  evolution 
and  energy  production  of  the  stars  gave  an  added  interest  to  spectroscopic 
studies  of  the  abundances  of  the  elements  participating  in  these  transformations 
in  stars  of  various  ages.  The  great  light-gathering  power  of  the  Hale  telescope 
and  the  unique  efficiency  of  its  spectrographic  equipment  have  made  possible 


the  intensive  study  of  faint  stars,  such  as  the  globular-cluster  stars  and  the 
white  dwarfs,  of  special  significance  in  evolutionary  theory. 

Extensive  studies  of  the  properties  of  the  stars  in  globular  and  galactic  clusters 
have  been  made  during  the  current  year.  All  the  stars  of  a  given  cluster  are 
at  essentially  the  same  distance  and  are  believed  to  have  about  the  same  age. 
Determinations  of  the  apparent  magnitude  and  color  of  the  stars  in  these 
clusters  yield  at  once  the  data  on  the  relationship  between  luminosity  and 
surface  temperature  for  groups  of  stars  of  various  ages  that  are  so  essential  for 
the  interpretation  of  stellar  evolution.  A  study  of  NGC  188  by  Sandage  and 
Dr.  Sidney  van  den  Bergh  of  the  Perkins  Observatory  gave  evidence  that  it  is 
about  1.5  billion  years  older  than  any  previously  known  galactic  cluster.  Thirty- 
two  white  dwarfs,  which  are  supposed  to  represent  the  final  stage  in  stellar 
evolution,  were  discovered  by  Baade  in  the  globular  cluster  M  67.  From  spec- 
troscopic studies,  Jesse  L.  Greenstein,  with  the  assistance  of  H.  Lawrence  Heifer 
and  George  Wallerstein,  found  that  the  abundance  of  the  metals  in  the  stars  of 
the  globular  cluster  M  92  is  only  0.5  per  cent  of  that  in  the  stars  in  the  solar 
neighborhood,  whereas  in  the  cluster  M  13  it  is  about  5  per  cent. 

Armin  J.  Deutsch  has  found  important  evidence  of  a  process  by  which  the 
massive  stars,  when  in  the  expanded  red-giant  stage,  eject  into  space  an  ap- 
preciable fraction  of  their  mass,  providing  a  probable  explanation  for  the  great 
differences  between  the  initial  mass  of  the  blue  giant  stars  and  their  final  mass 
in  the  white-dwarf  stage,  and  explaining  the  presence  of  the  large  number  of 
atoms  of  the  heavy  elements  observed  in  the  stellar  gas. 

A  most  important  field  of  investigation  was  opened  by  Hubble  in  the  third 
and  fourth  decades  of  the  century  with  the  discovery  of  the  linear  relationship 
between  the  radial  velocities  of  the  galaxies  and  their  distances.  In  the  1920's 
and  the  1930's  these  velocities  were  extensively  measured  with  the  100-inch 
telescope  by  Humason.  When  the  Hale  telescope  was  completed,  Humason 
used  this  instrument  and  its  prime-focus  spectrograph  to  measure  the  radial 
velocities  of  a  substantial  number  of  additional  galaxies.  In  particular,  he  ex- 
tended those  measurements  to  very  distant  objects  having  a  velocity  of  reces- 
sion about  20  per  cent  that  of  light.  At  the  same  time  Stebbins  and  Whitford 
of  the  Washburn  Observatory  and  Edison  Pettit  of  the  Mount  Wilson  and 
Palomar  Observatories  used  the  60-inch  and  the  100-inch  instruments  to  meas- 
ure photoelectrically  the  magnitudes  of  the  galaxies  whose  velocities  had  been 
determined.  In  1956  Humason,  Sandage,  and  Dr.  N.  Mayall  of  the  Lick  Ob- 
servatory published  a  list  of  the  velocities  of  600  galaxies  measured  at  Mount 
Wilson  and  Palomar,  together  with  300  galaxies  observed  at  the  Lick  Observa- 
tory. A  detailed  analysis  was  then  made  of  the  velocity-distance  relationship, 
based  on  the  observed  velocities  and  the  distances  deduced  from  magnitudes. 

A  new  method  for  the  measurement  of  the  redshifts  of  very  faint  and  distant 
galaxies  has  been  developed  by  Baum.    Employing  eight-color  photometry, 


using  wavelength  bands  distributed  uniformly  throughout  the  spectrum,  he 
has  determined  the  shape  of  the  curve  relating  radiated  energy  to  wavelength 
for  a  number  of  galaxies  at  a  wide  range  of  distances.  The  displacement  of 
this  curve  for  a  distant  galaxy  with  respect  to  that  of  a  near-by  galaxy  gives 
a  measure  of  the  redshift.  Redshifts  as  large  as  AX/X=0.4,  nearly  twice  the 
largest  definite  redshift  obtainable  from  spectroscopic  observations,  have  been 
found.  Indeed,  it  is  becoming  evident  that  the  primary  factor  that  sets  the  limit 
on  the  distance  to  which  the  Hale  telescope — or  any  telescope  of  similar  magni- 
tude— can  observe  a  galaxy  or  a  cluster  of  galaxies  is  not  the  falling  off  of  the 
light  because  of  the  increasing  distance.  Instead,  it  is  the  redshift  of  the  light 
from  the  galaxy  into  the  far  infrared,  where  photographic  plates  are  no  longer 
sensitive,  where  atmospheric  absorption  is  large  and  atmospheric  "night-sky" 
is  strong. 

Observations  have  been  made  to  investigate  various  properties  of  the  galaxies. 
Thus  Guido  Munch  has  made  spectroscopic  observations  of  M  81  to  determine 
its  rotation  and  mass.  Jan  Hendrick  Oort  and  Rudolph  L.  Minkowski  have 
investigated  the  rotation,  the  mass  distribution,  and  the  range  of  stellar  veloci- 
ties in  NGC  3115.  Multicolor  measurements  of  the  integrated  light  from  a 
large  number  of  globular  clusters  and  galaxies  have  been  made  by  William  G. 
Tifft  for  a  study  of  their  stellar  contents.  Similar  studies  of  the  energy  distri- 
bution within  galaxies  have  been  made  by  Code,  using  a  spectroscopic  scanner. 
Fritz  Zwicky  has  carried  out  studies  of  the  luminosity  function  and  the  distribu- 
tion and  clustering  of  galaxies,  and  has  also  investigated  the  faint  blue  stars 
and  other  objects  in  the  halo  about  our  own  and  other  galaxies. 

Following  the  discovery  of  strong  radio  sources  in  space,  Baade  and  Minkow- 
ski were  able  to  identify  several  of  the  strongest  of  these  with  peculiar  celestial 
objects  which  had  already  been  observed  optically.  Several  of  these  radio 
sources,  such  as  Cygnus  A,  NGC  1275,  and  probably  NGC  5128,  have  proved 
to  be  in  fact  pairs  of  galaxies  in  collision.  It  is  thought  that  collision  of  the  gas 
clouds  in  the  spiral  arms  of  these  pairs  of  galaxies  is  responsible  for  the  strong 
radio  waves  they  emit.  The  remains  of  two  supernovae  which  exploded  earlier 
— the  Crab  Nebula,  known  to  have  exploded  in  A.D.  1054,  and  a  faint  nebula 
in  Cassiopeia  first  observed  by  Minkowski  and  identified  by  him  as  the  remains 
of  the  Supernova  of  1572 — have  been  shown  to  be  identical  with  radio  sources. 
The  discovery  in  the  Soviet  Union  of  the  polarization  of  the  light  from  the 
Crab  Nebula,  followed  by  the  detailed  studies  of  this  polarization  by  Baade, 
confirms  the  suggestion  that  much  of  the  light  observed  from  this  object  is 
synchrotron  radiation  from  electrons  of  very  high  velocity  accelerated  in  a 
small  magnetic  field.  Similarly,  the  observations  by  Baade  of  the  polarization  of 
the  light  from  the  jet  of  the  Galaxy  M  87,  also  a  radio  source,  points  to  synchro- 
tron radiation  as  the  source  of  much  of  its  light. 

The  discovery  in  1946  by  H.  W.  Babcock  of  a  star  having  a  general  magnetic 


field  with  a  strength  of  several  thousand  gauss  opened  a  field  of  great  interest. 
It  has  been  followed  by  a  survey  of  many  similarly  magnetic  stars.  In  all,  305 
stars  have  been  examined  from  this  point  of  view.  Of  these,  84  show  a  definite 
magnetic  field.  It  is  probable  that  there  are  such  fields  in  55  additional  ones. 
Since  in  many  stars  the  magnetic  fields  fluctuate  more  or  less  regularly  with  a 
period  of  a  few  days,  it  has  been  necessary  to  make  extensive  observations  of 
many  of  these  objects  over  a  long  period  of  time.  The  spectroscopic  and  other 
properties  of  these  magnetic  stars  have  been  extensively  studied  by  Deutsch  in 
an  attempt  to  construct  a  model  to  explain  their  behavior. 

With  the  development  of  the  solar  magnetograph  by  H.  D.  and  H.  W.  Bab- 
cock,  it  has  become  possible  to  map  the  magnetic  fields  over  the  surface  of  the 
sun  with  an  accuracy  of  a  few  tenths  of  a  gauss.  Daily  maps,  taken  over  a 
period  of  a  few  years,  provide  definite  evidence  of  persistent  polar  fields  and  of 
fluctuating  fields  over  the  equatorial  regions  of  the  sun.  The  observed  fluctua- 
tions explain  why  the  early  attempts  of  Hale  and  of  others  to  measure  a  general 
field  were  inconclusive.  Sunspots,  prominences,  flares,  and  other  solar  phe- 
nomena have  been  investigated  by  Seth  B.  Nicholson  and  Robert  S.  Richardson 
on  photographs  and  spectroheliograms  taken  daily  at  the  solar  towers  on  Mount 

Extensive  studies  have  been  made  of  the  gas  clouds  within  our  own  Galaxy. 
Surveys  carried  out  by  Minkowski  have  nearly  doubled  the  number  of  known 
planetary  nebulae.  Spectroscopic  techniques  have  been  developed  by  Olin  C. 
Wilson  which  have  enabled  him  to  study  the  internal  motions  of  the  planetary 
nebulae  and,  with  Munch,  to  map  the  motions  of  the  gases  over  a  large  part  of 
the  brighter  areas  of  the  nebula  of  Orion.  Osterbrock  has  used  a  new  technique 
to  measure  the  densities  of  the  gases  in  various  nebulae. 

During  this  period  Ralph  E.  Wilson  completed  the  radial-velocity  program 
of  the  Observatories.  This  involved  the  measurement  of  several  thousand 
plates,  many  of  which  were  accumulated  during  World  War  II,  and  the  publi- 
cation of  the  radial  velocities  of  more  than  2400  stars.  Under  the  title  General 
Catalogue  of  Stellar  Radial  Velocities,  Wilson  has  published  a  compilation  of 
the  positions,  magnitudes,  spectral  types,  and  definitive  radial  velocities  of  all 
the  15,105  stars  whose  velocities  have  been  determined  at  any  observatory. 

Olin  C.  Wilson,  with  the  assistance  of  M.  K.  Vainu  Bappu,  has  made  a  de- 
tailed study  of  the  widths  of  the  emission  components  at  the  center  of  the 
H  and  K  lines  of  the  spectra  of  a  number  of  late-type  stars.  He  discovered  a 
linear  relationship  between  the  logarithm  of  this  width  and  the  absolute  magni- 
tude of  the  star  which  should  prove  a  powerful  tool  in  determining  the  dis- 
tances of  stars  of  these  types.  A  large  number  of  observations  have  been  made 
by  Paul  W.  Merrill,  Alfred  H.  Joy,  and  Roscoe  F.  Sanford  on  the  spectra  of 
variable  stars,  and  of  stars  with  prominent  emission  lines. 


A  major  program  of  mapping,  including  the  whole  sky  north  of  decimation 
— 27°,  carried  forward  under  the  sponsorship  of  the  National  Geographic 
Society,  occupied  most  of  the  present  decade.  Two  photographs  were  taken  of 
each  of  879  separate  fields,  one  in  blue  light  and  the  other  in  red  light.  Stars 
were  recorded  to  the  twenty-first  magnitude  on  the  blue  plates  and  to  the 
twentieth  on  the  red  plates,  achieving  a  penetration  to  a  distance  about  three 
times  farther  than  any  previous  surveys — representing  a  coverage  of  about 
twenty-five  times  the  volume  of  space.  More  than  120  copies  of  this  National 
Geographic-Palomar  Observatory  Sky  Survey,  containing  1758  photographic 
prints,  have  been  ordered  and  are  being  distributed  to  observatories  throughout 
the  world.  From  a  search  of  the  survey  plates  George  O.  Abell  compiled  a  list 
of  2712  galaxies,  of  which  but  a  few  dozen  had  previously  been  known. 

This  year  Sandage  and  E.  M.  Burbidge  have  found  that  two  faint  globular 
clusters  discovered  by  Abell  in  the  Survey  lie  at  a  distance  of  400,000  light-years. 
As  this  is  twice  the  distance  of  the  Magellanic  Clouds  these  clusters  are  clearly 

Two  very  unusual  asteroids,  named  Icarus  and  Geographos,  which  may 
come  unusually  close  to  the  earth,  were  discovered  just  before  and  during  the 
Survey  from  observations  made  with  the  48-inch  schmidt  camera.  Another 
discovery  of  a  new  object,  comparatively  near  by,  was  made  in  1951,  when 
Nicholson  detected  the  twelfth  satellite  of  Jupiter  on  plates  taken  with  the 
100-inch  telescope  on  Mount  Wilson. 

Shortly  before  the  initiation  of  the  joint  operation  of  the  Mount  Wilson  and 
Palomar  Observatories,  a  guest-investigator  program  was  inaugurated.  Its 
purpose  was  to  make  the  unique  facilities  of  the  Observatories  available — to  the 
extent  that  they  were  not  required  by  the  regular  staff — to  a  wider  group  of 
astronomers  from  other  institutions.  This  program  has  been  carried  forward 
consistently  on  a  cooperative  basis  throughout  the  decade.  The  institution  from 
which  the  visiting  astronomer  comes  provides  the  necessary  leave  and  travel 
expenses  for  him;  the  Observatories  furnish  the  telescope  time  and  the  photo- 
graphic plates  and  other  supplies  for  the  proposed  program.  During  the  decade, 
60  astronomers  from  23  institutions  in  the  United  States  have  made  more  than 
160  visits  to  the  Observatories  to  carry  on  work  directly  with  the  telescope,  or 
in  a  very  few  cases  to  study  plates  already  available  in  the  files.  In  this  same 
time  20  astronomers  from  18  institutions  in  12  other  countries  have  made  26 

The  optical  tests  of  the  Hale  telescope,  which  were  reported  early  in  the 
decade,  have  shown  that  the  instrument  performs  as  well  as  had  been  planned. 
But  the  real,  functional  test  of  any  great  new  instrument,  and  of  the  thinking 
and  planning  that  lie  behind  it,  is  the  degree  to  which  it  can  solve  new  prob- 
lems, the  extent  to  which  it  can  break  new  ground.   The  record  of  the  first 


decade  of  operation  of  the  200-inch  telescope  and  its  supporting  instruments  on 
Palomar  Mountain  and  Mount  Wilson — the  record  of  the  past  ten  years  of  the 
Mount  Wilson  and  Palomar  Observatories,  in  short — provides  the  final  and 
crowning  evidence  of  the  vision  of  George  Ellery  Hale,  embodied  in  the  great 
and  unique  instrument  that  bears  his  name. 

The  Department  of  Archaeology 

This  year  marks  a  major  occurrence  in  the  program  of  the  Carnegie 
Institution — the  completion  of  the  work  of  the  Department  of  Archaeology. 
It  signalizes  the  accomplishment  of  a  most  noteworthy — indeed  a  most  out- 
standing— undertaking,  which  took  its  origin  in  the  Institution  more  than 
fifty  years  ago,  the  records  of  which  fill  approximately  a  hundred  volumes  of 
Institution  publications  and  include  many  papers  published  elsewhere,  and 
which  culminated  in  the  extraordinary  pioneering  investigations  of  Middle 
American  aboriginal  culture  for  which  it  is  best  known. 

It  is  clear  that  consideration  was  given  to  research  in  various  phases  of 
archaeology  from  the  very  beginning  of  the  history  of  the  Carnegie  Institution. 
In  the  earliest  years,  indeed,  programs  were  formulated  and  grants  were  made 
for  archaeological  research  in  a  wide  variety  of  fields  and  a  wide  range  of  geo- 
graphic locations.  It  was  in  1913,  however,  with  the  publication  of  Reports 
upon  the  Present  Condition  and  Future  Needs  of  the  Science  of  Anthropology, 
that  a  direction  of  research  was  outlined  and  crystallized  which  was  to  set  the 
main  course  of  effort  of  the  Institution  in  this  field  for  the  succeeding  forty-five 
years — an  approach  determined  largely  by  the  able  advocacy  by  Sylvanus  G. 
Morley  of  the  Maya  civilization  of  Middle  America  as  an  appropriate  field  of 

With  the  approval  of  that  course  by  the  Trustees  of  the  Institution  there 
followed  a  preparatory  phase  of  exploration,  reconnaissance,  and  planning 
which  occupied  ten  years.  It  was  a  fitting  introduction  to  the  intensive  period  of 
field  research  that  followed.  Explorations  in  the  Peten  region  of  Guatemala,  in 
Honduras,  Nicaragua,  and  Costa  Rica  were  accompanied  by  preliminary  in- 
vestigation of  the  ruins  of  Copan  and  were  crowned  by  the  discovery  of  the 
ruins  of  Uaxactun,  with  "the  oldest  monument  yet  reported  from  the  Maya 
field"  in  1916  and  the  publication  of  Morley's  important  monograph  on  the 
inscriptions  at  Copan  in  1920.  This  exploratory  period  was  characterized  by 
another  notable  feature,  the  significance  of  which  to  further  research  it  is  hard 
to  overemphasize — the  development  of  two  basic  principles  of  procedure  in 
field  work.  First,  the  Institution  in  its  research  programs  should  refrain  from 
exporting  any  of  its  archaeological  finds  from  the  countries  of  their  origin  but 
rather  should  return  all  of  them  to  the  appropriate  government  authorities 
when  study  was  completed.  Second,  the  Institution  should  assume  an  obliga- 


tion  to  preserve  remains,  once  uncovered,  from  the  further  deterioration  by 
weathering  which  would  otherwise  be  inevitable.  This  philosophy  defined 
a  climate  that  has  been  maintained  throughout  the  entire  program  of  archaeol- 
ogy of  the  Institution — a  tone  that  has  fostered  and  expanded  a  continuing 
measure  of  good  will. 

It  was  in  1924  that  large-scale  excavations  at  Chichen  Itza  in  Yucatan,  recom- 
mended in  Morley's  original  report,  were  begun  in  earnest,  to  be  continued 
unabated  for  ten  years.  Two  years  later  investigations  of  similar  intensity  were 
begun  at  Uaxactun  and  carried  forward  for  a  dozen  years  thereafter.  This  was 
the  period  of  the  most  intensive  exploration  and  discovery — the  period,  possibly, 
when  our  vision  and  knowledge  of  the  true  nature  of  the  complex  of  Middle 
American  cultures  were  most  rapidly  expanded  and  brought  into  focus.  In  1928 
George  C.  Vaillant,  in  the  course  of  a  visit  of  but  a  few  days  to  Uaxactun,  sank 
a  pit  through  the  various  plaza  levels  of  the  ruin  into  the  soil  beneath.  The 
artifacts  that  appeared  from  this  level  were  not  Mayan  in  the  accepted  sense, 
but  rather  strongly  suggested  the  early  Archaic  pottery  of  the  Mexican  and 
Guatemalan  highlands.  This  important  discovery  opened  a  whole  new  horizon 
of  Maya  prehistory,  suggesting  a  major  concept  then  quite  new — the  idea  that 
there  might  have  been  connections  between  the  cultures  of  the  lowland  and 
the  highland  Maya  regions  at  a  very  early  period — that  indeed  there  might  have 
been,  in  some  sense,  a  common  base  linking  them  in  early  pre-Classic  Mayan 
times.  That  concept  led  on  to  a  program  of  extensive  investigation  in  the 
Guatemalan  highlands,  begun  in  1932  and  culminating  in  the  pivotally  im- 
portant excavations  of  two  inconspicuous  mounds  at  the  great  ruin  site  of 
Kaminaljuyu  outside  of  Guatemala  City  by  Alfred  V.  Kidder,  Jesse  D.  Jennings, 
and  Edwin  M.  Shook  in  1936,  1937,  1941,  and  1942.  That  work  resulted  in  a 
linkage  of  the  great  Classic  centers  of  Middle  American  culture  in  the  Valley 
of  Mexico,  the  Valley  of  Oaxaca,  the  Guatemalan  highlands,  and  the  lowlands 
of  Peten.  For  the  first  time,  the  high  cultures  of  Middle  America  were  brought 
into  focus  in  time  and  space,  and  it  was  revealed  that  they  formed  one  great 
area  of  relatively  homogeneous,  or  in  any  case  interdependent,  civilizations.  It 
was  established  that  the  Mayan  Classic  civilization  had  a  base  far  broader  and 
deeper  than  earlier  workers  had  imagined,  that  it  was  not  a  unique  development 
from  which,  alone,  high  culture  was  disseminated  to  adjacent  peoples;  that, 
on  the  contrary,  the  Classic  phases  of  a  number  of  other  cultures  were  roughly 
coeval  with  the  Maya. 

Perhaps  this  demonstration  of  the  homogeneity  and  the  massiveness  of  the 
Middle  American  cultures,  of  the  breadth  of  the  common  base  upon  which 
they  rested  in  the  cultures  of  the  Formative  period,  and  of  the  age,  the  com- 
plexity, the  high  development  characteristic  of  these  Formative  cultures  them- 
selves— emphasized  particularly  by  later  excavations  at  Kaminaljuyu  in  the 


period  1946-1950 — represented  among  the  most  important  of  all  the  results  of 
the  Mayan  program  in  the  Institution.  That  situation  was  further  underlined 
by  the  finding  of  sites  of  Formative  period  pottery  in  the  northern  part  of  the 
Yucatan  peninsula  in  1942  by  G.  W.  Brainerd,  proving  that  the  northern  cul- 
tures were  as  deeply  rooted  as  the  southern,  and  that  the  relatively  specific 
sequences  between  the  cultures  of  the  north  and  south  which  had  earlier  been 
visualized  were  no  longer  tenable.  The  massiveness  was  emphasized  by  the 
final  major  field  program  of  the  Department,  undertaken  in  1950,  at  the  site  of 
Mayapan,  the  last  great  aboriginal  Mayan  city.  This  was  an  intensive  investiga- 
tion, carried  forward  in  depth,  with  particular  emphasis  on  the  secular  aspects 
of  the  culture.  It  established  the  time  of  the  greatness  of  Mayapan  beyond 
reasonable  doubt.  It  also  established  another  finding  of  singular  importance. 
There  had  been  a  widespread  assumption  that  the  final  and  dramatic  break 
with  cultural  tradition  that  marked  the  close  of  Mayan  greatness — the  precipi- 
tous descent  from  order  and  high  culture  to  what  must  have  been  chaos  bor- 
dering on  ruin — occurred  with  the  rise  to  power,  under  foreign  domination, 
of  the  great  center  of  Chichen.  The  investigations  identified  that  precipitous 
decline  rather  with  the  end  of  the  greatness  of  Chichen  and  the  rise  to  power  of 
Mayapan  itself. 

These  are  but  the  broad  and  blurred  outlines  of  an  extensive  and  varied  pro- 
gram. In  the  very  wealth  of  its  content,  it  is  easy  to  overlook  facets  somewhat 
separated  from  the  main  stream  but  of  intrinsic  importance,  such  as  the  pro- 
gram of  study  of  the  Early  Basketmaker  II  culture  of  the  North  American 
Southwest,  carried  on  for  ten  years  in  Arizona,  New  Mexico,  and  southern 
Colorado  by  Earl  H.  Morris,  with  very  interesting  results.  It  is  easy  to  over- 
look special  but  central  aspects  of  the  work,  such  as  the  long  and  fertile  history 
of  research  in  Mayan  hieroglyphics  and  the  investigations  in  ceramics,  or  dra- 
matic specific  findings,  such  as  those  resulting  from  the  excavations  at  Copan 
or  the  spectacular  discovery  of  the  famous  wall  paintings  of  Bonampak,  in  the 
lowland  Maya  country.  Finally,  it  is  all  too  easy  to  neglect  or  underestimate 
one  of  the  most  important  consequences  of  the  whole  program  of  Middle 
American  research  in  the  Institution :  the  cordial  relations  that  were  established 
with  the  countries  in  which  the  investigations  were  made,  the  good  will  that 
was  uniformly  generated,  and  the  great  stimulus  to  local  efforts  of  research  in 
the  field  that  the  program  conferred.  The  significant  strengthening  of  the 
archaeological  interests  and  programs  of  the  Mexican  Government,  and  the 
establishment  of  a  museum  of  archaeology  and  ethnology  by  the  government 
of  the  Republic  of  Guatemala  and  the  organization  of  a  national  institute  of 
anthropology  and  history  there,  represent  but  two  of  the  tangible  consequences 
of  this  policy  and  the  climate  it  created  and  maintained. 

The  record  of  the  Department  of  Archaeology  which  closes  this  year  has  been 


a  great  one,  worthy  of  special  pride.  It  is  a  particular  satisfaction  that  its  closing 
does  not  mean  that  all  effort  in  this  field  will  be  terminated.  The  work  of 
the  Institution  will  continue  to  be  enriched  by  the  investigations  of  the  Director 
of  the  Department  and  of  certain  members  of  its  staff. 

The  Department  of  Terrestrial  Magnetism  has  continued  its  active  program 
in  radio  astronomy,  the  study  of  the  emission  of  energy  at  radio  frequencies  by 
celestial  objects,  of  which  the  Crab  Nebula  provides  so  striking  an  example. 
The  past  year  has  witnessed  solar  activity  of  unprecedented  magnitude.  This 
has  interfered  seriously  with  optical  redshift  measurements  of  the  far  distant 
galaxies  at  Mount  Wilson  and  Palomar.  It  has  also  made  it  necessary,  as  was 
mentioned  last  year,  to  postpone  for  the  time  being  the  continuing  program  of 
measurements  of  radio  frequencies  from  astronomical  objects  in  the  range  of 
12  to  15  mc  which  was  begun  at  the  Department  of  Terrestrial  Magnetism  in 
1955-1956.  Work  has,  however,  proceeded  rapidly  on  precision  equipment  de- 
signed to  measure  the  position  of  radio  sources  to  within  a  few  square  minutes 
of  arc.  This  is  a  400  mc/sec  linear  array,  consisting  of  a  pair  of  V-reflectors,  each 
614  feet  long,  arranged  on  an  east-west  baseline  with  a  spacing  of  1842  feet 
between  centers.  These  arrays  can  be  used  separately,  when  they  yield  a  fan 
beam  %°X20°  to  half-power  points,  or  they  can  be  employed  together  as  an 
interferometer  with  a  lobe  spacing  of  4  minutes  of  arc.  Calculations  indicate 
that  the  two  arrays  should  have  an  effective  collecting  area  equivalent  to  that 
of  two  90-foot  paraboloids.  At  the  close  of  the  year,  all  the  elements  of  this 
new  system  had  been  assembled  and  mounted  and  final  connections  were  being 
made.  A  test  had  also  been  completed  on  the  phase  stability  of  open-wire  trans- 
mission lines.  This  is  a  factor  vital  to  accuracy  in  the  final  measurements.  The 
limits  with  respect  to  the  conditions  of  temperature,  humidity,  and  frost  under 
which  the  system  will  operate  appear,  as  a  result  of  tests,  to  be  such  that  the 
system  will  be  operable  for  a  considerable  fraction  of  the  time — perhaps  as 
much  as  three-quarters  of  the  total. 

Last  year  an  antenna  array  specially  designed  for  a  detailed  examination  of 
the  radio  emission  of  the  sun  was  built  by  the  Department  at  its  River  Road  site 
near  Seneca,  Maryland,  and  preliminary  scanning  of  the  sun's  disk  revealed 
localized  bright  sources  which  traveled  across  it  as  the  sun  rotated.  By  the  end 
of  this  year,  the  equipment  had  been  operated  long  enough  to  provide  a  good 
description  of  the  events  occurring  on  the  sun  at  a  wavelength  of  340  mc.  The 
most  common  features  found  were  quiet  bright  spots,  already  detected  and  il- 
lustrated last  year.  Their  positions  on  the  disk  usually  agreed  with  that  of  large 
plages,  but  the  converse  was  by  no  means  always  true — other,  equally  large, 
plages  seemed  to  have  no  radio  bright  spots  associated  with  them.  The  spots 
persisted  for  several  days,  sometimes  for  an  entire  disk  passage. 


A  second  common  feature  of  the  solar  face  observed  at  340  mc  consisted  of 
active  spots  which,  though  persisting  for  several  days  like  the  quiet  spots, 
showed  changes  in  intensity  and  produced  frequent  small  bursts  lasting  a 
second  or  less.  Active  spots  may  be  much  more  intense  than  the  quiet  ones ;  the 
largest  detected  so  far  gave  a  steady  flux  of  60X10"22  watt/meter2/cycle/ 
second,  with  instantaneous  values  of  perhaps  twice  this  magnitude.  These  values 
are  to  be  compared  with  fluxes  from  about  5X10"23  watt/meter2/cycle/second 
(the  smallest  that  can  be  detected)  to  about  5X 10"22  watt/meter2/cycle/second 
for  the  quiet  bright  spots. 

At  the  close  of  the  International  Geophysical  Year,  it  is  interesting  to  recol- 
lect that  the  idea  actually  took  its  origin  in  discussions  among  several  members 
and  former  members  of  the  staff  of  the  Department  of  Terrestrial  Magnetism  in 
1950.  There  had  been  a  half-century  interval  between  the  two  previous  Interna- 
tional Polar  Years  (1883  and  1933).  It  was  generally  understood  that  the  next 
period  of  large-scale  international  collaboration  in  geophysical  studies  would 
come  in  1983.  But  the  tremendous  advances  in  exploration  made  possible  by 
modern  radio  communication  and  techniques  of  air  reconnaissance  and  airlift 
supply  suggested  that  a  shorter  interval  of  25  years  would  be  especially  attractive 
and  appropriate.  A  suggestion  to  this  effect  at  an  international  meeting  of  geo- 
physicists  led  quickly  to  wide  approval  and  support. 

It  seemed  most  appropriate,  therefore,  that  the  Department  participate  in 
the  world  program  which  it  did  so  much  to  initiate,  and  that,  in  accordance 
with  the  philosophy  of  the  Institution  as  a  whole,  this  participation  be  on  the 
basis  of  active  individual  research.  Accordingly,  three  noteworthy  projects 
have  been  carried  forward  during  the  past  year  as  a  part  of  the  IGY  program. 
All  of  them  represent  further  extensions  of  interests  already  developed  in  the 

The  first  of  these  projects  concerned  the  study  of  the  intense  band  of  electric 
current,  called  the  "electrojet,"  that  circulates  in  the  upper  atmosphere  in  the 
region  of  the  earth's  magnetic  equator.  Such  a  study  is  of  peculiar  interest 
during  periods  of  magnetic  disturbance.  Four  Askania  variographs  loaned  to 
the  Department  through  the  cooperation  of  the  U.  S.  Coast  and  Geodetic  Survey 
were  put  in  operation  at  temporary  magnetic  observatories  established  on  the 
coast  of  Peru  at  Talara,  Chiclayo,  Chimbote,  and  Yauca,  extending  from  geo- 
graphic latitude  4°  N.  to  15°  S.  The  project  was  initiated  in  collaboration 
with  the  Instituto  Geofisico  de  Huancayo.  A  permanent  magnetic  observatory 
was  also  established  for  the  University  of  Arequipa  by  the  Instituto  Geofisico 
for  which  the  Department  provided  a  la  Cour  vertical  intensity  variometer.  In 
this  multistation  network  spaced  on  both  sides  of  the  magnetic  equator  in  Peru 
the  "electrojet"  phenomenon  has  been  extensively  investigated  during  the  year 
and  magnetic  storm  and  disturbance  relationships  have  been  studied. 


The  second  project  in  which  the  Department  has  participated  with  the 
IGY  concerns  a  field  far  removed  from  considerations  of  galaxies  and  stars 
but  of  even  more  immediate  concern  and  appeal — the  nature  of  our  own  earth. 
For  many  years  the  Department  has  been  conducting  an  extensive  series  of 
seismic  and  gravity  studies  designed  to  increase  our  understanding  of  the 
structure  of  the  earth's  crust,  and  especially  of  large-scale  processes  which, 
operating  over  long  periods  of  time,  have  resulted  in  the  formation  of  continents 
and  ocean  deeps,  high  plateaus  and  mountain  ranges.  In  previous  years  this 
program  has  included  field  studies  on  the  Colorado  Plateau  and  in  Alaska. 
Last  year  it  was  reported  that  a  new  study,  of  similar  nature,  was  planned  in  the 
Andean  highlands,  utilizing  explosions  normally  set  off  in  the  operation  of 
large  open-pit  copper  mines.  During  the  past  summer  this  was  carried  out, 
with  results  both  interesting  and  somewhat  puzzling,  which  are  at  present 
under  study.  In  the  high  montane  and  alpine  plateau  regions  of  Peru,  Bolivia, 
and  Chile  some  indication  was  obtained  that  the  crustal  thickness  under  the 
Andes  may  be  10  to  20  km  greater  than  had  previously  been  found  in  North 
America.  Unusually  severe  attenuation  of  the  reflected  waves  was  encountered 
under  the  high  Andean  plateau,  however,  and  continuation  of  the  investiga- 
tions, probably  by  earthquake  observations  in  collaboration  with  university 
colleagues  in  Peru  and  Chile,  is  anticipated. 

The  third  area  of  interest  to  the  Department  which  was  expanded  and  inten- 
sified over  the  past  year  within  the  framework  of  the  IGY  is  one  which  it  has 
long  shared  with  the  Geophysical  Laboratory — the  measurement  of  the  ages  of 
rock  minerals  through  the  methods  provided  by  the  study  of  radioisotopes. 
Collections  of  samples  for  this  work  were  made  in  South  America,  northern 
Europe,  and  selected  additional  regions  of  the  United  States. 

The  program  on  mineral  ages  conducted  by  the  cooperative  age  group,  con- 
sisting of  L.  Thomas  Aldrich  and  George  W.  Wetherill  in  the  Department 
and  George  R.  Tilton  and  Gordon  L.  Davis  in  the  Geophysical  Laboratory, 
has  been  concerned  this  year  with  investigations  of  the  geographical  patterns  of 
the  ages  of  Precambrian  rocks  in  North  America.  A  comprehensive  study  of 
findings  made  by  the  Institution,  together  with  data  obtained  at  the  geology  de- 
partments of  the  University  of  Minnesota  and  the  Massachusetts  Institute  of 
Technology  and  at  the  Lamont  Geological  Observatory,  has  developed  a  rather 
definite  picture  of  the  broad  outlines  of  major  periods  of  mineral  formation 
for  those  parts  of  the  continent  that  have  now  been  measured.  A  large  area  with 
rocks  of  ages  exceeding  2500  million  years  extends  from  western  Quebec 
through  Ontario  to  eastern  Saskatchewan  and  northern  Minnesota.  It  reap- 
pears in  Montana  and  Wyoming.  A  second  large  area,  with  rocks  close  to  1850 
million  years  old,  occurs  in  the  western  United  States.  Rocks  of  this  age  are  now 
known  also  from  Missouri,  Wisconsin,  Michigan,  and  Ontario.  A  third  large 
region  having  similar-aged  rocks  extends  from  northern  Quebec  south  into 


Ontario,  New  York,  New  Jersey,  Virginia,  and  North  Carolina.  All  these  rocks 
were  formed  close  to  1000  million  years  ago.  Thus  this  period  too  appears  to 
have  been  one  in  which  active  processes  of  mineral  formation  were  proceeding 
over  large  areas  in  North  America. 

The  existence  of  five  different  nuclear  clocks  (decay  of  U238,  U235,  Th232,  Rb87, 
K40)  has  created  new  opportunities  for  research  on  the  time  sequences  of 
ancient  processes,  a  second  major  concern  of  the  cooperative  age  group.  A 
whole  series  of  pertinent  questions  may  now  be  fruitfully  asked  about  typical 
mountain  chains,  such  as  the  Appalachians.  How  old  are  the  ancient  crystal- 
line rocks  that  predate  the  orogeny?  How  did  the  Paleozoic  upheaval,  oc- 
curring 300  million  years  ago,  affect  the  apparent  ages  of  minerals  in  the  old 
rocks  ?  What  was  the  sequence  of  events  in  various  parts  of  the  belt  ? 

Intensive  study  has  been  given  to  the  metamorphic  Precambrian  rocks  of  the 
central  and  southern  Appalachians.  Measurements  of  mineral  ages  and  petro- 
graphic  examination  of  the  gneisses  of  this  region  suggest  the  possibility  that 
there  have  been  two  major  periods  of  mineral  formation.  During  the  first, 
dating  from  1000  to  1100  million  years  ago,  zircon  and  probably  potassium 
feldspar  were  formed.  The  age  of  zircon  from  the  Baltimore  gneiss,  a  part  of 
the  basement  complex  of  the  Appalachian  Piedmont,  has  been  determined  as 
1100  million  years  from  nearly  concordant  uranium-lead  ages.  During  the 
second  period  of  active  mineral  formation,  which  occurred  about  300  to  350 
million  years  ago,  it  would  appear  that  the  mica  of  the  gneiss  was  formed. 
Biotite  ages  of  about  300  million  years  have  been  established  by  both  rubidium- 
strontium  and  potassium-argon  methods.  It  appears  that  the  rock  may  have 
crystallized  about  1100  million  years  ago  and  that  the  biotite  age  is  related  to 
local  metamorphism  taking  place  some  800  million  years  later. 

A  general  picture  of  geologic  history  of  the  Appalachian  orogenic  belt 
emerges.  The  belt,  extending  along  the  Atlantic  coast  from  southeastern  Canada 
to  the  southeastern  United  States,  apparently  experienced  several  periods  of 
deformation  between  250  and  400  million  years  ago.  Great  thicknesses  of  sedi- 
ments may  have  been  deposited  on  a  basement  of  gneisses  and  granites.  Sub- 
sequently both  sediments  and  basement  sank  into  the  crust  and  were  subjected 
to  elevated  temperatures  and  pressures.  Here  some  of  the  crystalline  rocks  and 
sediments  were  altered  to  metamorphic  rocks,  and  granitic  rocks  were  at  the 
same  time  formed  or  intruded.  Finally,  uplift  and  erosion  exposed  the  resulting 
assemblage,  now  including  gneiss,  schist,  marble,  quartzite,  and  granite. 

Studies  of  mineral  ages  during  the  year  have  not  been  confined  to  rocks  of 
the  North  American  continent,  or  even  of  the  western  hemisphere.  During 
the  Andes  expedition  of  the  Department,  as  indicated  earlier,  samples  of  rocks 
were  obtained  from  Peru,  Chile,  and  Brazil.  Peruvian  specimens  have  now 
been  analyzed.  The  evidence  suggests  that  they  are  Paleozoic.  Zircon  obtained 


from  a  gneiss  at  Koli,  Finland,  which  occupies  a  position  there  analogous  to 
that  of  the  Baltimore  gneiss,  was  studied  for  comparison  with  the  Appalachian 
samples.  It  presents  an  interesting  discordance.  Thus  rubidium-strontium  and 
potassium-argon  determinations  by  J.  A.  O.  Kouvo  of  biotite  from  several 
gneisses  in  the  area  suggest  an  age  of  about  1800  million  years,  while  the  Pb207- 
Pb206  determinations  suggest  an  age  for  the  zircon  of  2600  to  2700  million  years. 
Possibly  the  gneiss  had  an  age  of  2600  to  2700  million  years  and  was  meta- 
morphosed some  1800  million  years  ago. 

When  the  ages  of  all  the  principal  regions  of  rocks  underlying  the  whole  of 
the  North  American  continent  have  been  determined,  it  may  be  possible  to 
evolve  a  critical  picture  of  the  way  in  which  the  continent  was  formed.  A 
principal  question  will  be  whether  the  continent  has  grown  from  an  ancient 
"nucleus"  by  the  successive  addition  of  materials  derived  from  depth  during 
successive  orogenies  or  whether  it  has  always  had  a  considerable  area.  The 
clarifying  of  this  picture  forms  one  major  goal  of  the  studies  of  the  ages  of 
rocks  in  North  America. 

Investigations  yielding  quite  different  kinds  of  information  bearing  on  the 
processes  of  mountain  building  continue  to  be  prosecuted  vigorously  in  the 
Geophysical  Laboratory.  They  are  concerned  with  the  behavior  of  mineral 
constituents  under  conditions  of  temperature  and  high  pressure  approximating 
those  obtaining  deep  within  the  earth's  crust.  To  this  end,  new  equipment  of 
rather  revolutionary  design  for  studying  geochemical  and  geophysical  phe- 
nomena has  been  constructed  at  the  Geophysical  Laboratory  by  Francis  R. 
Boyd,  Jr.,  and  Joseph  L.  England,  and  is  currently  under  test.  Pressures  ap- 
proaching 100,000  atmospheres  have  been  attained  at  temperatures  of  300°  to 
400°  C  with  the  "squeezer"  equipment  described  in  last  year's  report.  It  is 
desirable,  however,  to  work  at  much  higher  temperatures.  With  the  new 
"single-stage"  apparatus,  pressures  of  50,000  atmospheres  combined  with  tem- 
peratures up  to  1700°  C  have  been  achieved.  With  still  newer  equipment,  desig- 
nated as  the  two-stage  design,  in  which  the  piston  of  the  single-stage  apparatus 
is  supported,  successful  experimental  runs  of  65,000  atmospheres'  pressure  have 
been  made  at  a  temperature  of  1100°  C.  The  single-stage  equipment  is  being 
used  to  study  the  influence  of  pressure  on  the  melting  points  of  minerals  that 
might  exist  80  miles  below  the  earth's  surface. 

Data  on  a  number  of  systems  are  beginning  to  accumulate.  Of  particular 
interest  is  information  on  the  change  in  melting  point  of  various  silicates  with 
pressure — data  of  special  importance  since  they  define  a  limit  to  the  geothermal 
gradient  within  the  earth's  mantle. 

Studies  of  phase  equilibria  are  being  continued  in  other  connections.  Hans 
P.  Eugster  and  Dr.  Charles  Milton,  of  the  U.  S.  Geological  Survey,  have  been 
interpreting  phase-equilibria  data  on  alkaline  carbonates  to  explain  some  of  the 


bizarre  mineral  assemblages  of  the  Green  River  shales  of  Colorado,  Wyoming, 
and  Utah.  These  shales,  of  Eocene  age,  contain  the  world's  largest  reserves 
of  hydrocarbons.  Associated  with  them  are  a  number  of  unusual  alkaline 
minerals,  such  as  trona  (NaHCCVNasCOs^HsO).  This  highly  soluble  salt 
is  found  in  a  bed  10  feet  thick  and  several  miles  wide.  If  ordinary  river  water 
is  concentrated  by  evaporation  in  a  large  lake  basin,  an  alkaline  solution  results. 
Trona  is  known  to  be  precipitated  above  about  35°  C  from  concentrated  brines 
in  equilibrium  with  air  containing  300  to  400  parts  per  million  of  carbon 
dioxide.  The  formation  of  trona  in  Wyoming  can  be  explained  by  the  evapora- 
tion of  brines,  in  a  shallow  warm  basin,  equilibrated  with  air  having  approxi- 
mately the  same  carbon  dioxide  content  as  that  of  the  present  atmosphere. 

Geothermometry  continues  to  be  of  great  interest  also  in  quite  another  con- 
text: the  study  of  ore  deposits.  Ore  minerals  have  been  under  intensive  investi- 
gation during  the  year  in  the  Geophysical  Laboratory  by  Gunnar  Kullerud, 
R.  G.  Arnold,  H.  L.  Barnes,  L.  A.  Clark,  and  E.  H.  Roseboom,  Jr.  More  than  a 
score  of  systems  are  under  study,  the  data  from  which  will  ultimately  be  ap- 
plicable to  ore-deposit  geothermometry. 

Hatten  S.  Yoder  has  continued  and  expanded  his  studies  on  the  effect  of  water 
on  the  melting  relations  of  rock-forming  silicates,  while  Eugster,  D.  R.  Wones, 
and  A.  C.  Turnock  have  studied  pressure-temperature  stability  characteristic 
of  hydrous  iron-bearing  micas  and  chlorites. 

Investigations  in  experimental  petrology  have  made  active  progress  during 
the  year.  They  include  a  study  of  cordierites  by  Yoder,  J.  Frank  Schairer,  and 
W.  F.  Schreyer,  of  pyroxenes  by  Schairer  and  N.  Morimoto,  and  of  amphiboles 
by  W.  G.  Ernst,  and  a  comparable  study  of  spinels  by  Wones  and  Turnock.  A 
particularly  interesting  accomplishment  has  been  the  synthesis  of  a  low-tempera- 
ture form  of  cordierite  identical  with  samples  of  the  natural  mineral  obtained 
in  Albany  County,  Wyoming,  and  Guilford,  Connecticut. 

It  will  be  recalled  that  one  of  the  high  points  in  the  research  of  the  Geo- 
physical Laboratory  recorded  last  year  was  the  work  of  Felix  Chayes,  in  which 
optical  analogues  were  developed  to  simulate  diffraction  patterns  obtained  in 
X-ray  studies  of  crystal  structure.  By  varying  the  analogue,  Chayes  was  able 
to  study  the  kinds  of  patterns  that  might  be  obtained  in  various  types  of  order- 
disorder  in  crystals,  laying  particular  emphasis  on  the  "short-range"  ordering 
concerned  with  nearest-neighbor  pairings.  This  approach  has  been  continued, 
refined,  and  expanded  during  the  current  year.  Developments  have  concerned 
the  design  of  random-layered  sequences  characterized  by  arbitrary  levels  of 
short-range  ordering,  the  production  of  experimental  diffraction  masks  based 
on  such  sequences,  the  generation  of  diffraction  transforms  from  these  masks, 
and  direct  calculation  of  the  diffraction  effects  by  high-speed  computation. 

The  Geophysical  Laboratory  has  for  a  number  of  years  been  conducting  a 


vigorous  program  in  crystallography,  of  central  importance  to  many  fields 
of  research.  Almost  forty-one  years  ago  the  first  account  of  a  crystal  structure  to 
have  been  determined  in  the  United  States  was  published  from  the  Throop  Col- 
lege of  Technology,  the  parent  of  the  California  Institute  of  Technology.  It  was 
a  determination  of  the  structure  of  chalcopyrite,  and  it  was  made  possible  by  a 
grant  from  the  Carnegie  Institution  of  Washington  to  A.  A.  Noyes,  at  whose 
suggestion  the  research  was  undertaken.  It  is  therefore  of  particular  interest 
that  this  year  an  intensive  investigation  of  the  nature  of  chemical  bonding  in 
chalcopyrite  by  neutron  diffraction  procedures  has  been  undertaken  collabora- 
tively by  G.  Donnay,  J.  D.  H.  Donnay  at  Johns  Hopkins,  and  Drs.  L.  M.  Corliss, 
Julius  M.  Hastings,  and  Norman  Elliott  at  the  Brookhaven  National  Labora- 
tory. The  data  on  structure  already  available  from  X-ray  diffraction  studies  could 
not  give  complete  information  on  the  nature  of  the  bonding.  The  neutron  dif- 
fraction work  has  now  measured  the  magnetic  properties  of  the  mineral.  In- 
terpretation of  the  data  indicates  that  the  substance  is  best  represented  by  the 
formula  Cu+Fe+++S2.  In  an  associated  research  Morimoto  has  conducted  a 
crystallographic  study  of  arsenopyrite,  one  of  the  common  sulfide  minerals, 
using  X-ray  diffraction  methods. 

During  the  year  Willard  F.  Libby  has  carried  on  a  series  of  researches  on  the 
geochemistry  of  fission  products.  In  particular  he  has  been  concerned  with 
practical  aspects  of  soil  contamination  with  strontium  90  and  the  ways  in  which 
it  may  be  alleviated.  It  has  been  shown  that  potassium,  when  introduced  into 
soil  at  as  low  a  level  as  about  60  pounds  per  2  million  pounds  of  soil,  reduces 
the  observed  uptake  of  radioactive  strontium  in  experimental  plants  by  some- 
thing like  40  per  cent. 

The  Geophysical  Laboratory  has  made  particular  efforts  to  counteract  a 
situation  that  plagues  scientific  research  everywhere,  but  is  perhaps  unusually 
serious  in  geochemical  and  geophysical  research.  Publications  in  these  fields 
are  so  widely  scattered  that  many  of  them  must  be  effectively  lost  to  most 
graduate  and  professional  students  of  the  earth  sciences.  A  particularly  signifi- 
cant contribution  to  the  alleviation  of  this  problem  was  made  during  the 
academic  year  1957-1958  in  the  holding  of  a  series  of  weekly  seminars  and 
discussions  on  the  topic  "Researches  in  Geochemistry"  at  the  Laboratory  and  at 
the  Johns  Hopkins  University.  The  participants  included  some  of  the  leading 
geochemists  of  the  country.  The  manuscripts  have  been  edited  at  the  Laboratory 
and  will  be  issued  as  a  symposium  volume  early  in  1959.  Much  new  work,  as 
yet  unpublished,  will  be  included. 

Few  aspects  of  modern  research  are  more  challenging  than  those  of  theoreti- 
cal biology.  The  Carnegie  Institution  is  deeply  concerned  with  this  frontier. 
Five  of  the  seven  departments  include  in  their  programs  investigations  in  the 
life  sciences.  They  range  in  aspect  from  the  organization  of  molecular  units 


in  vital  processes  through  the  nature  and  function  of  somewhat  larger  but  still 
microscopic  or  submicroscopic  intracellular  entities  like  cell  walls,  microsomes, 
chromosomes,  ribonucleoprotein  particles,  and  chlorophyll-protein  complexes, 
to  the  processes  underlying  the  differentiation  and  organization  of  cells  them- 
selves in  the  developing  many-celled  embryo,  and  finally  to  the  structure  and 
interaction  of  natural  populations  of  many-celled  organisms,  with  the  questions 
of  experimental  taxonomy  and  of  evolution  which  they  invoke. 

In  the  Geophysical  Laboratory  Philip  H.  Abelson  has  this  year  investigated 
the  detection  of  organic  materials  in  Precambrian  rocks.  Shales  and  slates  of  a 
low  grade  of  metamorphism  like  the  Huronian  Rove  slate  and  the  Keeweena- 
wan  Nonesuch  shale  have  been  examined.  Yields  were  very  small  when  ex- 
traction methods  using  an  ethyl  alcohol-benzene  mixture  in  a  Soxhlet  apparatus 
were  applied  to  the  latter  and  negligible  for  the  former.  Exploratory  experi- 
ments involving  hydrogenation,  carried  out  on  Swedish  Kolm  shale,  of  Cam- 
brian age,  Nevadan  Vanini  shale  of  Ordovician  age,  and  Colorado  Green  River 
shale  of  the  Eocene,  showed  that  this  method  is  capable  of  increasing  the 
organic  yields  from  these  processed  shales  very  significantly.  Values  as  high  as 
70  per  cent  of  the  total  organic  matter  present  in  the  Kolm  shales  were  ob- 
tained, in  contrast  to  only  1.6  per  cent  obtained  from  an  aliquot  sample  ex- 
tracted without  heating,  and  8  per  cent  from  another  sample  extracted  after 
heating  at  375°  C  for  5  hours  in  a  nitrogen  atmosphere.  Hydrogenation, 
therefore,  is  very  effective  in  rendering  organic  matter  more  extractable,  and 
may  have  an  important  application  in  the  difficult  problem  of  investigating 
organic  material  in  the  Precambrian  sediments.  The  approach  seems  attractive 
enough,  indeed,  to  encourage  further  investigation  of  its  applicability  to  studies 
of  organic  sediments  of  all  ages. 

Studies  particularly  concerned  with  tracing  the  pathways  of  synthesis  of 
proteins  and  nucleic  acids  in  microorganisms  have  been  under  way  in  the  bio- 
physics group  of  the  Department  of  Terrestrial  Magnetism  for  a  number  of 
years.  Recently  attention  has  been  focused  particularly  on  the  critical  im- 
portance in  metabolism  and  growth  of  the  elaborate  and  precise  fine  structure 
of  many  of  the  constituents  of  living  cells.  Among  these  constituents,  the  lipo- 
protein membranes  and  ribonucleoprotein  particles  which  together  comprise 
the  microsomal  fraction  seem  of  special  significance  and  interest.  Such  entities, 
which  range  from  molecular  weights  of  one  to  four  million,  appear  to  be  es- 
sential in  the  synthesis  of  protein.  They  are  under  intensive  investigation. 

It  already  seems  clear  that  particles  of  this  type,  obtained  from  the  bacterium 
Escherichia  coli,  are  complex  structures,  composed  of  ribonucleic  acid  and  a 
number  of  different  proteins  held  together  by  hydrogen  bonds.  One  of  these 
proteins  is  the  enzyme  ribonuclease,  which  is  inactive  while  held  within  the 
particle  but  when  released  by  breakage  of  the  hydrogen  bonding  can  exert  its 


enzymic  capacity,  hydrolyzing  the  nucleic  acid.  In  the  analytical  centrifuge 
the  particles  show  a  spectrum  of  sizes.  The  distribution  of  particles  within  that 
spectrum,  interestingly  enough,  depends  strongly  on  the  rate  of  cellular  growth. 
Bacteria  rapidly  growing  in  a  broth  medium  contain  more  of  the  smaller 
particles  than  bacteria  growing  in  a  glucose-salt  medium,  and  cells  that  have 
been  treated  with  chloramphenicol  to  halt  protein  synthesis  contain  mostly  one 
class  of  ribosome.  These  findings  suggest  that  the  different  sizes  may  corre- 
spond to  successive  stages  of  formation  of  the  particles  and  that  the  spectrum 
of  ribosome  size  distribution  can  serve  as  an  indicator  of  the  protein-synthesiz- 
ing ability  of  the  cell.  Studies  using  the  incorporation  of  tracer  molecules 
indicate  that  the  larger  particles  may  be  formed  by  incorporating  protein  and 
nucleic  acid  macromolecules.  These  early  results  emphasize  that  an  under- 
standing of  the  role  of  nucleoprotein  particles  may  constitute  a  major  step  in 
working  out  the  mechanisms  of  the  synthesis  of  proteins  and  nucleic  acids. 

A  very  different  approach  to  the  problem  of  protein  synthesis  has  been  pro- 
vided by  the  use  of  amino  acid  analogues.  It  will  be  recalled  that  last  year 
an  investigation  carried  out  in  collaboration  with  Dr.  Georges  N.  Cohen  of  the 
Institut  Pasteur  showed  that  selenomethionine  can  replace  methionine  and 
support  exponential  growth  in  a  methionine-requiring  mutant  of  the  bacterium 
E.  coli.  Study  of  such  analogues  has  been  continued  extensively  during  the  year 
in  collaboration  with  Dr.  Cohen  and  with  H.  de  Robichon-Szulmajeter  of  the 
National  Institutes  of  Health.  Particular  emphasis  has  been  laid  on  determining 
whether  the  analogues  are  contained  in  radically  different  molecular  species  or 
in  proteins  similar  to  those  normally  synthesized.  Such  investigation  requires 
analogues  that  will  substitute  for  only  one  naturally  occurring  amino  acid. 
Norleucine,  which  substitutes  for  methionine  in  the  proteins  of  E.  colt,  has 
proved  especially  useful  for  this  purpose. 

The  proteins  incorporating  the  analogue  appeared  to  be  only  slightly  altered. 
Furthermore,  analogues  are  incorporated  into  different  proteins  in  the  same 
proportion,  suggesting  that  the  mechanism  for  amino  acid  selection  does  not 
differ  from  one  protein  to  another.  With  some  analogues,  enzymes  retained 
their  usual  biological  activity;  with  others,  they  were  inactive.  This  observation 
suggests  that  the  amino  acid  complement  of  the  active  sites  of  an  enzyme  can 
be  examined  by  studying  the  sensitivity  of  the  enzyme  to  a  spectrum  of 
analogues.  Analysis  showed  that  each  of  the  many  proteins  that  could  be  re- 
solved by  ion-exchange  chromatography  incorporated  the  analogue  to  the  same 
extent.  There  thus  appear  to  be  no  differences  in  selectivity  among  the 
mechanisms  that  make  up  the  different  proteins.  Such  a  result  would  be  ex- 
pected if  the  selection  of  the  amino  acid  is  determined  by  the  order  of  the 
nucleotides  in  ribonucleic  acid. 


One  of  the  central  concerns  of  the  Department  of  Plant  Biology  has  long 
been  the  nature  and  mode  of  action  of  chlorophyll  as  it  occurs  in  living  plants — 
as  it  is  involved  in  that  economically  supremely  important  natural  reaction, 
photosynthesis.  Early  in  this  century  it  was  hoped  that  the  mode  of  action  of 
chlorophyll  in  photosynthesis  would  become  evident,  once  its  chemical  struc- 
ture was  established.  Time  has  proved  that  the  problem  is  far  more  complex 
and  subtle  than  this.  When  knowledge  of  structure  did  not  establish  an  ade- 
quate basis  for  comprehending  the  mechanism,  it  was  hoped  that  isolation  of 
functional  units  of  chlorophyll  and  examination  of  their  properties  would 
greatly  aid  an  understanding  of  chlorophyll  action.  It  turned  out  that  func- 
tional units  of  active  chlorophyll  with  simple  proportions  of  pigment  and  pro- 
tein were  not  isolable  from  mature  plants. 

A  much  more  fertile  line  of  attack  was  developed  a  number  of  years  ago 
by  James  H.  C.  Smith  at  the  Department  in  the  study  of  the  freshly  formed 
chlorophyll  complex.  This  chlorophyll-protein  complex,  known  as  the  chloro- 
phyll holochrome,  seems  to  be  a  definite  chemical  individual  of  the  protein- 
prosthetic  group,  isolable  in  particles  of  closely  uniform  size,  and  amenable 
to  study  outside  the  green  leaf.  Investigation  of  its  properties  has  occupied  an 
important  place  in  the  work  of  the  Department  for  a  number  of  years. 

The  search  for  an  explanation  of  the  very  different  properties  of  chlorophyll 
in  the  natural  complex  and  in  its  pure  form  continues  to  be  very  much  in  the 
foreground.  A  basic  problem  is  to  establish  the  range  of  variation  of  the  measur- 
able properties  of  the  natural  chlorophyll-^  complex.  How  many  recognizable 
forms  of  chlorophyll  a  exist  in  plants  ? 

It  is  quite  generally  believed  that  two  different  natural  forms  of  chlorophyll 
a  occur  together  in  most  green  plants.  Neither  has  been  isolated  as  a  chemical 
substance,  their  existence  being  deduced  from  the  relative  efficiency  of  photo- 
synthesis at  different  wavelengths  of  light.  The  derivative  spectrophotometer 
designed  and  built  several  years  ago  in  the  Department  by  C.  Stacy  French  and 
his  colleagues  has  proved  a  particularly  useful  tool  for  this  investigation.  An 
extensive  survey  has  been  undertaken  during  the  year  just  past  of  the  derivative 
absorption  spectra  of  chlorophyll  in  numerous  algae  and  other  plants.  The 
survey  seems  to  indicate  that,  contrary  to  the  common  assumption,  there  must 
be  more  than  two  forms  of  chlorophyll  a.  One  new  form,  very  different  from 
those  previously  recognized,  has  an  absorption  maximum  at  the  very  long 
wavelength  695  mfx.  It  is  found  in  old — but  not  in  young — cultures  of  the 
green  alga  Euglena.  It  appears  to  be  chlorophyll  a  combined  in  an  unusual 

Within  the  last  two  years  there  has  been  developed  in  several  laboratories 
a  physical  concept  of  the  mode  of  action  of  the  natural  chlorophyll  complex 
that  may  ultimately  prove  one  of  the  most  illuminating  to  our  understanding 


of  photosynthesis.  A  pioneer  in  the  development  of  these  ideas  has  been  Dr. 
William  Arnold  of  the  Oak  Ridge  National  Laboratories,  and  the  second  Fel- 
low of  the  Carnegie  Institution  under  the  program  made  possible  by  the 
Carnegie  Corporation.  The  concept  envisions  the  functional  units  within  a 
chloroplast  as  in  effect  photocells  whose  behavior  in  energy  trapping  and  energy 
transfer  can  be  considered  in  terms  of  the  theory  of  semiconductors.  Few  de- 
partures in  the  field  of  theoretical  photosynthesis  in  the  past  few  years  have  been 
as  radical,  as  provocative,  or  as  suggestive  as  this.  It  has  already  led  to  several 
kinds  of  new  experiments  and  provides  a  description  in  terms  less  specifically 
related  to  chemical  structure  than  other  models  that  have  been  attempted. 

The  investigations  of  Smith  on  the  photochemical  formation  of  chlorophyll 
in  leaves  from  its  precursor  protochlorophyll,  carried  on  for  many  years,  have 
been  actively  continued.  This  year  attention  has  been  focused  particularly 
on  the  amount  of  light  required  to  form  a  molecule  of  chlorophyll.  Proto- 
chlorophyll with  its  carrier  protein  can  now  be  removed  from  dark-grown 
leaves  in  active  form  and  purified  to  a  reasonable  degree.  The  absorption  co- 
efficients of  protochlorophyll  and  of  chlorophyll  a  are  known  within  a  few 
per  cent.  It  seemed  feasible,  therefore,  to  measure  the  efficiency  of  the  photo- 
chemical conversion  of  protochlorophyll  to  chlorophyll.  Extracts  of  proto- 
chlorophyll were  illuminated,  and  the  incident  light  absorbed  by  active  pro- 
tochlorophyll in  the  solution  was  calculated.  After  the  exposure  the  amount 
of  chlorophyll  formed  was  measured  spectroscopically.  The  yields  ranged  from 
about  0.5  to  0.7  molecule  of  chlorophyll  formed  per  quantum  of  light  absorbed, 
with  an  average  of  0.6  molecule.  It  is  tempting  to  think  that  two  quanta  of 
light  may  produce  one  molecule  of  chlorophyll.  Experiments  are  being  con- 
tinued and  methods  improved  to  obtain  more  precise  values. 

In  attempting  to  elucidate  the  situation  in  the  higher  green  plants  a  com- 
parison of  the  characteristic  chlorophyll  with  the  chlorophyll  in  the  purple 
bacteria  that  are  capable  of  photosynthesis  is  particularly  useful.  An  interesting 
investigation  of  the  reversible  oxidative  bleaching  of  chlorophyll  by  chemical 
treatment  and  by  light  in  several  species  of  photosynthetic  bacteria  has  been 
undertaken  during  the  year  by  J.  C.  Goedheer.  The  study  has  shown  that  the 
chemical  properties  of  the  pigment  are  markedly  influenced  by  its  incorpora- 
tion in  natural  structures.  Several  coexisting  forms  of  bacteriochlorophyll  have 
been  found  to  have  differing  chemical  reactivity.  The  form  that  absorbs  light 
at  the  longest  wavelength  (890  m(ji)  is  the  most  active.  This  is  the  form  that 
L.  N.  M.  Duysens  earlier  discovered  to  be  the  receiver  of  energy  absorbed  by 
other  pigments;  it  is  also  the  fluorescent  form.  The  fact  that  the  absorption 
bands  of  the  different  forms  of  chlorophyll  in  living  purple  bacteria  are  so 
much  more  widely  separated  than  in  the  green  plants  makes  them  peculiarly 
advantageous  as  a  reference  point  in  work  of  this  kind. 


Possibly  there  are  no  more  fundamental  aspects  of  theoretical  biology  than 
those  involved  in  the  study  of  genetic  mechanisms.  They  underlie  the  whole 
field,  and  so  must  be  of  essential  concern  to  each  of  the  departments  occupied 
with  biological  matters.  Questions  involving  the  structure  and  function  of  the 
materials  of  heredity  have,  however,  occupied  a  central  position  for  many  years 
in  the  research  of  the  Department  of  Genetics.  In  the  course  of  this  work  many 
different  organisms  have  been  and  are  currently  being  employed:  maize — the 
plant  which  more  than  fifty  years  ago  enabled  G.  H.  Shull  in  the  Department 
to  work  out  the  basic  principles  of  hybrid  vigor;  the  fruit  fly  Drosophila,  long 
a  classic  tool  of  geneticists ;  and  more  recently  and  more  especially,  those  most 
modern  tools  of  genetics,  the  usefulness  of  which  has  been  adequately  recog- 
nized and  exploited  only  in  very  recent  years — bacteria  and  viruses. 

In  the  field  of  virus  genetics,  Alfred  D.  Hershey  has  been  conducting  re- 
searches of  great  interest,  designed  especially  to  elucidate  the  mechanisms  of 
heredity  in  certain  strains  of  bacteriophages.  During  the  past  year  his  ex- 
periments, and  those  of  his  co-workers,  have  shown  that  both  genetic  re- 
combination and  chromosomal  replication  take  place  in  bacteriophages  in  the 
absence  of  protein  synthesis.  On  the  other  hand,  quantitative  analysis  of  the 
effects  of  ultraviolet  light  on  genetic  recombination  strongly  suggests  that  such 
recombination  is  directly  linked  to  the  synthesis  of  deoxyribonucleic  acid. 

For  several  years  M.  Demerec  and  his  group  have  been  deeply  concerned 
with  what  may  be  called  the  "fine  structure"  of  the  bacterial  chromosome,  as 
investigated  by  the  methods  of  biochemical  genetics.  It  is  now  possible  to 
analyze  the  structure  of  certain  segments  of  the  chromosome  concerned  with 
biochemical  syntheses  to  such  a  degree  of  resolution  as  to  distinguish  active 
"sites"  located  not  more  than  a  few  wavelengths  of  light  apart.  An  important 
test  object  in  this  research  has  been  the  bacterium  Salmonella  typhimurium, 
the  agent  of  mouse  typhoid.  Stocks  of  this  organism  have  been  constructed 
embodying  combinations  of  genetic  markers  mediating  the  metabolic  in- 
corporation of  a  number  of  amino  acids,  such  as  cystine  and  tryptophan.  Par- 
ticular use  has  been  made  of  the  phenomenon  of  transduction,  now  well  known.  • 
A  fragment  of  bacterial  genetic  material,  incorporated  within  an  infecting 
particle  of  bacteriophage,  can  be  carried  by  this  phage  to  a  second  bacterial  host 
and  there  incorporated  intact  into  the  new  genetic  complex.  It  is  frequently 
possible  for  the  larger  bacteriophages  to  carry  a  fragment  of  bacterial  chromo- 
some long  enough  to  include  a  number  of  linked  and  separately  identified 
metabolic  sites.  This  situation  offers  the  opportunity  of  presenting  the  receiv- 
ing bacterium  with  a  dilemma.  It  can  be  made  to  include  within  its  genome  a 
pair  of  sites — or  a  site  and  its  recessive  allelomorph — at  points  where  it  would 
normally  contain  but  one.  The  question  then  arises  which  of  the  two  alterna- 
tive sites  will  be  perpetuated  when  the  bacterium  divides :  which,  in  short,  will 
be  reproduced.  This  phenomenon,  known  as  "copy  choice,"  is  of  great  interest 


in  itself,  but  it  is  also  important  because  of  the  light  it  sheds  on  the  detailed 
mechanism  of  chromosome  replication. 

By  means  of  such  studies,  using  five  linked  genetic  markers  that  can  be  car- 
ried together  in  one  transducing  fragment,  Demerec  and  his  group  have  this 
year  obtained  information  suggesting  that  a  definable  copy-choice  mechanism 
operates  during  the  formation  of  the  recombinant  offspring  of  an  infected  bac- 
terium. According  to  their  interpretation,  recombination  occurs  at  the  time  of 
chromosome  replication,  when,  presumably,  a  transduced  fragment  is  closely 
synapsed  with  the  homologous  region  of  the  chromosome  of  the  recipient  bacte- 
rium. Chromosome  replication  proceeds  by  the  formation  of  copies  of  small  parts 
of  the  chromosome  and  the  joining  of  these  replicas  in  a  zipperlike  manner. 
In  the  duplicated  regions,  replicas  may  be  modeled  either  on  the  site  already 
present  in  the  bacterium  or  on  that  received  in  the  transducing  fragment.  But, 
once  a  choice  has  been  made — after  the  first  replica  of  one  member  of  a  dupli- 
cate set  has  been  formed — the  choices  for  subsequent  replicas  do  not  occur  at 
random.  There  is  a  higher  probability  that  consistency  will  be  maintained — a 
better  chance  that,  however  the  choice  was  made  on  the  first  replication,  so  it 
will  remain  thereafter.  The  experimental  data  indicate  that  the  frequency  of 
occurrence  of  nonconsistent  switches  depends  on  the  genetic  constitution  of  the 
chromosome  regions  involved.  These  findings  may  prove  of  great  importance 
to  our  understanding  of  the  detailed  mechanisms  by  which  that  most  funda- 
mental of  all  the  properties  of  life — the  duplication  of  chromosome  material — 

In  a  related  research  directed  to  a  slightly  different  end,  Allan  Campbell  has 
been  investigating  a  transduction  system  in  which  genes  of  the  bacterium 
Escherichia  coli  concerned  with  the  metabolism  of  galactose  are  carried  from 
one  cell  to  another  by  the  bacteriophage  lambda.  Earlier  work  had  indicated 
that  the  crucial  intermediate  in  this  process  is  a  hybrid  genetic  structure  con- 
taining information  derived  partly  from  the  phage  and  partly  from  the  bac- 
terial host.  The  work  of  the  past  year  has  strengthened  this  hypothesis.  An 
array  of  such  hybrid  structures  has  been  accumulated,  each  arising  from  a 
separate  primary  transductional  event,  and  they  have  been  found  to  differ 
significantly  in  the  exact  content  of  genetic  loci.  Thus  some  of  the  details  of 
the  whole  process  are  becoming  clearer. 

The  remarkable  program  of  Barbara  McClintock  in  the  genetics  of  maize, 
which  has  been  in  progress  for  a  number  of  years,  is  continuing,  with  new  and 
exceedingly  interesting  findings.  It  will  be  recalled  that  this  work  involved 
the  detection,  and  the  subsequent  investigation,  of  systems  of  elements  in  the 
cell  nucleus  that  control  the  action  of  particular  genetic  loci  without  themselves 
taking  a  specific  part  in  the  shaping  of  cell  function  or  morphology  in  the 
manner  classically  associated  with  the  action  of  genes.  Last  year  the  finding  of 


a  system  was  reported  which  controls  gene  action  at  two  known  loci,  not  di- 
rectly related  to  the  Ds~Ac  system  earlier  described. 

One  of  the  consequences  of  this  situation,  earlier  discussed,  is  that  the  ac- 
tivity of  a  particular  gene  may  vary  during  the  development  of  a  maize  plant. 
Thus  the  mature  plant  may  exhibit  entirely  different  patterns  of  expression  of 
the  same  gene  in  its  various  parts — a  matter  of  the  greatest  interest  to  those 
concerned  with  cell  differentiation  in  the  many-celled  embryo.  In  some  cases, 
both  the  pattern  type  and  the  inheritance  of  pattern  type  in  subsequent  progeny 
appear  to  be  very  irregular.  Analysis  of  one  such  case  this  year  has  revealed  the 
underlying  mechanism  clearly.  The  complexities  were  found  to  be  a  result  of 
alternating  cycles  of  action  and  inaction  in  one  particular  controlling  element, 
resulting  in  changes  in  expression  of  the  gene.  Further  analyses  of  actions  of 
this  type  can  be  of  the  widest  biological  importance. 

The  Department  of  Genetics  shares  with  the  biophysics  group  in  the  Depart- 
ment of  Terrestrial  Magnetism  an  intense  interest  in  the  intracellular  deoxy- 
ribonucleases.  A  proper  understanding  of  their  properties  seems  essential  to 
a  solution  of  general  problems  concerned  with  the  metabolism  of  deoxyribo- 
nucleic acid  and  its  role  in  cell  division.  During  the  year  Margaret  R. 
McDonald  has  continued  her  studies  directed  toward  the  isolation  and  char- 
acterization of  these  enzymes.  Last  year  she  discovered  that  salmon  testes  offer 
an  important  natural  source  of  experimental  material.  Her  efforts  this  year 
have  been  concentrated  on  ascertaining  the  best  procedures  for  extracting 
and  purifying  the  deoxyribonuclease  of  these  tissues.  An  enrichment  of 
activity  of  more  than  six  hundred  fold  (measured  in  units  per  milligram  of 
protein)  has  been  attained  without  any  appreciable  loss  in  total  enzymatic  ac- 
tivity. The  best  product  so  far  obtained,  however,  is  not  homogeneous.  It  is 
composed  of  at  least  three  proteins.  One  has  been  crystallized,  but  not  yet 
analyzed  for  biological  specificity.  Methods  are  at  present  under  study  to 
permit  still  further  purification  of  the  deoxyribonuclease  from  salmon  testes, 
in  anticipation  of  a  detailed  study  of  its  properties  and  its  mode  of  action. 

The  level  of  biological  organization  involving  the  combining  of  cells  into 
tissues,  and  the  problems  of  cell  differentiation,  regulation,  and  coordination 
that  they  involve,  are  the  special  concern  of  the  Department  of  Embryology. 
Research  in  the  Department  during  the  year  has  proceeded  most  actively  and 
on  so  many  fronts  that  only  a  few  can  be  cited  here. 

Countless  students  of  biology  have  been  intrigued  by  the  movements  of 
spermatozoa.  It  is  clear  that  motility  may  be  ascribed  to  processes  occurring 
in  the  sperm  tail,  involving  a  synchronous  cycle  of  contraction  and  relaxation. 
In  studying  the  basis  of  this  rhythmic  motility,  David  W.  Bishop  has  made 
what  may  prove  to  be  an  important  advance  in  the  discovery  that  the  mecha- 
nisms of  contractility  and  coordination  can  be  experimentally  disengaged,  thus 
permitting  separate  analyses  of  the  processes.   The  extraction  of  mammalian 


sperm  with  glycerine  divests  them  of  their  normal  properties  of  permeability 
and  irritability.  Simultaneously  they  lose  their  ability  to  beat  in  a  coordinated, 
spiral  manner.  Such  cell  models  beat  rhythmically,  but  only  in  one  plane. 
Their  only  forward  movement  is  slow  and  jerky,  and  occasionally  they  may 
even  move  backward.  But  the  two-dimensional  beat  is  vigorous  and  may 
actually  be  of  higher  frequency  than  that  of  the  normal  unextracted  sperm. 
These  extracted  cells  have  thus  retained  their  mechanism  for  contraction,  but 
have  lost  the  coordinating  mechanism  that  normally  provides  for  a  contraction 
sequence  in  adjacent  fibrils. 

On  the  other  hand,  unextracted  spermatozoa,  cells  with  relatively  long  and 
flexible  tails  like  the  sperm  of  the  common  squid,  can  be  slowed  down  by  dilu- 
tion of  the  sperm  mass  or  by  the  gradual  depletion  of  the  energy  reserves  of  the 
cell  and  the  substrates  that  supply  these  reserves.  In  one  sense  this  treatment 
"uncouples"  the  two  kinds  of  sperm  motility.  Stationary  cells,  and  even  those 
drifting  backward  in  the  medium,  can  rotate  rapidly  about  their  longitudinal 
axes,  owing  to  a  spiral  flagellation  of  the  tail  elements.  The  frequency  of  this 
type  of  fibrillar  motion  is  high,  and,  for  any  given  sperm,  the  direction  of 
rotation  remains  constant.  The  second,  more  violent,  but  less  rhythmic  beat 
in  these  retarded  sperm  is  a  two-dimensional  lashing  of  the  tail.  This  contrac- 
tion-relaxation cycle  has  a  time  constant  distribution  of  about  three  to  one,  with 
a  definite  pause  intervening  between  contractions.  The  experimental  separa- 
tion of  the  contraction-relaxation  mechanism  from  the  coordination  mechanism 
raises  a  number  of  provocative  questions  and  clearly  constitutes  a  problem  de- 
manding further  attention. 

James  D.  Ebert  has  recently  presented  an  analysis  of  progress  in  the  field  of 
immunoembryology  under  the  title  "The  acquisition  of  biological  specificity" — 
a  title  reflecting  one  of  the  principal  themes  of  research  in  the  Department. 
It  is  concerned  with  the  mechanisms  of  synthesis  and  interaction  of  specific 
macromolecules.  It  embraces  problems  of  tissue  specificity  and  individual 
specificity.  Although  tissue  specificity  may,  in  part,  reflect  differences  in  associa- 
tions and  numbers  of  macromolecules  which  are  alike  in  all  tissues,  immuno- 
chemical and  physiological  techniques  are  capable  of  distinguishing  molecular 
types  characteristic  of  each  tissue.  The  difference  between  two  individuals,  how- 
ever, is  based  on  more  than  the  difference  in  the  sum  of  their  tissue-specific  pro- 
teins, or  other  antigenic  molecules.  It  rests  more  broadly  on  the  existence  in 
each  individual  (or  in  all  individuals  of  identical  genotype,  such  as  identical 
twins  and  members  of  a  highly  inbred  strain)  of  specific  molecules  not  re- 
stricted to  certain  tissues  but  common  to  all,  or  most,  of  its  parts.  (A  third,  but 
less  secure,  assumption  states  that  similarly  there  are  antigens  common  to  all 
members  of  a  species,  the  species-specific  antigens.) 

In  recent  reports  of  the  Department,  emphasis  has  been  placed  on  studies  of 
tissue  specificity.   During  the  past  year  important  progress  was  recorded  in 


analyzing  the  chemistry  and  patterns  of  synthesis  of  the  contractile  proteins, 
including  actomyosin  and  tropomyosin,  of  hemoglobin,  and  of  the  retinal 
photopigments.  But  what  of  the  recognition  mechanisms  whereby  an  embryo 
develops  the  ability  to  distinguish  isoantigens  and  tissue-specific  antigens,  and 
to  make  antibodies  to  them  ? 

The  year  has  been  marked  by  several  promising  findings  in  this  area.  We 
know  that  exposure  of  embryos  late  in  their  development,  and  of  newborn  ani- 
mals, to  living  adult  cells  induces  a  state  of  tolerance,  for  when  these  animals 
are  challenged  with  living  homologous  cells  in  later  life  they  are  incapable  of 
producing  an  immune  reaction.  The  mechanism  for  producing  an  immune  re- 
action to  homografts  develops  only  in  the  late  embryonic  stages,  and  during 
this  period  it  is  subject  to  modification.  But  what  of  the  mechanisms  for  the 
production  of  antibodies  to  purified,  nonliving,  tissue-specific  antigens?  If 
tolerance  can  also  be  acquired  to,  say,  the  antigens  of  spleen,  liver,  or  heart, 
the  embryologist  has  a  powerful  tool,  for,  by  paralyzing  the  immune  mecha- 
nisms for  a  number  of  "common"  antigens,  he  should  be  able  to  enhance  the 
specificity  of  antiorgan  sera  of  his  choice.  During  the  year  Charles  Wyttenbach 
has  made  substantial  progress  toward  this  end.  Although  the  results  of  his  first 
large-scale  experiments  must  be  considered  tentative,  pending  amplification 
and  verification,  it  appears  that  tolerance  of  a  high  order  is  produced  by  the 
injection  of  newborn  rabbits  with  antigens  of  liver  and  spleen.  These  promis- 
ing experiments  are  being  continued. 

When  embryos  are  exposed  to  living  grafts  of  adult  tissues,  which  themselves 
have  the  capacity  to  produce  antibodies,  an  entirely  different  picture  emerges. 
Last  year  the  consequences  of  the  reaction  of  grafts  of  adult  spleen  and 
lymphoid  cells  against  animals  experimentally  deprived  of  the  ability  to  produce 
antibodies  were  described — the  so-called  graft-versus-host  reaction.  The  results 
were  clouded  by  the  inability  to  establish  that  the  host's  immune  mechanisms 
had  not  recovered  from  the  treatment.  Ebert  and  his  co-worker  Louis  E. 
DeLanney  believe  that  they  are  now  well  on  the  way  toward  providing  the 
critical  evidence  required.  Experiments  have  been  carried  out,  using  both  chick 
and  salamander  embryos,  in  which  grafts  of  adult  spleen  were  made  before 
the  host's  antibody-forming  tissues  had  begun  to  develop.  Well  in  advance  of 
the  critical  period  for  the  onset  of  immune  reactions  the  embryo  is  killed  as  a 
result  of  a  widespread  destruction  of  the  vascular  bed.  This  critical  program 
was  initiated  during  DeLanney's  tenure  as  a  Fellow  of  the  Carnegie  Institution 
of  Washington.  Continuing  long-range  cooperation  between  the  Department 
and  DeLanney  at  Wabash  College  is  most  gratifying. 

Another  significant  contribution  to  the  Department's  program  was  made 
during  the  year  by  a  visiting  investigator  from  New  Zealand's  University  of 
Otago  School  of  Medicine,  William  E.  Adams,  who  utilized  the  Bluntschli  Col- 
lection in  a  rewarding  study  of  the  development  of  the  adrenal  gland  and  the 


sympathetic  paraganglia  in  insectivores.  Especially  noteworthy  is  his  obser- 
vation that  the  paraganglionic  tissue  is  not  confined  to  the  adrenal  region  but 
extends  alongside  the  sympathetic  chain  as  a  distinctive  column,  reaching  to  the 
base  of  the  skull. 

Several  new  investigations,  begun  during  the  year,  have  proved  to  be  re- 
warding. Using  the  large  native  silkworm  moth,  Hyalophora  cecropia,  as 
experimental  material,  Hans  Laufer  has  made  a  study  of  factors  regulating  pro- 
tein synthesis  in  development.  Employing  immunochemical  techniques,  he 
has  shown  that  certain  antigenic  constituents  of  the  blood  of  the  insect  are  re- 
generated after  experimental  bleeding.  The  regenerated  antigens  are  proteins 
— and  he  has  been  able  to  demonstrate,  through  an  ingenious  combination  of 
immunochemical  and  histochemical  techniques,  that  at  least  one  of  them  has 
enzymatic  activity.  It  is  of  unusual  interest  that  an  enzyme  can  be  identified 
histochemically  while  in  combination  with  its  antibodies.  Suggestive  studies 
of  this  kind  pose,  and  attempt  in  some  part  to  answer,  the  question  why  some 
kinds  of  molecules  are  replaced,  but  not  others.  Their  primary  purpose  is  not 
to  make  a  chemical  inventory  of  the  proteins  and  their  changes  in  development 
(although  this  is  an  obvious  first  step  if  there  is  to  be  subsequent  progress)  but 
to  gain  further  insight  into  the  reasons  for  the  changes. 

The  final  level  of  biological  organization,  that  of  many-celled  individuals 
into  species  and  populations,  is  of  particular  interest  to  the  Department  of  Plant 
Biology,  whose  work  in  experimental  taxonomy,  extending  over  many  years, 
has  become  classic.  This  year  has  been  notable  for  these  programs.  There  has 
been  what  almost  amounts  to  the  initiation  of  a  new  field  of  investigation.  The 
year  has  also  seen  the  maturation  and  completion  of  work  in  an  older  one. 

For  many  years  the  program  of  experimental  taxonomy  has  been  concerned 
with  the  manner  in  which  populations  of  plants  become  adjusted  to  many  en- 
vironments (living  as  well  as  nonliving),  viewing  the  changes  they  undergo  as 
processes  of  genetic  transformation,  reflected  in  morphological  and  to  a  degree 
in  physiological  change.  The  emphasis  in  these  studies  of  plant  relationships 
has  progressed  in  recent  years  from  primarily  cytogenetic  and  transplant  in- 
vestigations to  those  in  the  area  of  comparative  physiology.  Why  some  plants 
thrive  in  a  particular  environment  while  other,  closely  related  forms  require  a 
very  different  climate,  even  for  survival,  has  become  one  of  the  main  questions 
under  study. 

Different  species  and  strains  of  the  monkey  flower  Mimulus  have  been  found 
particularly  suitable  for  comparison  of  physiological  behavior  as  measured  in 
the  laboratory.  Not  only  are  they  able  to  grow  in  the  contrasting  environments 
of  the  three  experimental  gardens  of  the  Department  at  Palo  Alto,  Mather,  and 
Timber  line,  but  their  size  and  adaptability  make  them  excellent  subjects  for 
laboratory  investigation.  A  number  of  climatic  races  of  Mimulus  plants  are 
being  grown  at  the  three  altitude  stations  as  material  for  laboratory  studies 


of  the  way  in  which  their  rates  of  photosynthesis  and  respiration  vary  in  re- 
sponse to  changes  of  temperature  and  light  intensity.  Recorded  patterns  of 
photosynthesis  and  respiration  from  these  climatic  races,  taken  under  controlled 
conditions  of  light  and  temperature,  will  be  compared  with  the  observed 
morphological  growth  responses  of  the  same  plants  at  the  different  altitudes  and 
in  experimental  growth  chambers  in  which  temperature,  light,  and  humidity 
can  be  accurately  controlled. 

The  rates  of  photosynthesis  and  respiration  of  clones  of  Mimulus  taken 
originally  from  diverse  latitudes  and  altitudes  have  been  explored  in  a  pre- 
liminary survey.  The  results  obtained  to  date  indicate  marked  differences 
among  them.  The  interpretation  of  these  differences  in  terms  of  climatic  races 
and  their  significance,  if  any,  in  natural  selection  in  different  environments  is 
an  ultimate  objective  of  the  investigations. 

In  segregating  progenies  of  the  cross  of  two  species  of  Mimulus  taken  from 
different  altitudes  (M.  cardinalis  and  M.  lewisii)  the  frequencies  of  certain 
flower  colors  were  found  to  vary  greatly  with  the  altitude  at  which  the  prog- 
enies had  been  established.  At  high  elevations  types  resembling  the  alpine 
parent  were  predominant,  while  types  resembling  the  lowland  parent  were 
favored  in  the  milder  climates.  There  appears  to  be  a  genetic  linkage  between 
floral  characters  and  the  physiological  characters  that  determine  survival  under 
extreme  conditions. 

Studies  have  been  begun  of  the  germination  of  seedling  populations  of  con- 
trasting parental  and  hybrid  lines  of  Mimulus  under  crossed  gradients  of  two 
controlled  variables — temperature  and  light  intensity.  This  technique  is  also 
being  used  to  compare  differences  in  germination  capacity  in  races  from  con- 
trasting habitats  and  in  their  segregating  F2  progeny.  Individuals  among  the 
segregates  which  appear  to  differ  in  their  temperature  and  light  requirements 
may  be  selected  and  subjected  to  further  field  and  laboratory  tests. 

This  year  marks  the  completion  in  the  Department  of  the  range-grass  pro- 
gram described  in  1954,  involving  a  series  of  breeding  and  testing  experiments 
undertaken  in  cooperation  with  the  Agricultural  Research  Service  of  the  De- 
partment of  Agriculture.  They  were  designed  to  test  the  responses  of  apomictic 
strains  of  key  parental  and  hybrid  bluegrasses  in  widely  divergent  climates  at 
various  experiment  stations  strategically  located  throughout  the  United  States. 
The  results  illustrate  to  a  striking  degree  the  climatic  specificity  of  the  regional 
responses  of  the  apomictic  strains.  This  is  the  first  time  that  such  a  series  of 
widely  distinct  parental  species  and  their  stabilized  apomictic  hybrid  derivatives 
have  been  systematically  studied  over  a  great  range  of  climates  on  what  amounts 
to  a  nation-wide  basis.  These  tests  therefore  are  of  general  biological  as  well  as 
of  more  specific  agronomic  interest.  The  extraordinary  wealth  of  information 
accumulated  in  the  Department  about  the  grasses  of  the  genus  Poa  is  being 
analyzed  and  summarized  preparatory  to  publication. 


LOSSES  .  .  . 

The  death  of  Ernest  Orlando  Lawrence  and  of  Howard  E.  Tatel  during  the 
year  brought  losses  to  the  Institution  which  can  never  be  repaired.  Dr.  Lawrence 
had  served  as  a  Trustee  of  the  Institution  for  fourteen  years.  Dr.  Tatel  had  con- 
ducted distinguished  research  in  geophysics  and  seismology  in  the  Department 
of  Terrestrial  Magnetism,  where  he  was  chairman  of  the  earth  physics  section, 
for  the  past  ten  years. 

Dr.  Lawrence  was  born  in  Canton,  South  Dakota,  on  August  8,  1901.  He 
attended  the  University  of  South  Dakota,  from  which  he  was  graduated  in  1922, 
taking  his  master's  degree  at  the  University  of  Minnesota  and  his  doctorate  at 
Yale,  where  he  became  an  assistant  professor.  In  1928  he  moved  to  the  Univer- 
sity of  California. 

It  was  in  these  years  that  Lawrence  conceived  the  idea  of  the  cyclotron, 
which  can  be  regarded  as  having  initiated  the  modern  era  of  experimental  in- 
vestigation in  nuclear  physics  and  for  which — together  with  the  great  program 
of  research  that  accompanied  the  development  and  was  made  possible  by  it — he 
was  awarded  the  Nobel  prize  in  1939.  In  1936  he  established  the  Radiation 
Laboratory  at  the  University  of  California,  where  his  research  was  continued 
and  of  which  he  served  as  Director  for  the  remainder  of  his  life.  That  Labora- 
tory created  and  continues  to  occupy  a  unique  position  in  the  whole  fabric  of 
American  research  in  the  physical  sciences. 

During  World  War  II  and  thereafter  Dr.  Lawrence  rendered  services  to  the 
nation  that  can  perhaps  never  be  adequately  estimated,  especially  in  the  fields 
of  atomic  development,  of  national  self-defense,  and  of  international  relations. 
Death  came  at  the  close  of  a  particularly  arduous  period  of  public  service,  as 
a  representative  of  the  United  States  at  the  international  conference  on  scientific 
detection  of  nuclear  explosions  held  at  Geneva  during  the  current  summer. 

Dr.  Tatel  died  on  November  15,  1957,  in  Washington,  D.  C.  He  had  just  re- 
turned from  the  seismic  expedition  of  the  Department  of  Terrestrial  Magnetism 
to  the  high  Andes.  He  was  born  in  New  York  City  on  December  22,  1913. 
He  attended  the  Massachusetts  Institute  of  Technology,  where  he  took  both  his 
undergraduate  and  his  master's  degrees,  and  did  his  doctoral  work  at  Stanford 
University.  He  spent  the  following  two  years  as  a  research  associate  in  nuclear 
physics  at  the  University  of  Michigan. 

During  World  War  II  Dr.  Tatel  was  engaged  in  research  and  development 
work  at  the  Applied  Physics  Laboratory  of  the  Johns  Hopkins  University, 
where  he  pioneered  in  the  development  of  the  proximity  fuse  and  in  the  early 
developmental  phase  of  ram  jet  propulsion.  In  1947  he  came  to  the  Department 
of  Terrestrial  Magnetism,  where  his  research  in  seismology  and  geophysics  was 
of  great  importance.  Especially  noteworthy  were  his  field  work  with  colleagues 
of  the  Institution  on  the  thickness  of  the  earth's  crust,  and  his  laboratory  model 


work,  which  provided  illuminating  analyses  of  seismic  phenomena.  In  recent 
years  he  devoted  much  attention  to  the  development  of  special  equipment  for 
radio  astronomy  and  to  measuring  the  hydrogen  clouds  of  our  Galaxy.  Dr. 
Tatel  will  be  long  remembered  by  his  associates,  not  only  for  his  high  profes- 
sional accomplishments,  but  even  more  for  the  quality  which  his  personality 
brought  to  everyone  associated  with  him  and  to  every  enterprise  in  which  he 
was  engaged. 

Dr.  Walter  Baade,  astronomer  at  the  Mount  Wilson  and  Palomar  Observa- 
tories, retired  on  June  30,  1958,  completing  a  career  of  research  embodying 
contributions  to  astronomy  of  the  highest  order. 

On  coming  to  the  Mount  Wilson  Observatory  in  1931,  Dr.  Baade  first  at- 
tacked a  series  of  photometric  problems,  including  the  establishment  of  photo- 
metric standards  in  Selected  Areas,  investigations  of  the  light-curves  of  super- 
novae,  and  studies  of  the  magnitudes  of  variables  in  globular  clusters  and  in 
galaxies  and  their  use  for  distance  determinations.  During  World  War  II  he 
took  advantage  of  the  darker  skies  consequent  on  the  black-out  in  the  Los 
Angeles  area  and  of  the  development  of  faster  red-sensitive  plates  to  undertake 
those  critical  galactic  investigations  that  led  to  the  formulation  of  the  concept 
of  Population  I  and  Population  II  stars  discussed  elsewhere  in  this  report ;  with 
colleagues  he  made  detailed  studies  of  the  color-magnitude  relationships  in 
representative  samples  of  the  two  populations.  The  results  of  these  investiga- 
tions provided  the  observational  basis  for  present  theories  of  stellar  evolution. 

With  the  completion  of  the  200-inch  Hale  telescope  Baade  continued  his 
studies  of  the  Andromeda  galaxy  and  the  other  members  of  the  local  group, 
paying  special  attention  to  cepheid  variables  and  other  stellar  types  that  might 
be  used  as  indicators  for  fixing  the  distances  of  these  objects.  This  work  repre- 
sented the  first  step  in  the  precise  determination  of  the  distances  of  all  objects 
outside  the  Milky  Way.  The  observations  provided  much  of  the  evidence  for  a 
revision  of  the  absolute  magnitudes  of  the  cepheid  variables  and  a  resultant 
increase  in  the  distance  scale  of  all  objects  outside  our  Galaxy  by  a  factor  of 
nearly  3. 

In  collaborative  work  with  Dr.  Minkowski,  Dr.  Baade  also  identified  many 
of  the  celestial  radio  sources  with  optically  observed  objects  and  contributed 
much  to  the  physical  interpretation  of  the  nature  of  these  sources.  From  1953 
until  his  retirement,  Dr.  Baade  served  on  the  Observatory  Committee  of  the 
Mount  Wilson  and  Palomar  Observatories. 

.  .  .  AND  GAINS 

The  fellowship  program  in  the  natural  sciences,  made  possible  by  a  generous 
gift  from  the  Carnegie  Corporation  of  New  York  and  described  in  last  year's 
report,  has  brought  to  the  Institution  a  group  of  distinguished  investigators. 


During  the  year  fellowships  were  awarded  to  Dr.  Hessel  de  Vries  of  the  Uni- 
versity of  Leiden,  Dr.  David  G.  Catcheside  of  the  University  of  Birmingham, 
Dr.  Mogens  Westergaard  of  the  University  of  Copenhagen,  and  Dr.  Evelyn 
E.  B.  Smith  of  the  University  of  Glasgow. 

It  is  a  special  pleasure  to  report  that  Dr.  Vannevar  Bush,  retired  President  of 
the  Institution,  received  the  New  England  award  of  the  Engineering  Societies 
of  New  England  on  November  12,  1957.  The  award  is  presented  annually  to 
an  engineer  in  New  England  who  "merits  recognition  of  his  accomplished 
work  as  well  as  his  character." 

Dr.  George  W.  Corner,  the  former  Director  of  the  Department  of  Embryol- 
ogy, received  the  Passano  Foundation  award  of  the  American  Medical  Associa- 
tion at  its  convention  in  San  Francisco  on  June  25,  1958.  The  award  was  made 
in  recognition  of  "his  long  and  continuing  researches  and  for  many  fruitful 
contributions  to  the  better  understanding  of  mammalian  anatomy  and  physiol- 
ogy, with  particular  emphasis  on  human  reproduction."  Both  the  University 
of  Chicago  and  the  Woman's  Medical  College  of  Philadelphia  conferred  hon- 
orary degrees  on  Dr.  Corner  during  1958. 

It  gives  me  much  pleasure  to  announce  the  following  honors  that  have  been 
received  during  the  year  by  the  members  of  the  staff  of  the  Institution : 

Dr.  Horace  W.  Babcock,  the  assistant  director  of  the  Mount  Wilson  and 
Palomar  Observatories,  was  presented  with  the  Henry  Draper  medal  of  the 
National  Academy  of  Sciences  on  April  28,  1958,  for  his  "original  and  out- 
standing work  leading  to  the  discovery  of  magnetic  fields  in  stars  and  also  the 
general  magnetic  field  of  the  sun."  He  was  also  awarded  the  Eddington  medal 
of  the  Royal  Astronomical  Society  for  his  research  on  the  magnetic  fields  of 
early-type  stars  and  of  the  sun. 

Dr.  Allan  Sandage,  astronomer  at  the  Observatories,  received  the  Helen 
Warner  prize,  given  by  the  American  Astronomical  Society  for  outstanding  re- 
search by  younger  members  of  the  Society.  The  prize  was  given  especially  for 
his  investigations  of  the  extragalactic  distance  scale. 

Dr.  Hatten  S.  Yoder,  petrologist  at  the  Geophysical  Laboratory,  and  Dr. 
Alfred  D.  Hershey,  microbiologist  at  the  Department  of  Genetics,  were  elected 
to  membership  in  the  National  Academy  of  Sciences  on  April  29,  1958. 

Dr.  Barbara  McClintock,  cytogeneticist  at  the  Department  of  Genetics,  re- 
ceived the  degree  of  Doctor  of  Science,  honoris  causa,  from  Smith  College  on 
June  8,  1958.  She  also  received  on  August  28,  1957,  a  Certificate  of  Merit  from 
the  Botanical  Society  of  America  "in  recognition  of  distinguished  achievement 
in  and  contributions  to  the  advancement  of  botanical  science." 

Caryl  P.  Has\ins 












Operated  by  Carnegie  Institution  of  Washington 

and  the  California  Institute  of  Technology 

Pasadena,  California 

IRA  S.  BOWEN,  Director 
HORACE  W.  BABCOCK,  Assistant  Director 


Ira  S.  Bowen,  Chairman 
Walter  Baade  x 
Horace  W.  Babcoek 

Robert  S.  Bacher 
Jesse  L.  Greenstein 
Earnest  C.  Watson 

1  Replaced  by  Rudolph  Minkowski,  July  1,  1958. 



Introduction    51 

Observing   Conditions    56 

Solar  Observations 56 

Solar  photography  56 

Sunspot  activity   56 

Magnetic  polarities    57 

Solar  magnetic  fields 57 

Corpuscular  radiation  58 

Planets   58 

Rotation  of  Venus 58 

Comets 59 

Stellar  Spectroscopy 59 

Chemical    composition    of    stellar    at- 
mospheres      59 

Carbon  stars    60 

Blue  stars  in  the  galactic  halo 60 

Central  star  of  the  Crab  Nebula 60 

Line  blanketing   61 

Variable  stars   61 

Magnetic  stars    62 

Mass  loss  from  late-type  giants 62 

The  relationship  of  H  and  K  emission 

to  absolute  magnitude 64 

Radial  velocities  64 

Gaseous  Nebulae  64 

Internal  motions 64 

Radial  velocities   65 

Densities  of  nebulae 66 

Expanding   shells   around   novae   and 

supernovae    66 

Globular  and  Galactic  Clusters 67 

White  dwarfs  in  M  67 70 

Absolute  magnitudes  of  cepheids 70 

Dynamics  of  clusters 70 

Galaxies   71 

The  local  group 71 

Rotation  and  dispersion  of  stellar  ve- 
locities      72 

Magnitudes  and  energy  curves  of  gal- 
axies     72 

Distance  scale  73 

Velocities  of  galaxies 74 

Supernovae    74 

Density  of  gas  clouds 74 

Catalogue  of  galaxies  and   cluster  of 

galaxies    75 

Multiple  galaxies   75 

Spectra  of  intergalactic  luminous  mat- 
ter    75 

Radio  Sources 76 

Theoretical  Studies   77 

Instrumentation    77 

Guest  Investigators    79 

Staff  and  Organization 83 

Bibliography   85 

Carnegie  Institution   of  Washington  Year  Boo\  57,  1957-1958 


The  Hale  200-inch  telescope  on  Palo- 
mar  Mountain  was  dedicated  on  June  3, 
1948,  and  the  agreement  for  the  joint  op- 
eration of  the  Mount  Wilson  and  Palomar 
Observatories  became  effective  on  April  1, 
1948.  As  the  tenth  anniversaries  of  both 
these  events  fall  in  the  present  report  year, 
it  is  appropriate  to  review  briefly  the  ob- 
servational programs  carried  out  during 
this  first  decade  of  joint  operation. 

Between  the  two  World  Wars  Hubble, 
using  the  100-inch  on  Mount  Wilson,  was 
able  to  make  preliminary  measurements 
of  the  distances,  diameters,  and  luminosi- 
ties of  the  galaxies  and  thereby  to  provide 
definite  evidence  that  the  galaxies  are  huge 
systems  of  stars  similar  to  our  Milky  Way, 
and  the  major  units  in  which  the  mass  of 
the  universe  is  distributed.  One  of  the 
major  projects  planned  for  the  Hale  tele- 
scope was  a  thorough  restudy  of  the 
near-by  galaxies,  in  particular  a  re-exami- 
nation of  the  procedure  for  fixing  their 
distances  in  order  to  eliminate  the  very 
large  uncertainties  that  were  known  to 
exist  in  these  initial  measurements.  In 
general  this  procedure  consists  of  the  com- 
parison of  the  apparent  brightness  of  some 
distance  indicator,  such  as  a  star,  in  the 
galaxy  with  its  absolute  brightness  or  lu- 
minosity as  determined  from  near-by  ex- 
amples in  our  own  Galaxy.  The  inverse- 
square  law  for  the  falling  oflF  of  apparent 
brightness  with  distance  then  gives  the 
required  answer.  For  use  as  such  a  dis- 
tance indicator,  the  star  or  other  object 
must  be  unusually  bright  so  that  it  can  be 
observed  at  large  distances,  and  it  must 
have  some  characteristic  by  which  it  can 
be  identified  as  a  particular  type  of  star 
or  object  having  the  same  absolute  bright- 
ness as  the  examples  measured  in  our  own 
system.  In  his  original  measurements 
Hubble  used  cepheid  variables  as  the  chief 
distance  indicator,  since  they  could  be 
identified  by  their  light-fluctuations. 

In  1952  Baade,  from  observations  with 

the  200-inch,  Thackeray  and  Wesselink, 
from  observations  of  the  Magellanic 
Clouds  at  Pretoria,  and  Mineur  as  well  as 
Blaauw  and  H.  R.  Morgan,  from  a  theo- 
retical study  of  proper  motions  of  cepheids 
in  our  own  Galaxy,  independently  arrived 
at  the  conclusion  that  the  classical  cephe- 
ids are  about  1.5  magnitudes  brighter  than 
the  value  assumed  in  the  earlier  studies. 
A  careful  remeasurement  was  also  made 
of  the  apparent  magnitudes  of  a  substan- 
tial number  of  cepheids  in  several  of  the 
galaxies  close  enough  for  their  observa- 
tion. Since  the  absolute  luminosity  of  a 
cepheid  variable  is  a  function  of  its  period, 
it  is  necessary  to  accumulate  several  dozen 
plates  of  each  field,  distributed  over  a 
sufficient  length  of  time  to  give  a  reliable 
light-curve  of  each  variable.  For  this  pur- 
pose Baade  obtained  sets  of  plates  for  four 
fields  in  the  Andromeda  galaxy  and  for 
several  other  galaxies  in  the  local  group, 
including  NGC  185,  NGC  205,  and  the 
dwarf  galaxies  in  Draco,  Ursa  Minor, 
and  Leo  I  and  II. 

At  the  same  time  steps  were  taken  to 
investigate  the  properties  of  other  unusu- 
ally bright  objects  that  might  be  used  as 
distance  indicators  to  check  the  distances 
found  from  the  cepheids  and,  if  possible, 
to  extend  the  measures  to  more  distant 
galaxies.  Hubble  and  Sandage  measured 
the  brightest  stars  in  the  Andromeda  gal- 
axy, including  the  very  bright  irregular 
variables.  Sandage  investigated  the  inte- 
grated magnitude  of  the  gaseous  nebulae 
or  H  II  regions  surrounding  several  of 
these  very  luminous  stars.  Code  and  Dr. 
T.  E.  Houck  have  compared  the  excep- 
tionally bright  blue  stars  in  NGC  6822, 
M  33,  the  Large  Magellanic  Cloud,  and 
our  own  Galaxy.  Arp  made  a  very  ex- 
tensive study  of  the  magnitudes,  light- 
curves,  and  frequency  of  occurrence  of 
ordinary  novae  in  the  Andromeda  galaxy. 
For  this  purpose  fields  in  the  galaxy  were 
photographed  with  the  60-inch  on  every 



clear  moonless  night  during  two  observ- 
ing seasons.  A  total  of  thirty  novae  was 
found  during  this  period.  Measurements 
of  the  distribution  in  magnitude  of  the 
globular  clusters  in  the  Andromeda  galaxy 
were  made  by  Baum.  Humason  and  San- 
dage  made  observations  of  the  red  super- 
giants  in  M  33.  Hubble,  Sandage,  and 
Baade  observed  various  distance  indicators 
in  the  more  distant  galaxies  outside  the 
local  group,  including  M  51,  members  of 
the  M  81  and  M  101  groups  of  galaxies, 
and  the  galaxies  in  the  Ursa  Major  cloud 
and  the  Virgo  cluster. 

Finally,  it  was  necessary  to  recalibrate 
the  magnitudes  of  the  stars  in  Selected 
Areas  that  had  been  used  as  standards  of 
magnitude  for  the  very  faint  stars  in  these 
galaxies.  This  recalibration  with  photo- 
electric techniques  was  started  by  Drs. 
J.  Stebbins  and  A.  E.  Whitford  of  the 
Washburn  Observatory  using  the  100-inch 
telescope,  and  was  completed  by  Baum 
with  the  200-inch.  These  observers  found 
that  the  earlier  photographically  deter- 
mined magnitudes  of  very  faint  stars  were 
too  bright  by  as  much  as  1  magnitude. 

When  these  corrections  are  made,  the 
Andromeda  galaxy  turns  out  to  be  nearly 
3  times  as  distant  as  was  indicated  by  the 
earlier  measurements.  Likewise,  the  di- 
ameter and  mass  of  this  galaxy  must  be 
increased  by  a  similar  factor,  and  the  lu- 
minosity by  a  factor  of  about  8.  The  ap- 
plication of  the  new  values  of  the  magni- 
tudes of  various  distance  indicators  to  the 
preliminary  observations  of  the  more  dis- 
tant galaxies  points  to  an  even  larger  cor- 
rection to  the  distances  and  dimensions  of 
these  objects.  Thus  Sandage  estimates  that 
the  old  distance  of  the  Virgo  cluster  and  all 
more  distant  galaxies  should  be  increased 
by  a  factor  of  between  5  and  10.  If  further 
measurements  confirm  this  estimate,  the 
size  of  the  observable  universe  must  be 
increased  by  the  same  factor. 

During  World  War  II  Baade  had  ob- 
served with  the  100-inch  that  the  most 
conspicuous  stars  in  the  nucleus  of  the 
Andromeda   galaxy   and   in   the  elliptical 

galaxies  are  red  giants  whereas  in  the 
spiral  arms  of  the  Andromeda  galaxy  and 
in  the  neighborhood  of  the  sun  in  our 
own  Galaxy  the  brightest  stars  are  blue. 
This  observation  led  to  the  concept  of 
population  types,  the  stars  in  our  own 
neighborhood  being  designated  as  Popu- 
lation I  and  those  in  the  galactic  nuclei 
and  in  the  elliptical  galaxies  as  Population 
II.  This  concept  raised  many  new  ques- 
tions as  to  the  cause  of  the  different  char- 
acteristics of  the  two  population  types.  It 
also  introduced  many  new  uncertainties 
into  the  distance-scale  problem  by  ques- 
tioning the  validity  of  the  assumption  that 
the  magnitude  of  a  given  stellar  type  used 
as  a  distance  indicator  is  the  same  in  all 
objects.  This  last  point  was  especially  seri- 
ous, since  in  many  of  the  older  measure- 
ments the  absolute  magnitude  was  deter- 
mined from  examples  in  one  population 
and  the  apparent  magnitude  in  a  distant 
object  from  examples  of  the  other  popu- 

An  extensive  study  of  the  properties  of 
stars  of  the  two  population  types  was 
therefore  instituted.  Measurements  of  the 
color-magnitude  relationships  were  made 
for  various  groups  of  stars,  including  sev- 
eral globular  clusters  and  Galactic  clusters, 
observed  by  Arp,  Baum,  Osterbrock,  San- 
dage, Schmidt,  and  Walker,  and  a  few 
members  of  the  local  group  of  galaxies  in- 
vestigated by  Baade  and  Sandage.  Studies 
were  also  made  of  the  period-luminosity 
function  of  cepheid  and  cluster-type  vari- 
ables in  these  objects. 

From  these  studies  the  picture  soon 
emerged  that  Population  I  is  characterized 
by  very  young  stars  that  have  only  re- 
cently condensed  from  clouds  of  dust  and 
gas.  This  assumption  was  confirmed  by 
the  observation  that  stars  with  Population 
I  characteristics  are  invariably  associated 
with  such  gas  and  dust  clouds.  On  the 
other  hand  Population  II  contains  only  old 
stars.  Further  confirmation  of  this  picture 
was  obtained  from  parallel  theoretical  and 
laboratory  developments  which  showed 
that  the  energy  radiated  by  the  stars  comes 


from  the  transformation  of  hydrogen  into 
helium  and  the  heavier  elements  in  the  hot 
stellar  core.  The  detailed  analysis  of  the 
nuclear  processes  occurring  in  these  stars 
was  developed  by  Hoyle  and  G.  R.  Bur- 
bidge  in  collaboration  with  Drs.  W.  A. 
Fowler  and  E.  M.  Burbidge  of  the  Kellogg 
Radiation  Laboratory.  The  effect  of  the 
depletion  of  the  hydrogen  fuel  on  the 
characteristics  of  the  star  was  investigated 
theoretically  by  Hoyle  and  by  Dr.  Martin 
Schwarzschild  and  others. 

In  general,  when  a  mass  of  predomi- 
nantly hydrogen  gas  condenses  into  a  star 
on  the  main  sequence,  the  luminosity  and 
surface  temperature  on  this  sequence  are 
fixed  by  the  mass  condensed:  the  luminos- 
ity varies  approximately  as  the  cube  of  the 
mass,  and  the  surface  temperature  in- 
creases slowly  but  continuously  with  the 
mass.  Because  of  the  rapid  expenditure  of 
the  hydrogen  fuel  by  the  very  massive 
stars  their  fuel  soon  becomes  exhausted. 
As  depletion  of  the  fuel  progresses  the  star 
expands  and  its  surface  cools  while  the 
total  radiation  increases;  that  is,  the  star 
leaves  the  main  sequence  at  the  critical 
stage  of  fuel  depletion.  When  the  fuel  is 
finally  exhausted  the  radiation  rapidly  falls 
off  and  the  star  probably  ends  its  career  as 
a  faint  white  dwarf. 

On  this  picture  a  young  group  of  stars 
is  characterized  by  a  continuous  progres- 
sion of  properties  from  the  massive,  ex- 
tremely luminous,  blue  stars  to  the  reddish 
faint  dwarf  stars.  As  the  group  becomes 
older  the  very  luminous  blue  stars  use  up 
their  fuel  and  drop  out,  and  the  brightest 
stars  in  such  a  group  are  the  stars  of 
moderate  initial  luminosity  which  because 
of  approaching  depletion  of  fuel  have  ex- 
panded and  cooled  of?  but  have  become 
temporarily  more  luminous.  Such  a  stellar 
group,  a  few  billion  years  old,  should  have 
the  distribution  in  luminosity  and  color 
found  to  exist  among  the  globular  clusters 
and  the  elliptical  galaxies. 

The  discovery  of  the  role  played  by  nu- 
clear transformations  in  the  energy  pro- 
duction and  the  evolution  of  the  stars  gave 

an  added  interest  to  spectroscopic  studies 
of  abundances  of  the  elements  partici- 
pating in  these  transformations  in  stars 
of  various  ages.  Fortunately  the  great 
light-gathering  power  of  the  200-inch  and 
the  efficiency  of  its  spectrographs  made 
possible  the  study  of  the  faint  stars  such 
as  the  globular-cluster  stars  and  the  white 
dwarfs  which  are  of  special  significance  in 
evolutionary  theory. 

Deutsch,  Greenstein,  and  Olin  Wilson 
carried  out  extensive  observations  of  the 
old  stars  in  several  globular  clusters  that 
are  typical  representatives  of  Population 
II.  In  general  these  studies  indicated  that 
the  metal  content  of  the  Population  II 
stars  is  lower  than  that  of  the  Population  I 
stars,  although  the  abundance  of  the  met- 
als in  different  clusters  appears  to  range 
from  nearly  equality  to  that  of  Population 
I  in  some  clusters  down  to  less  than  1  per 
cent  of  this  in  other  clusters  such  as  M  92. 
Similarly,  although  the  color-luminosity 
curves  of  these  clusters  in  general  follow 
the  pattern  predicted  by  theory,  small  un- 
explained differences  between  different 
clusters  of  nearly  the  same  age  are  also 

Greenstein  also  made  spectroscopic 
analyses  of  some  dozens  of  the  faint  white 
dwarfs,  which,  as  mentioned  above,  prob- 
ably represent  the  final  stage  of  evolution 
of  a  star.  He  found  striking  differences  in 
their  spectra  and  presumably  in  the  com- 
position of  their  surface  atmospheres. 
Greenstein  also  investigated  the  differences 
in  chemical  composition  between  dwarfs, 
subgiants,  and  giant  stars.  The  composi- 
tions of  stars  of  high  and  low  velocity 
were  also  compared,  as  there  is  evidence 
that  they  may  be  representatives  of  the 
two  population  types. 

Deutsch  has  investigated  the  processes 
by  which  the  massive  stars,  when  in  the 
expanded  red-giant  stage,  eject  into  space 
an  appreciable  fraction  of  their  mass. 
These  processes  provide  a  probable  ex- 
planation for  the  large  differences  between 
the  initial  mass  of  the  blue  giant  stars  and 
their  final  mass  in  the  white-dwarf  stage. 


At  the  same  time  they  explain  the  pres- 
ence of  observable  amounts  of  the  heavy 
elements  in  the  interstellar  gas. 

One  phase  of  the  study  of  galaxies  initi- 
ated in  the  1920's  and  1930's  with  the  100- 
inch  was  the  measurement  of  their  radial 
velocities  by  Humason  and  the  discovery 
by  Hubble  of  the  linear  relationship  be- 
tween these  velocities  and  the  distances. 
On  its  completion  Humason  used  the  200- 
inch  telescope  and  its  prime-focus  spectro- 
graph to  measure  the  velocities  of  a  sub- 
stantial number  of  additional  galaxies,  and 
in  particular  to  extend  the  measurements 
out  to  objects  having  a  velocity  of  reces- 
sion about  20  per  cent  that  of  light.  At 
the  same  time  Drs.  J.  Stebbins  and  A.  E. 
Whitford  of  the  Washburn  Observatory 
and  Pettit  used  the  60-inch  and  100-inch 
instruments  to  measure  photoelectrically 
the  magnitudes  of  the  galaxies  whose 
velocities  have  been  determined.  In  1956 
Humason,  Sandage,  and  Dr.  N.  Mayall 
of  the  Lick  Observatory  published  a  list 
of  the  velocities  of  600  galaxies  measured 
at  the  Mount  Wilson  and  Palomar  Ob- 
servatories and  of  300  galaxies  observed  at 
the  Lick  Observatory.  These  authors  then 
made  a  detailed  analysis  of  the  velocity- 
distance  relationship  based  on  the  observed 
velocities  and  the  distances  as  deduced 
from  the  magnitudes. 

By  means  of  eight-color  photometry 
Baum  has  determined  the  shape  of  the 
curve  of  radiated  energy  versus  wave- 
length for  a  number  of  galaxies  at  a  wide 
range  of  distances  and  has  measured  the 
radial  velocities  from  the  shifts  in  these 
curves.  Redshifts  as  great  as  AA/A  =  0.4, 
well  beyond  the  range  of  definite  spectro- 
scopic observations,  have  been  found.  In- 
deed it  is  now  becoming  evident  that  the 
primary  factor  that  sets  the  limit  on  the 
distance  to  which  a  large  telescope  like  the 
Hale  200-inch  can  observe  a  galaxy  or 
cluster  of  galaxies  is  not  the  falling  off  of 
the  light  because  of  increasing  distance  but 
the  redshift  of  light  from  the  galaxy  into 
the  far  infrared  where  photographic  plates 

are  no  longer  sensitive  and  where  atmos- 
pheric absorption  is  large  and  atmospheric 
"night-sky"  radiation  is  strong. 

Various  properties  of  galaxies  have  been 
investigated.  Thus  Munch  has  made  spec- 
troscopic observations  of  M  81  to  deter- 
mine its  rotation  and  mass.  Oort  and 
Minkowski  have  investigated  the  rotation, 
the  mass  distribution,  and  the  range  of 
random  stellar  velocities  in  NGC  3115. 
Multicolor  measurements  of  the  integrated 
light  of  a  large  number  of  globular  clus- 
ters and  galaxies  have  been  made  by  Baum 
and  Tifft  for  study  of  their  stellar  con- 
tents. Code  has  made  similar  studies  of 
the  energy  distribution  in  galaxies  with  a 
spectroscopic  scanner  for  the  same  purpose. 
Studies  of  the  luminosity  function  and  of 
the  distribution  and  clustering  of  galaxies 
were  carried  out  by  Zwicky.  Zwicky  also 
investigated  the  faint  blue  stars  and  other 
objects  in  the  halo  about  our  own  and 
other  galaxies. 

After  the  discovery  of  strong  radio 
sources  in  space,  Baade  and  Minkowski 
were  able  to  identify  several  of  the  strong- 
est of  the  sources  with  peculiar  optically 
observed  objects.  Several  sources,  Cygnus 
A,  NGC  1275,  and  probably  NGC  5128, 
turn  out  to  be  pairs  of  galaxies  in  collision. 
It  is  thought  that  the  strong  radio  waves 
emitted  have  their  origin  in  the  collision 
of  the  gas  clouds  present  in  the  spiral  arms 
of  these  pairs  of  galaxies.  The  remains  of 
two  explosions  of  supernovae,  the  Crab 
Nebula  (Supernova  of  A.D.  1054)  and  a 
faint  nebula  in  Cassiopeia,  first  observed 
by  Minkowski  and  identified  by  him  as 
the  remains  of  the  Supernova  of  1572,  were 
shown  to  be  identical  with  radio  sources. 
The  discovery  in  the  USSR  of  the  polar- 
ization of  the  light  from  the  Crab  Nebula 
followed  by  detailed  studies  of  the  polar- 
ization by  Baade  confirms  the  suggestion 
that  much  of  the  observed  light  from  this 
object  is  synchrotron  radiation  from  elec- 
trons of  very  high  velocity  accelerated  in 
a  small  magnetic  field.  Similarly,  the  ob- 
servations by  Baade  of  the  polarization  of 


the  light  from  the  jet  of  the  Galaxy  M  87 
which  is  also  a  radio  source  points  to 
synchrotron  radiation  as  the  origin  of 
much  of  the  light  from  this  object. 

The  discovery  by  H.  W.  Babcock  in 
1946  of  a  star  having  a  general  magnetic 
field  with  a  strength  of  several  thousand 
gauss  has  been  followed  by  a  survey  of 
many  stars  of  similar  types.  Of  305  stars 
examined,  84  show  a  definite  magnetic 
field  and  an  additional  55  probably  have 
such  a  field.  Since  in  many  stars  the  mag- 
netic fields  fluctuate  more  or  less  regularly 
with  a  period  of  a  few  days,  it  has  been 
necessary  to  make  extensive  observations 
of  many  of  these  objects  over  a  long  period 
of  time.  The  spectroscopic  and  other  prop- 
erties of  these  magnetic  stars  have  been 
studied  by  Deutsch  in  an  effort  to  build 
up  a  model  to  explain  their  behavior. 

With  the  development  of  the  solar  mag- 
netograph  by  H.  D.  and  H.  W.  Babcock 
it  has  been  possible  to  map  the  magnetic 
fields  over  the  surface  of  the  sun  with  an 
accuracy  of  a  few  tenths  of  a  gauss.  Daily 
maps  over  a  period  of  a  few  years  provide 
definite  evidence  for  persistent  polar  fields 
and  for  fluctuating  fields  over  the  equa- 
torial regions  of  the  sun.  The  observed 
fluctuations  explain  why  early  attempts  by 
Hale  and  others  to  measure  a  general  field 
of  the  sun  led  to  such  inconclusive  results. 
Sunspots,  prominences,  flares,  and  other 
solar  phenomena  have  been  investigated 
by  Nicholson  and  Richardson  on  photo- 
graphs and  spectroheliograms  taken  daily 
at  the  solar  towers  on  Mount  Wilson. 

Extensive  studies  have  been  made  of  the 
gas  clouds  in  our  own  Galaxy.  Minkowski 
has  carried  out  surveys  that  have  nearly 
doubled  the  number  of  known  planetary 
nebulae.  Olin  Wilson  has  developed  spec- 
troscopic techniques  that  have  enabled  him 
to  study  the  internal  motions  of  the  plane- 
tary nebulae  and,  with  Munch,  to  map  the 
motions  of  the  gases  over  a  large  part  of 
the  brighter  areas  of  the  Orion  nebula. 
Osterbrock  has  used  a  new  technique  to 
measure  densities  of  the  gas  in  various 

During  the  decade  Ralph  Wilson  com- 
pleted the  radial-velocity  program  of  the 
Observatories.  This  program  included  the 
measurement  of  several  thousand  plates, 
many  of  which  were  accumulated  during 
World  War  II,  and  the  publication  of  the 
radial  velocities  of  over  2400  stars.  A  com- 
pilation listing  the  positions,  magnitudes, 
spectral  types,  and  definitive  radial  veloci- 
ties of  all  the  15,105  stars  whose  velocities 
have  been  determined  at  any  observatory 
was  then  published  by  Wilson  with  the 
title  General  Catalogue  of  Stellar  Radial 

Olin  Wilson  with  the  assistance  of 
Bappu  made  a  detailed  study  of  the  widths 
of  the  emission  components  at  the  center 
of  the  H  and  K  lines  of  stars  later  than 
GO.  Wilson  found  a  linear  relationship 
between  the  logarithm  of  this  width  and 
the  absolute  magnitude  of  the  star  that 
should  prove  a  powerful  tool  for  the  de- 
termination of  the  luminosities,  and  there- 
fore of  the  distances,  of  stars  of  these  types. 

A  large  number  of  observations  were 
made  by  Merrill,  Joy,  and  Sanford  of  the 
spectra  of  variable  stars  and  of  stars  with 
prominent  emission  lines. 

In  a  project  sponsored  by  the  National 
Geographic  Society  the  48-inch  schmidt 
camera  was  used  during  most  of  the 
decade  to  map  the  whole  sky  north  of  dec- 
lination —  27°.  Two  photographs  were 
taken,  one  in  blue  light  and  the  other  in 
red  light,  of  each  of  879  fields.  Stars  were 
recorded  to  the  21st  magnitude  on  the  blue 
plates  and  to  the  20th  magnitude  on  the 
red  plates.  This  mapping  represents  a 
penetration  to  a  distance  about  three  times 
farther  than  previous  surveys,  or  a  cover- 
age of  about  twenty-five  times  the  volume 
of  space.  Over  120  copies  of  this  National 
Geographic  Society-Palomar  Observatory 
Sky  Survey  in  the  form  of  1758  photo- 
graphic prints  have  been  ordered  and  are 
being  distributed  to  observatories  through- 
out the  world.  From  a  search  of  the  survey 
plates  Abell  compiled  a  list  of  2712  clusters 
of  galaxies,  of  which  only  a  few  dozen 
were  known  before. 


Just  before  and  during  the  observations 
for  the  Survey  two  very  unusual  asteroids, 
named  Icarus  and  Geographos,  that  may 
come  unusually  close  to  the  earth,  were 
discovered  with  the  48-inch  schmidt.  In 
1951  Nicholson  discovered  the  twelfth 
satellite  of  Jupiter  on  plates  taken  with  the 
100-inch  telescope. 

The  above  describes  the  major  programs 
that  have  been  carried  out  by  the  perma- 
nent staff  of  the  Observatories,  the  Car- 
negie Fellows  and  research  fellows  who 
spent  one  or  two  years  each  at  the  Ob- 
servatories, and  a  number  of  graduate  stu- 
dents from  the  Department  of  Astronomy 
at  the  California  Institute.  Because  of  the 
limited  size  of  the  staff,  the  programs  have 
not  required  all  the  available  time  of  the 
telescopes.  Shortly  before  the  start  of  joint 
operation  of  the  two  Observatories  a  guest- 
investigator  program  was  inaugurated  by 
which  such  facilities  of  the  Observatories 
as  were  not  needed  by  its  own  staff  were 
made  available  to  astronomers  of  other 
institutions.  This  program  has  been  car- 
ried out  on  a  cooperative  basis,  the  visiting 
astronomers'  own  institution  providing  the 

necessary  leave  and  travel  expenses  and 
the  Observatories  furnishing  the  telescope 
time  and  the  photographic  plates  and 
other  supplies. 

During  the  past  decade  60  astronomers 
from  23  institutions  in  the  United  States 
made  over  160  visits  to  the  Observatories 
to  carry  out  observations  with  the  tele- 
scopes or  in  a  very  few  cases  to  study  plates 
already  available  in  the  files.  In  this  same 
period  20  astronomers  from  18  institutions 
in  12  other  countries  made  26  visits. 

The  optical  tests  of  the  Hale  telescope 
reported  early  in  the  decade  showed  that 
technically  the  instrument  performs  fully 
as  well  as  had  been  planned.  The  real  test 
of  the  soundness  of  design  of  an  instru- 
ment, its  optical  perfection,  and  the  effi- 
ciency of  its  auxiliary  equipment  comes 
from  its  ability  to  solve  astronomical  prob- 
lems. The  foregoing  record  of  the  first 
decade  of  operation  of  the  200-inch  tele- 
scope and  its  supporting  instruments  on 
Palomar  Mountain  and  Mount  Wilson 
provides  the  final  answer  as  to  the  success 
of  Hale's  dreams  and  hopes  for  the  tele- 
scope that  now  bears  his  name. 

For  the  second  time  in  the  past  eleven 
years  precipitation  was  above  normal  on 
Mount  Wilson,  with  a  rainfall  of  56.99 
inches.  Because  of  the  large  number  of 
cloudy    days    observing    conditions    were 


poor.  Solar  observations  were  made  on 
327  days,  and  observations  were  made  on 
292  nights  with  the  100-inch  telescope 
and  on  272  nights  with  the  60-inch  tele- 


Solar  Photography 

Solar  observations  were  made  by  Cragg, 
Hickox,  Nicholson,  Richardson,  and  Sey- 
fert.  The  numbers  of  photographs  of 
various  kinds  taken  between  July  1,  1957, 
and  June  30,  1958,  were  as  follows: 

Direct  photographs 648 

Ha  spectroheliograms,  60-foot  focus .  .  543 

Ha  spectroheliograms,  18-foot  focus.  .  1,200 

K2  spectroheliograms,  18-foot  focus.  .  909 

K2  spectroheliograms,  7-foot  focus .  .  .  55,600 

K  prominences,  18-foot  focus 1,008 

Sanspot  Activity 

The  magnetic  classification  and  study 
of  sunspots  and  related  phenomena  have 
been  continued  by  Nicholson  and  Cragg. 
Cooperative  programs  have  been  carried 
out  with  the  U.  S.  Naval  Observatory,  the 
University  of  Michigan,  the  Observatory 
of  Kodaikanal,  the  Meudon  Observatory, 
the  Central  Radio  Propagation  Laboratory, 
and  the  Naval  Research  Laboratory.  Dur- 
ing the  calendar  year  1957,  solar  observa- 
tions were  made  on  310  days,  on  none  of 


which  was  the  sun  without  spots.  The 
total  number  of  spot  groups  observed  in 
1957  was  855,  compared  with  642  in  1956 
and  208  in  1955.  Previous  to  1957  the 
largest  number  of  groups  observed  here  in 
one  year  was  663  in  1947.  The  largest 
number  of  groups  ever  observed  here  on 
one  day  was  27  on  December  31,  1957. 
Previous  to  this  cycle  the  highest  average 
number  of  sunspot  groups  per  month  was 
16.8  in  May  1947.  This  number  was  ex- 
ceeded in  four  consecutive  months  of  this 
cycle,  October,  November,  and  December 
1957,  and  January  1958.  The  current 
maximum  is  unquestionably  the  highest  in 
recorded  sunspot  history. 

The  number  of  sunspot  groups  in  high 
latitudes  has  been  exceptionally  large  in 
recent  years.  In  the  65  years  from  1878  to 
1943  only  four  groups  were  seen  on  more 
than  one  day  farther  than  40  degrees  from 
the  equator.  In  the  10  years  from  1943  to 
1953,  eight  such  groups  were  observed, 
and  in  the  present  cycle  twenty-seven  have 
already  been  observed,  four  in  1955,  seven 
in  1956,  eleven  in  1957,  and  live  in  the  first 
half  of  1958. 

A  large,  active  prominence  that  erupted 
on  June  15,  1958,  was  photographed  on 
June  12,  13,  14,  and  15.  On  June  15  during 
the  eruption  72  exposures  were  made. 
The  prominence  reached  a  height  of  1.7 
solar  radii  (735,000  miles)  and  a  terminal 
(maximum)  velocity  of  175  miles  per 

The  monthly  means  of  the  number  of 
groups  observed  daily  for  the  past  two  and 
one-half  years  are  shown  in  table  1. 

Magnetic  Polarities 

Magnetic  polarities  in  each  spot  group 
have,  if  possible,  been  measured  at  least 
once.  The  classification  of  groups  observed 
between  July  1,  1957,  and  June  30,  1958,  is 
indicated  in  table  2.  "Regular"  groups  in 
the  northern  hemisphere  are  those  in 
which  the  preceding  members  have  N 
(north-seeking)  polarity;  in  the  southern 
hemisphere  the  polarities  are  reversed. 

Solar  Magnetic  Fields 

Observations  with  the  solar  magneto- 
graph  have  been  continued  by  Harold  D. 
Babcock  at  the  Hale  Solar  Laboratory  on 
nearly  all  clear  days.  Much  attention  has 
been  paid  to  the  weak,  high-latitude  mag- 
netic fields  of  the  sun,  and  the  sequence 
of    observations    accumulated    since    1952 


Daily  Number  of  Sunspot  Groups 





January    6.4 

February   9.3 

March    10.6 

April    10.8 

May    9.9 

June   9.3 

July 9.3 

August 13.5 

September  15.2 

October   12.9 

November  11.8 

December    13.8 



















Yearly  mean 




Hemisphere    Regular  Irregular  Unclassified 

North   323  9  144 

South 267  7  149 

Whole  sun..  590  16  293 

shows  that  systematic  changes,  undoubt- 
edly related  to  the  progress  of  the  main 
solar  cycle,  have  taken  place. 

Howard  employed  the  solar  magneto- 
graph  which  is  installed  at  the  150-foot 
tower  to  investigate  magnetic  fields  and 
Doppler  shifts  on  the  surface  of  the  sun 
with  good  definition.  An  entrance  aper- 
ture 10  seconds  of  arc  square  is  normally 
used.  From  some  tracings  made  in  Oc- 
tober it  was  discovered  that  the  photo- 
spheric  magnetic  field  in  active  regions, 
which  is  of  the  order  of  30  gauss,  takes 
very  nearly  the  form  of  the  calcium  plage 


regions.  Observations  were  made  later  to 
determine  whether  deviations  of  the  mag- 
netic field  from  the  exact  shape  of  the 
plage  regions,  which  sometimes  occur,  may 
be  due  to  the  fact  that  the  magnetograph 
measures  only  the  component  of  the  mag- 
netic field  in  the  line  of  sight  and  that  all 
the  lines  of  force  may  not  be  parallel.  A 
number  of  regions  near  sunspots  have  been 
traced,  and  it  is  hoped  that  the  regions 
where  the  lines  of  force  of  the  sunspot 
fields  re-enter  the  photosphere  can  be 

A  number  of  traces  of  the  magnetic-field 
strength  along  the  sun's  equator  made  in 
October  showed  that  magnetic  features  of 
the  order  of  5  gauss  exhibit  some  changes 
over  a  period  of  a  few  hours.  Whether 
these  changes  represent  real  growth  and 
decay  of  the  field,  or  whether  the  fields 
moved  slowly  out  of  the  path  of  the  traces, 
could  not  be  determined.  More  observa- 
tions designed  to  differentiate  between  the 
two  possibilities  are  planned. 

A  method  was  devised  to  measure  mag- 
netic fields  and  motions  in  sunspots.  An 
aperture  5  seconds  square  is  used.  A  num- 
ber of  measures  were  made  of  a  spot  as  it 
crossed  the  solar  disk.  The  object  of  this 
investigation  is  to  determine  whether  or 
not  the  outward  motion  (Evershed  effect) 
in  sunspot  penumbrae  is  parallel  to  the 
lines  of  force  of  the  magnetic  field.  A  de- 
vice designed  by  Dr.  Robert  Leighton  is 
being  used  in  conjunction  with  the  mag- 
netograph to  obtain  an  accurate  determina- 
tion of  the  differential  rotation  of  the  sun. 
The  accuracy  obtained  for  relative  Doppler 
shifts  in  one  day's  observations  is  roughly 
0.03  km/sec.  Observations  can  be  made 
clear  to  the  poles. 

Corpuscular  Radiation 

Wildey  found,  from  a  theoretical  study 
of  the  transmission  of  solar  corpuscular 
radiation  (100  km/sec)  in  the  interplane- 
tary dust  cloud,  that  the  attenuation  was 
virtually  zero. 


Rotation  of  Venus 

Observations  of  Venus  for  rotation  were 
made  by  Richardson  in  1956-1957  at  the 
Snow  telescope  with  a  coelostat  and  con- 
cave-mirror system,  which  produced  an 
image  of  the  planet  on  the  slit  from  1.6  to 
3.9  mm  in  diameter,  during  eastern  elonga- 
tion. A  grating  with  900  lines  per  mm  was 
used  in  a  constant-temperature  pit  with  a 
Littrow  spectrograph  of  18  feet  focal 
length.  This  arrangement  used  in  the  2nd 
order  gave  a  dispersion  of  0.84  A/mm. 
With  a  slit  width  of  0.1  mm  a  satisfactory 
exposure  could  be  obtained  in  45  minutes, 
although  more  than  an  hour  was  desirable 
if  time  permitted.  The  emulsion  was  East- 
man 103a-E(2)  used  with  Corning  filter 
3486,  which  gives  80  per  cent  transmission 
longward  of  5570  A. 

Observations  when  Venus  was  west  of 
the  sun  were  made  at  the  coude  focus  of 
the  100-inch  telescope  with  the  114-inch 
camera.   Measures  on  102  lines  from  8  of 

the  best  plates  were  selected  for  discussion. 
The  straight  mean  of  the  102  measures 
gave  a  limb  velocity  of  —0.032  km/sec, 
corresponding  to  a  rotation  period  of  14 
days,  retrograde.  The  standard  error  of 
the  mean  is  ±0.033  km/sec. 

Checks  on  the  measures  at  the  Snow 
telescope  were  obtained  on  the  sun  by  us- 
ing a  short-focus  lens  that  gave  a  solar 
image  about  the  size  of  Venus.  Observa- 
tions for  solar  rotation  were  made  from 
heliocentric  latitude  3°  to  71°.  They 
agreed  closely  with  the  solar  rotation  in 
these  latitudes  obtained  by  Adams.  Checks 
at  the  100-inch  were  obtained  from  spectra 
of  Mars  which  gave  a  rotation  period  of 
24.2  hours,  as  compared  with  the  known 
rotation  of  24.6  hours. 

That  observations  were  not  taken 
pole-on  was  checked  from  the  inclination 
of  the  axis  of  Venus  as  determined  by 
G.  P.  Kuiper  and  Richardson,  on  the 
assumption  that  the  ultraviolet  cloud  belts 
are  parallel  to  the  planet's  equator. 




A  series  of  direct  photographs  of  the 
bright  comet  Mrkos  (1957d)  was  obtained 
with  the  48-inch  schmidt  telescope  by 
Kearns  under  the  supervision  of  Oster- 
brock.  The  plates  were  taken  for  a  physi- 
cal study  of  the  material  in  the  tail,  and 
they  have  been  sent  to  Professor  Biermann 
of  the  Max  Planck  Institut,  who  will 
supervise  their  measurement  and  reduc- 
tion. A  preliminary  survey  of  the  plates 
by  Kearns  resulted  in  the  discovery  of  a 
feature  which  not  only  moved  from  the 
head  of  the  comet  away  from  the  sun 
along  the  tail,  but  also  appeared  to  move 
from  one  side  of  the  tail  to  the  other.  This 
observation  shows  that  the  comet  tail 
either  rotated  or  oscillated,  with  a  period 
of  the  order  of  four  days. 

Greenstein  obtained  spectra  of  Comet 
Mrkos  (1957d)  with  the  highest  resolution 
ever  employed  for  cometary  spectra, 
namely,  18  A/mm  in  the  blue  and  27 
A/mm  in  the  red.  The  analysis  of  the 
A3883  CN  band  permitted  a  very  detailed 
confirmation  of  the  Swings  fluorescence 
mechanism.  In  addition,  small  velocity 
differences  along  the  radius  vector  of  the 
sun  shifted  individual  rotational  lines  suffi- 
ciently, with  respect  to  the  solar  absorption- 

line  spectrum,  to  produce  intensity  changes 
between  the  side  of  the  comet  head  facing 
the  sun  and  the  side  away  from  it.  The 
structure  of  the  Na  I  lines  showed  a  simi- 
lar fore-and-aft  asymmetry,  being  strong 
but  sharply  cut  off  on  the  side  facing  the 
sun,  and  streaming  away  more  uniformly 
toward  the  tail.  An  atlas  of  the  visual 
spectrum  (AA4900-6800)  has  been  pre- 
pared, and  about  400  lines,  largely  C2  and 
NH2,  have  been  measured.  The  red  CN 
system  does  not  appear.  The  identification 
of  [O  I],  made  in  collaboration  with  Dr. 
P.  Swings  of  the  University  of  Liege, 
seems  definite.  Detailed  analysis  of  these 
spectra  will  be  carried  out  at  Liege. 

Periodic  Comet  Wolf  I  (1951  VI)  was 
recovered  on  plates  exposed  by  Baum  on 
the  night  of  June  12  at  the  prime  focus  of 
the  200-inch.  This  work  was  done  in  co- 
operation with  Dr.  Elizabeth  Roemer  of 
the  U.  S.  Naval  Observatory  at  Flagstaff, 
who  supplied  the  information  required 
for  undertaking  recovery  and  who  meas- 
ured the  plates  to  obtain  corrected  posi- 
tions. The  visual  magnitude  of  the  comet 
as  determined  indirectly  was  20.4  on  the 
night  of  observation. 


During  the  report  year  828  spectrograms 
were  made  with  the  200-inch  telescope, 
965  with  the  100-inch,  and  932  with  the 

Chemical    Composition    of  Stellar 

An  extensive  project  on  the  chemical 
composition  of  stellar  atmospheres  is  in 
progress  under  the  supervision  of  Green- 
stein with  the  financial  support  of  the 
Office  of  Scientific  Research  of  the  U.  S. 
Air  Force.  One  major  investigation  has 
been  the  determination  of  abundance  dif- 
ferences between  very  old  and  very  young 
stars.  Spectra  of  the  brightest  red  giants 
in  M  92  and  M  13  taken  by  Greenstein 

have  been  compared  with  those  of  the 
field  high-velocity  star  HDE  232078  and 
a  young  K3  II  star  in  the  Galactic  cluster 
M  41.  Heifer  and  Wallerstein  find  abun- 
dance deficiencies  of  the  metals  largest  in 
M  92  (factor  of  200),  next  in  M  13  and 
HDE  232078  (factor  of  20),  and  small  or 
absent  in  the  M  41  star.  All  metals  seem 
to  share  about  equally  in  this  deficiency. 

Two  dwarf  stars  in  the  Hyades,  a  rela- 
tively young  Galactic  cluster,  have  been 
compared  with  the  sun,  and  with  the  old, 
high-velocity  star  85  Pegasi.  This  investi- 
gation is  being  carried  out  using  relatively 
unblended  lines  in  the  yellow-red,  with 
many  Mount  Wilson  coude  plates  per  star, 
to  see  whether  any  small  abundance  dif- 



ferences  of  the  metals  can  be  detected, 
caused  by  a  difference  of  at  least  3.5  billion 
years  in  epoch  of  formation  of  the  stars. 

The  peculiar  red  supergiant  VY  Canis 
Majoris  has  been  observed  by  Wallerstein. 
The  outer  envelope  of  this  star  contains 
emission  lines  of  sodium,  calcium,  and  po- 
tassium (the  last  observed  for  the  first 
time  in  any  star) . 

The  reddish  white  dwarf  van  Maanen  2 
has  been  investigated  very  extensively  by 
Weidemann.  Models  have  been  computed 
for  values  of  the  effective  temperature 
6600°,  5700°,  and  5000°  K,  with  log  g= 
+  75,  8.0,  8.4.  The  profiles  of  the  blended 
and  broadened  ultraviolet  Fe  I  lines 
have  been  resolved  into  their  compo- 
nents, taking  into  account  the  depend- 
ence on  depth  and  the  pressure  broadening 
of  the  van  der  Waals  type.  The  gas  pres- 
sure is  found  to  be  about  1000  atmospheres, 
much  higher  than  expected  with  normal 
abundances.  Allowing  the  abundance  of 
hydrogen  ch  and  of  the  metals  cm  to  vary, 
the  observed  damping  constant  and  the 
central  depths  of  the  lines  yield  a  decrease 
of  log  £h=—  1.6,  and  of  log  cm^—  3.8, 
with  respect  to  the  sun.  The  enormous  de- 
crease in  metal  abundances  is  hardly  af- 
fected by  allowing  for  the  formation  of 
molecules,  although  the  hydrogen  defi- 
ciency is  reduced.  The  theoretical  predic- 
tion of  exhaustion  of  the  hydrogen  and 
low  surface  abundances  of  the  metals  is 
apparently  confirmed.  Much  remains  to 
be  done  in  this  field  of  phenomena  at  very 
high  pressure. 

Bonsack  has  made  an  extensive  study 
of  the  neutral  lithium  lines  in  47  G8  and 
Ml  stars.  About  half  the  stars  show  Li  I 
lines  at  7  A/mm.  Curves  of  growth  have 
been  constructed,  and  the  Li  abundance 
has  been  determined.  The  search  for  an- 
other element  important  in  theories  of 
nucleogenesis,  beryllium,  has  been  aban- 
doned because  of  the  technical  difficulties 
of  observation  at  A3130. 

An  elaborate  spectrophotometric  study 
of  the  standard  star,  o  Bootis,  F2  V,  is  be- 
ing carried  out  by  Greenstein.  It  is  hoped 

that  this  star,  observed  at  Mount  Wilson, 
Palomar,  Victoria,  and  elsewhere,  can  be 
used  to  provide  standards  of  equivalent 
width  for  other  investigators. 

Carbon  Stars 

A  survey  by  Greenstein  of  high-disper- 
sion spectra  of  the  early  carbon  stars  has 
revealed  a  number  of  differences  in  both 
the  line  and  the  band  spectra.  The  high- 
velocity  star  HD  201626  seems  to  be  of 
early  type,  related  to  the  "barium"  stars, 
and  has  very  strong  C2,  CH,  CN  bands. 
Excellent  spectra  have  been  obtained  of 
the  peculiar  carbon  emission  object,  in- 
volved in  a  nebula,  V348  Sagittarii.  Some 
lines  are  complex  in  structure,  double  or 
of  the  Be  or  P  Cygni  type.  The  barium 
star  £  Capricorni  has  been  found  to  show 
weak  C2  and  C  I  lines. 

Blue  Stars  in  the  Galactic  Halo 

Spectra  of  faint  blue  stars  in  the  Galac- 
tic halo  have  been  classified  by  Greenstein 
in  an  attempt  to  systematize  the  bewilder- 
ing variety  of  objects  found  there.  In 
addition  to  runaway  Population  I  objects, 
many  stars  of  Population  II  are  found, 
including:  (1)  weak-line  B  and  A  stars, 
in  which  hydrogen  lines  are  strong,  deep, 
and  sharp,  but  He  I  and  Mg  II  are  weak; 
these  may  be  horizontal-branch  stars;  (2) 
subdwarfs,  either  of  type  O  or,  less  easily 
recognizable,  of  type  B,  with  relatively 
shallow  lines.  The  search  for  new  types  of 
white  dwarfs  continues.  The  star  LDS 
749B  is  another  member  of  the  helium- 
rich  type,  DB.  The  object  HZ  34  has  an 
unusual  and  as  yet  unclassifiable  spectrum. 

Central  Star  of  the  Crab  Nebula 

Excellent  spectra  of  the  supposed  central 
star  of  the  Crab  Nebula  have  been  ob- 
tained by  Zwicky  with  the  200-inch  tele- 
scope prime-focus  spectrograph.  They  are 
entirely  continuous,  no  absorption  or  emis- 
sion features  being  recognizable  in  them. 
Further  attempts  will,  however,  be  made 
to   record   good   spectra   over   the   whole 



available  wavelength  range,  particularly 
in  the  ultraviolet,  and  to  eliminate  as  much 
of  the  background  light  from  the  super- 
posed nebulosity  as  possible  by  decreasing 
the  width  of  the  spectrograph  slit. 

Line  Blanketing 

Melbourne  has  nearlv  completed  ob- 
servations for  a  study  of  line  blanketing  in 
various  types  of  stars.  He  used  the  photo- 
electric scanner  to  obtain  monochromatic 
fluxes,  and  the  coude  spectrograph  to  de- 
termine the  effects  of  the  lines.  The  cor- 
rected fluxes  are  compared  with  predicted 
values  from  model  stellar  atmospheres.  A 
total  of  15  stars  from  09  V  to  G4  V, 
F5  III  to  G8  III,  and  2  subdwarfs  are  in- 
cluded in  the  program. 

Variable  Stars 

In  the  hope  of  obtaining  concurrent  rec- 
ords of  one  of  its  frequent  but  unpre- 
dictable outbursts,  continuous  photoelectric 
observations  of  the  light-changes  of  UV 
Ceti  were  made  on  four  nights,  October 
4-7,  1957,  by  Paul  Roques  at  the  Griffith 
Observatory  in  Los  Angeles  with  the 
12-inch  refractor,  and  on  the  same  nights 
successive  spectrograms  were  obtained 
with  the  60-inch  reflector  on  Mount  Wil- 
son by  Joy. 

No  variations  in  brightness  greater  than 
0.2  magnitude  were  recorded  while  the  sky 
in  Los  Angeles  permitted  observations. 
On  Mount  Wilson,  with  excellent  seeing 
and  clear  sky,  12  spectrograms  of  the 
fainter  star  of  the  pair  (magnitudes  12.5, 
13.00;  separation  <  2";  types  dM5.5e, 
dM5.5e)  were  obtained  with  average  ex- 
posures of  100  minutes.  On  October  7 
during  the  third  exposure  the  observer 
noted  a  sudden  flare  of  more  than  a  mag- 
nitude, and  the  star  returned  to  normal 
brightness  more  slowly  in  about  10  min- 
utes. Unfortunately,  at  this  time  the  Los 
Angeles  sky  was  covered  with  haze  and 
fog.  This  spectrogram  showed  much 
wider  and  stronger  emission  hydrogen 
lines  than  the  other  plates  of  the  series,  and 

the  bright  lines  H  and  K  of  Ca  II  were 
somewhat  strengthened.  This  observation 
confirms  the  results  previously  obtained 
for  the  same  star  at  the  time  of  the  much 
greater  flare  of  September  25,  1948. 

On  the  same  nights,  three  spectrograms 
each  of  V  371  Ononis  and  20C1191  (types 
dM3e,  dM3e)  were  obtained  by  Joy.  Ve- 
locity variations  were  confirmed,  but  no 
certain  changes  in  brightness  or  spectrum 
were  noted. 

Deutsch  has  the  following  variables  un- 
der spectroscopic  observation  at  the  coude : 
VV  Cephei,  which  is  now  in  the  egress 
phase  of  its  chromospheric  eclipse;  RU 
Camelopardalis,  a  cepheid  with  carbon 
bands;  p  Cassiopeiae,  which  has  many 
double  lines;  and  R  Cygni.  An  analysis 
by  Merrill  of  the  spectrum  of  this  last  star 
near  maximum  light  has  shown  the  pres- 
ence of  a  great  many  narrow  bright  lines 
of  metallic  elements. 

A  brief  summary  of  the  work  of  the  late 
R.  F.  Sanford  on  the  complex  and  variable 
spectrum  of  AB  Aurigae,  type  Aep,  has 
been  prepared  by  Merrill. 

Tiflft  has  completed  his  photoelectric  and 
spectroscopic  investigation  of  a  c-cluster- 
type  variable,  T  Sextantis,  and  is  reducing 
data  on  several  additional  cluster-type  vari- 
ables of  classes  a  and  c. 

Wallerstein  observed  U,  B,  V  colors  of 
long-period  variables  with  broad  flat 
minima:  U  Cancri,  Z  Cassiopeiae,  UX 
Cygni,  Z  Puppis,  Z  Tauri,  RU  Aurigae. 
Merrill  has  suggested  that  such  flattened 
light-curves  might  be  caused  by  faint  com- 
panions. Wallerstein  concludes,  however, 
that  these  stars  cannot  have  faint  blue 
companions  like  Mira  B;  for  Mira  in  the 
1956  minimum  the  colors  were  clearly 
composite,  B  -  V  =  + 1™41,  U  -  B  = 
—  0™85,  while  none  of  the  above  list 
showed  any  such  ultraviolet  excess. 

Two  symbiotic  stars  have  been  found  by 
Greenstein  at  high  Galactic  latitudes, 
+  24°2742,  +37°2318.  Both  have  P  Cygni 
characteristics  and  a  variable  Hot.  The 
former  shows  strong  Ca  II  emission,  the 



latter  a  composite,  early-plus-late-type  spec- 
trum. The  G  star  HD  117555,  also  at  high 
latitude,  has  been  extensively  observed; 
discovered  by  Merrill  to  be  rapidly  rotat- 
ing, it  has  very  broad  Ha  emission  which 
varies  in  intensity  and  profile  from  night 
to  night. 

Wildey  has  made  a  preliminary  study  of 
the  bright  variable  in  NGC  7006,  and  its 
location  in  the  color-magnitude  diagram. 

Magnetic  Stars 

The  catalogue  of  magnetic  stars  re- 
ported last  year  has  been  followed  by  a 
discussion  by  H.  W.  Babcock  of  the  avail- 
able data  on  magnetic  phenomena  among 
the  stars  of  spectral  type  A.  These  stars 
have  been  classified  into  four  groups  ac- 
cording to  the  type  of  magnetic  variation: 
the  a  group,  snowing  cyclically  varying 
fields  of  large  amplitude  with  nearly  sym- 
metrical reversals  of  polarity;  the  3  group, 
with  irregular  variations  and  seemingly 
random  reversals  of  polarity;  and  the  y 
group,  showing  fluctuations  of  magnetic 
intensity  but  always  the  same  polarity;  in 
addition,  a  new  (S)  group  of  magnetic 
stars  that  are  also  irregular  spectrum  vari- 
ables has  been  identified.  Ten  of  the  70 
known  magnetic  A  stars  are  spectroscopic 
binaries,  of  which  6  are  newly  identified. 
Only  1  of  the  a  variables  is  a  close  binary, 
however,  and  in  no  others  is  the  magnetic 
variation  related  to  the  binary  motion. 
It  is  becoming  increasingly  clear  that  mag- 
netic forces  are  dominant  in  governing  the 
disposition  of  gaseous  material  in  the  vi- 
cinity of  a  star,  and  that  these  forces  must 
be  taken  into  account  in  the  evolution  of 
binary  systems  and  in  general  cosmogoni- 
cal  problems. 

The  diverse  phenomena  observed  in  the 
magnetic  stars  invite  further,  more  inten- 
sive investigation,  and  observations  have 
been  continued  by  Babcock  on  a  regular 
basis  with  the  200-inch.  Three  stars  that 
may  be  singled  out  as  of  particular  interest 
are:  (1)  53  Camelopardalis,  an  a  variable 
with  an  8-day  period  and  a  field  varying 

between  the  limits  +3700  and  —5100 
gauss,  the  strongest  yet  observed  in  any 
star;  (2)  HD  32633,  an  a  variable  with  a 
4-day  period  and  a  decidedly  irregular 
magnetic  amplitude;  and  (3)  HD  187474, 
unique  in  that  its  magnetic  field  has  been 
invariant,  and  moderately  strong,  for  over 
a  year  since  its  discovery. 

Bonsack  has  analyzed  the  velocity  varia- 
tions of  different  lines  in  56  Arietis.  He 
found  that  the  rigid,  oblique,  rotating 
model  would  satisfactorily  explain  the  ob- 
served phenomena. 

Among  the  peculiar  A  stars,  the  periodic 
magnetic  spectrum  variables  HD  98088 
and  53  Camelopardalis  are  under  observa- 
tion for  harmonic  analysis  by  Deutsch. 
Spectrograms  of  the  ultra-sharp  line  "man- 
ganese star"  Iota  Coronae  Borealis  are  be- 
ing assessed  for  line  identification  at  the 
Perkins  Observatory. 

Mass  Loss  from  Late-Type  Giants 

Coude  spectrograms  at  dispersions  of 
10  A/mm,  or  brighter,  are  being  accumu- 
lated by  Deutsch  to  obtain  evidence  bear- 
ing on  mass  loss  from  late-type  giants  and 
supergiants.  The  observations  at  hand 
show  that  in  different  red  giants  the  cir- 
cumstellar  components  of  strong  resonance 
lines  occur  with  widely  different  strengths. 
At  10  A/mm,  the  circumstellar  spectrum 
can  generally  be  recognized  as  such  in  stars 
that  populate  the  Hertzsprung-Russell  dia- 
gram above  a  line  running  from  Mv—  —3 
at  M0  to  Mv  =  0  at  M5.  In  these  objects, 
the  H  and  K  lines  of  Ca  II  show  deep, 
rectangular  absorption  cores  (H3,  K3) 
superposed  on  emission  features  (H2,  K2). 
The  absorption  cores  of  other  resonance 
lines  correlate  well  in  strength  with  the 
strength  of  H3  and  K3,  indicating  com- 
parable ionization  in  all  these  circumstellar 

Circumstellar  features  are  now  recog- 
nized also  in  some  K-type  supergiants,  and 
in  a  few  long-period  variables  with  types 
later  than  M5.  A  recent  10  A/mm  spectro- 
gram of  x  Cygni  at  maximum  brightness 


shows  conspicuous  doubling  of  all  strong 
resonance  lines  in  absorption.  Circum- 
stellar  features  probably  occur  as  well  in 
the  spectra  of  some  N  and  S  stars  with 
effective  temperatures  and  luminosities 
that  are  comparable  with  those  of  M  giants 
and  supergiants.  Despite  these  correlations 
with  type  and  luminosity,  an  appreciable 
dispersion  exists  in  the  circumstellar-line 
strength  at  any  point  in  the  Hertzsprung- 
Russell  diagram. 

Observations  at  4.5  A/mm  of  a  few  M 
giants  of  earlier  type  and  lower  luminosity 
show  a  composite  structure  at  K3.  This  is 
typically  a  shallow  absorption  reversal  in 
K2,  with  a  deep  but  very  weak  core  at  the 
violet  edge.  The  shallow  absorption  fea- 
ture occurs  alone  in  normal  giants  earlier 
than  MO.  In  spectroscopic  binaries,  it 
moves  with  the  lines  from  the  reversing 
layer.  It  evidently  arises  not  in  a  "de- 
tached" circumstellar  envelope,  but  rather 
in  the  stellar  chromosphere,  at  or  near  the 
levels  responsible  for  K2.  The  deep,  violet- 
displaced  core  represents  the  first  detectable 
manifestation  of  the  true  circumstellar 

A  by-product  of  the  coude  survey  of 
late-type  giants  and  supergiants  is  the  dis- 
covery that  many  show  a  weak,  wide  emis- 
sion line  of  Co  I  in  the  longward  wing  of 
Ki,  and  some  show  an  emission  line  of 
Fe  II  in  the  same  region.  These  lines  ap- 
pear with  much  greater  strength  at  certain 
phases  in  long-period  variables.  In  the 
(relatively)  nonvariable  M  stars  they  are 
probably  related  to  the  emission  lines  in 
the  A3200  region,  and  they  are  evidence 
for  hot  and  deep  chromospheres  in  these 

For  evidence  on  the  dimensions  of  the 
circumstellar  envelopes,  the  behavior  of 
H3  and  K3  is  under  study  in  several  spec- 
troscopic binaries  containing  late-type 
giants.  In  each  of  the  following  systems, 
Hs  and  K3  have  been  found  to  show  sta- 
tionary components  that  are  not  interstel- 
lar, or  to  yield  a  motion  different  from  that 
of  the  reversing  layer:   £  Cygni  (K5  lb), 

RR  Ursae  Majoris  (gM5),  v\  Geminorum 
(gM3),  22  Vulpeculae  (cG4),  A  Androm- 
edae  (G7  III),  and  VV  Cephei  (M2  + 

Visual  binaries  comprising  late-type 
supergiants  are  also  being  surveyed  by 
Deutsch  for  evidence  of  the  a  Herculis 
phenomenon,  where  the  circumstellar  en- 
velope of  the  primary  impresses  its  absorp- 
tion lines  on  the  spectrum  of  the  visual 
companion.  Antares  probably  shows  the 
effect;  the  B3  V  companion  exhibits  pairs 
of  sharp  absorption  lines  at  H  and  K.  The 
weaker  pair  is  probably  interstellar,  the 
stronger  circumstellar  and  due  to  the  M 
supergiant.  In  the  unique  nebula  found 
by  Struve  to  surround  the  B  star,  sharp 
emission  lines  of  Si  II  have  been  identified. 
These  lines  are  evidence  in  support  of 
O'Keefe's  theory  that  quartz  crystals  con- 
dense in  the  gaseous  ejecta  from  the  M 
star,  then  volatilize  again  in  the  vicinity 
of  the  B  star.  The  emission  lines  of 
[Fe  II]  reveal  evidence  of  systematic  mo- 
tions superposed  on  the  turbulence  noted 
by  Struve.  The  problem  of  the  excitation 
of  this  nebula  is  still  unsolved. 

Mira  is  another  visual  binary  that  prob- 
ably shows  the  a  Herculis  phenomenon  at 
H  and  K.  In  this  system,  the  hot  com- 
panion lies  near  Mv=+7,  in  the  same 
region  of  the  Hertzsprung-Russell  diagram 
as  the  explosive  variables  in  the  binary 
systems  SS  Cygni,  AE  Aquarii,  and  T 
Coronae  Borealis.  The  observed  photo- 
metric and  spectroscopic  instability  of 
Mira  B  suggests  that  this  star  may  be  inter- 
acting with  matter  ejected  from  the  late- 
type  star,  in  a  process  analogous  to  that 
suggested  by  Kraft  and  Crawford  for  the 
other  three  systems.  The  possibility  exists 
that  all  these  explosive  variables  are  really 
white  dwarfs,  with  their  excess  luminosity 
deriving  from  accretion  heating.  There 
are  obvious  theoretical  reasons  for  antici- 
pating violent  instabilities  in  a  white  dwarf 
subjected  to  accretion  of  hydrogen-rich 
material  ejected  from  an  aging  red-giant 
companion.  At  the  1957  minimum  of  the 
long-period  variable,  Mira  B  was  seen  at 



nearly  the  same  position  and  brightness  as 
at  the  preceding  several  minima. 

The  Relationship  of  H  and  K  Emission  to 
Absolute  Magnitude 

A  total  of  303  stars  has  been  observed 
by  Wilson  for  his  investigation  of  the  rela- 
tion between  the  width  of  the  H  and  K 
emission  lines  and  the  absolute  magnitude. 
The  great  majority  of  these  stars  show  H 
and  K  in  emission. 

Pending  the  possible  derivation  of  a  the- 
oretical expression  for  the  width-luminos- 
ity relationship  which  might  conceivably 
by-pass  the  need  for  empirical  calibration, 
one  of  the  immediate  needs  is  to  improve 
the  latter.  For  this  purpose  the  four 
Hyades  K-type  stars  have  been  extensively 
observed.  When  the  accidental  measuring 
errors  are  reduced  in  this  fashion,  it  is 
found  that  the  K-line  widths  place  these 
four  stars  in  the  proper  order  of  luminosity 
although  the  total  range  in  their  absolute 
magnitudes  is  only  about  0.3  magnitude. 
Also  the  mean  of  measures  of  the  Praesepe 
stars,  less  extensive  than  those  of  the 
Hyades,  gives  a  distance  modulus  for 
Praesepe  within  0.1  magnitude  of  that  de- 
rived by  the  photometric  observers.  Thus, 
so  far,  the  method  appears  to  be  capable 
of  excellent  accuracy.  It  is  being  used 
therefore  to  construct  a  color-magnitude 
diagram  for  late-type  stars  in  the  solar 

A  preliminary  plot  of  this  type,  based  on 
observation  of  stars  with  modern  photo- 
electric colors  which  have  been  published 
in  various  lists,  shows  the  following  fea- 
tures: (1)  The  lower  and  right-hand  edges 
of  the  distribution  appear  to  be  fairly 
sharply  defined.  (2)  Except  for  a  scatter- 
ing of  supergiants,  most  of  the  plotted 
points  fall  within  the  area  bounded  by  the 
curves  for  M  67  and  M  3.   (3)  The  M  67 

curve  lies  fairly  close  to  the  lower  boundary 
of  the  distribution  except  near  B  — V~1.0, 
where  a  small  number  of  stars  form  a 
nearly  vertical  group  well  below  the  M  67 

Confirmation  of  Wilson's  photographic 
measurements  of  the  half-widths  of  emis- 
sion H  and  K  in  two  stars  was  obtained  by 
Code  and  Wilson  using  the  photoelectric 
spectrum  scanner  described  under  instru- 

Radial  Velocities 

Over  140  radial-velocity  spectrograms 
of  a  number  of  faint  M  dwarfs  and  a  few 
standard  stars  were  obtained  by  Woolley 
at  a  dispersion  of  80  A/mm  with  the  60- 
inch  telescope.  They  were  taken  to  fill  cer- 
tain gaps  in  the  radial-velocity  data  neces- 
sary for  a  statistical  study  of  the  space 
velocity  of  stars  within  20  parsecs  of  the 

An  investigation  of  the  radial  velocities 
of  distant  stars  in  the  Galactic  region  be- 
tween longitude  340°  and  longitude  30° 
has  been  carried  out  by  Luis  and  Guido 
Munch  to  provide  data  on  the  rotation  of 
the  inner  parts  of  the  Galactic  system.  The 
preliminary  discussion  of  the  material  has 
shown  that  the  observed  motion  of  these 
stars  is  in  good  agreement  with  the  rota- 
tional curve  of  the  Galactic  system  pro- 
posed by  M.  Schmidt  from  21-cm  observa- 
tions. In  the  course  of  this  work  it  was 
found  that  the  B0  la  star  HD  173438  is  a 
single-line  spectroscopic  binary  with  a 
period  around  240  days  and  a  90  km/sec 
range.  The  mass  function  of  the  system 
is  then  around  15  O,  and,  for  a  mass  ratio 
2,  the  mass  of  the  primary  would  be  200  O . 
The  09  star  HD  173783  has  been  found 
to  be  also  a  spectroscopic  binary,  but  its 
period  and  range  have  not  been  deter- 
mined with  certainty. 


Internal  Motions 
Wilson  has  continued  his  investigation 
of  the  internal  motions  of  nebulae  using 

the  multislit  technique,  and  a  dispersion  of 
4.5  A/mm  at  the  200-inch  coude.  Most  of 
the  planetary  nebulae,  except  for  the  ir- 



regular  ones,  are  more  or  less  elliptical  or 
ovoid  in  outline.  One  may  ask  whether 
the  present  shape  of  such  objects  is  a  mani- 
festation of  nonspherical  velocity  distribu- 
tion at  the  time  of  ejection  and,  if  so, 
whether  such  a  velocity  distribution  can 
be  revealed  by  observation. 

Imagine  two  ellipsoidal  nebulae,  one  of 
which  is  viewed  at  right  angles  to  the 
major  axis  while  the  other  is  seen  at  an 
angle  of  45°;  suppose  that  both  are  ob- 
served with  a  multislit,  and  that  the  central 
slit  is  set  on  the  nucleus  of  each.  Then  it 
follows  that  for  the  first  nebula  the  double 
line  produced  by  the  central  slit  is  sym- 
metrical, as  are  also  the  double  lines  on 
either  side  of  it.  The  same  is  not  true, 
however,  for  the  nebula  viewed  at  45°. 
Here  the  central  double  line  is  again  sym- 
metrical but  the  lines  on  either  side  are 
asymmetrical  and  in  the  opposite  sense  on 
the  two  sides  of  the  central  one.  Moreover, 
the  asymmetry  is  considerably  enhanced 
by  the  assumption  that  the  velocity  in  the 
shell  at  any  point  is  proportional  to  the 
radius  vector  from  the  nucleus  to  the  point. 

Spectral  lines  have  been  computed  ac- 
cording to  the  foregoing  picture,  and  com- 
pared with  observation.  Since  the  angle 
between  the  major  axis  and  the  line  of 
sight  is  not  known  for  any  nebula,  the 
comparison  must  be  qualitative,  but  within 
this  limitation  the  agreement  between  ob- 
served and  calculated  lines  is  excellent. 
Thus  it  is  found  that  NGC  7009  is  oriented 
so  that  the  line  of  sight  is  nearly  perpen- 
dicular to  its  true  major  axis.  NGC  7662 
on  the  other  hand  is  viewed  at  a  consider- 
able angle  to  its  major  axis,  and  the  as- 
sumption of  an  ellipsoidal  velocity  dis- 
tribution gives  the  best  agreement  between 
calculated  and  observed  lines. 

The  observations  of  the  Orion  nebula 
with  the  multislit  and  the  measurements 
of  the  plates  have  been  completed  by 
Munch  and  Wilson.  This  study  provides 
data  on  the  motions  of  the  gases  over  the 
brighter  parts  of  the  nebula.  A  large  num- 
ber of  additional  observations  were  made 

of  the  shapes  of  the  emission  lines  in  the 
Orion  nebula  by  Code  and  Wilson  using 
the  photoelectric  spectrum  scanner  on  the 
100-inch  telescope,  special  attention  being 
given  to  Ni,  N2,  and  H(3.  The  wings  of 
these  lines  were  traced  out  to  the  0.1  per 
cent  level  in  order  to  determine  the  ve- 
locity dispersion  implied  by  the  line  wings. 
A  continuous  emission  feature  at  approxi- 
mately 5034  A  with  a  half-width  of  about 
10  A  was  found  in  the  Orion  nebula  which 
cannot  be  identified  with  any  blend  of 
atomic  lines.  In  this  region  faint  lines  were 
identified  at  A5041  and  A5057  with  the 
42P°-42D  multiplet  of  Si  II  and  at  A5048 
with  the  He  I  line  from  the  21P°-41S 
transition.  These  lines  are  exceedingly 
weak,  representing  an  intensity  of  only  a 
few  per  cent  of  He  I  5015.7,  which  in  turn 
is  only  a  few  per  cent  of  the  Ni  line  of 
[O  III].  These  relative  intensities  indicate 
one  of  the  promising  advantages  of  the 
scanning  technique.  In  addition,  well  re- 
solved measurements  of  the  components 
of  3727  [O  II]  were  obtained  in  faint  ex- 
tensions of  the  nebulosity,  and  the  intensity 
ratios  were  in  good  agreement  with  Oster- 
brock's  photographic  determinations.  H(3 
was  traced  out  beyond  the  obvious  nebu- 
losity down  to  emission  measures  of  the 
order  of  300. 

Radial  Velocities 

Minkowski  has  now  finished  his  part  of 
a  program  of  radial  velocities  of  planetary 
nebulae  that  was  undertaken  as  a  joint 
program  by  him  and  by  Dr.  N.  U.  Mayall 
at  the  Lick  Observatory.  Radial  velocities 
of  142  new  planetaries  have  been  obtained 
from  observations  with  the  100-inch  and 
the  200-inch  telescopes.  A  total  of  nearly 
250  new  radial  velocities  for  faint  and 
small  planetaries  has  now  been  determined 
at  the  Mount  Wilson  and  Palomar  Ob- 
servatories, or  at  the  Lick  Observatory,  or 
at  both.  A  plot  of  them  shows  that  very 
large  velocity  dispersion  occurs  around  the 
Galactic  center,  where  the  velocities  range 
from  —250  to  +300  km/sec.  The  large  dis- 



persion  hides  a  regular  radial-velocity  vari- 
ation of  8  km/sec  per  degree  of  longitude 
from  negative  to  positive  values,  as  the 
longitude  increases  and  passes  that  of  the 
Galactic  center,  for  the  planetaries  clustered 
around  the  nuclear  region.  Generally 
negative  velocities  appear  near  the  longi- 
tude of  the  apex,  and  positive  ones  around 
the  antapex.  This  pronounced  reflection 
of  the  sun's  Galactic  rotation,  together 
with  the  inconspicuous  radial-velocity  vari- 
ation in  the  nuclear  region,  may  be  an 
indication  that  the  system  of  planetary 
nebulae  has  relatively  small  Galactic  ro- 
tation, but  the  results  of  a  complete  analy- 
sis, which  will  be  undertaken  by  M. 
Schmidt  at  Leiden,  have  to  be  awaited 
before  such  a  conclusion  can  be  definitely 

Densities  of  Nebulae 

Minkowski  and  Osterbrock  made  point- 
by-point  density  determinations  in  NGC 
650/1  and  NGC  6720,  using  the  [O  II] 
A3727  intensity-ratio  method,  to  study  the 
spatial  structure  of  these  two  planetary 
nebulae.  The  outer  shell  of  NGC  6720, 
originally  discovered  by  Duncan,  has  a 
lower  density  and  also  an  even  more  pro- 
nounced filamentary  structure  than  the 
classical  bright  ring  and  center.  These 
inner  parts,  in  which  the  mean  electron 
density  is  of  the  order  of  103/cm3,  are, 
however,  also  extremely  filamentary.  NGC 
650/1  appears  from  these  observations  to 
be  a  nebula  rather  similar  in  structure  to 
NGC  6720,  but  seen  in  a  different  projec- 
tion on  the  plane  of  the  sky. 

Observations  by  Osterbrock  of  the  A3727 
intensity  ratio  in  the  Cygnus  Loop  showed 
that  the  electron  density  in  the  visible  fila- 
ments of  this  nebula  is  of  the  order  of  a 
few  hundred  per  cubic  centimeter.  The 
total  mass  of  the  whole  visible  filamentary 
nebula  is  probably  of  the  order  of  0.1  solar 
mass.  The  mass  of  un-ionized  and  there- 
fore dark  matter  inside  the  loop  must  be 
considerably  larger  than  the  mass  of  the 
bright  filamentary  system. 

Observations  of  the  A3727  ratios  in  the 
small  T  Tauri  nebulosity  and  in  the  bright- 
est Herbig-Haro  object  showed  that  their 
electron  densities  are  of  the  order  of 
104/cm3.  These  nebulae,  both  probably 
connected  with  stars  in  early  stages  of 
their  evolution,  differ  from  typical  plane- 
taries in  having  both  [O  I]  and  [O  II] 
forbidden  lines  in  their  spectra.  They 
therefore  contain  neutral  oxygen,  and 
hence  neutral  hydrogen,  and  the  well 
known  theory  of  radiative  ionization, 
worked  out  by  Stromgren,  shows  that  they 
cannot  be  ionized  in  the  ordinary  way  by 
ultraviolet  radiation.  It  is  possible,  though 
not  certain,  that  the  mechanism  of  ioniza- 
tion is  high-energy  corpuscular  radiation 
released  by  the  involved  stars. 

Osterbrock  and  Miss  Flather  analyzed 
the  electron-density  distribution  in  the 
Orion  nebula.  A  model  was  constructed, 
in  which  the  density  varies  smoothly  with 
distance  from  the  center,  and  which  re- 
produces in  the  mean  the  observations 
not  only  of  A3727  ratios  but  also  of  the 
[O  III]/[0  II]  and  H(3/[0  II]  ratios. 
However,  this  preliminary  model  predicts 
much  higher  thermal  radio  emission  by  the 
Orion  nebula  than  the  published  observa- 
tions indicate.  The  apparent  discrepancy 
shows  that  there  must  be  very  pronounced 
density  fluctuations  in  the  nebula,  and 
though  there  is  not  enough  observational 
information  to  specify  these  fluctuations 
completely,  a  model  can  be  constructed 
that  satisfies  all  the  observations. 

Expanding  Shells  around  Novae  and 
Super  novae 

Since  1935  Baade  has  kept  under  ob- 
servation at  the  100-inch  and  later  at  the 
200-inch  the  expanding  shells  around  Nova 
T  Aurigae  (1890),  Nova  Persei  (1901), 
Nova  Aquilae  (1918),  Nova  Cygni  (1920), 
and  Nova  Herculis  (1934).  In  order  to 
conclude  the  series  these  novae  have  been 
reobserved  under  the  best  conditions  dur- 
ing the  past  year.    Together  with  simul- 


taneous  measures  of  the  Doppler  compo- 
nents of  the  expansion,  these  observations 
should  furnish  reliable  data  for  the  abso- 
lute magnitudes  of  five  novae  and  provide 
a  valuable  check  for  the  absolute  magni- 
tudes of  novae  based  on  the  distance  modu- 
lus of  the  Andromeda  galaxy. 

Baade  also  reobserved  at  the  200-inch 
the  remnants  of  the  supernova  Ophiuchi  of 
1604.  Intercomparison  with  a  plate  taken 
at  the  200-inch  in  1950  clearly  showed  mo- 
tions in  the  nebulosity,  but  a  much  longer 
time  interval  will  be  necessary  before  a 
study  of  the  motions  becomes  profitable. 


For  several  years  an  extensive  program 
has  been  in  progress  for  the  study  of  the 
color-magnitude  relations  and  other  prop- 
erties of  the  stars  in  a  number  of  globular 
and  Galactic  clusters.  Such  observations 
provide  the  basic  data,  about  the  properties 
of  groups  of  stars  of  different  ages,  that  are 
necessary  to  trace  the  evolution  of  stars 
having  various  initial  masses. 

Two  extremely  distant  clusters,  found  by 
Abell  on  the  National  Geographic  Society- 
Palomar  Observatory  Sky  Survey,  have 
been  studied  by  E.  M.  Burbidge  and  San- 
dage.  The  clusters  are  numbers  3  and  4 
of  AbelPs  list  of  13  new  clusters.  A  series 
of  plates  of  both  clusters  was  obtained 
with  the  200-inch  in  four  observing  seasons 
beginning  in  1953.  These  plates,  together 
with  photometric  transfers  from  Selected 
Areas  51  and  57,  gave  color-magnitude 
diagrams  that  clearly  identify  the  clusters 
as  globular.  But  the  diagrams  were  ab- 
normal along  the  horizontal  branches.  In 
neither  cluster  does  the  horizontal  branch 
extend  to  the  blueward  side  of  the  RR 
Lyrae  domain;  the  stars  are  bunched  at 
the  red  side  of  the  horizontal  branch.  In 
this  respect  the  clusters  fit  into  the  se- 
quence of  globular  clusters  ordered  in  the 
following  way:  M  13,  M  10,  M  2,  M  92, 
M  15,  M  5,  M  3,  Abell  number  3,  Abell 
number  4.  Clusters  near  the  M  13  end  of 
the  sequence  have  most  of  their  horizontal- 
branch  stars  blueward  of  the  RR  Lyrae 
domain.  Clusters  near  the  M  3  end  have 
horizontal  branches  equally  populated 
blueward  and  redward  of  the  RR  Lyrae 
domain.  If  Mv  =  0.0  for  the  horizontal- 
branch  stars,  then  the  clusters  Abell  num- 
bers 3  and  4  are  125,000  parsecs  distant. 

This  is  twice  the  distance  of  the  Magellanic 
Clouds  and  makes  both  clusters  inter- 

Arp  has  pointed  out  that  the  sequence 
M  13,  M  10,  M  2,  M  92,  M  15,  M  5,  and 
M  3  also  puts  the  globular  clusters  in  order 
of  decreasing  mean  period  of  their  RR 
Lyrae  stars.  For  example,  the  mean  period 
of  the  Bailey  type  a  and  b  variables  in 
clusters  like  M  2,  M  92,  and  M  15  is  0.64 
day,  but  in  clusters  like  M  5  and  M  3  the 
mean  period  is  only  0.53  day.  Sandage  has 
developed  a  partial  theory  to  explain  this 
and  other  differences  in  the  RR  Lyrae 
variables  from  cluster  to  cluster.  It  is  be- 
lieved that  the  absolute  magnitudes  of  the 
horizontal  branches  of  clusters  are  not  the 
same.  Therefore  the  absolute  luminosities 
of  RR  Lyrae  stars  differ  from  cluster  to 
cluster.  Sandage's  theory  predicts  that  the 
difference  in  absolute  magnitude  is  related 
to  the  difference  in  the  mean  period  by 
AMv  =  3.0Alog  P.  Therefore  ^MV  be- 
tween the  RR  Lyrae  stars  in  M  3  and  M  15 
must  be  0.25  magnitude  to  explain  the 
mean-period  difference.  Differences  in  the 
period-light  amplitude  and  period-color 
relations  predicted  from  the  theory  can  be 
checked  by  observation  of  light-curves  in 
the  two  colors  B  and  V  for  variables  in 
M  3  and  M  15.  Such  a  study  for  M  3  was 
completed  and  reported  three  years  ago. 
The  comparison  study  of  variables  in  M  15 
was  started  this  year  by  Dr.  Mary  Con- 
nelley,  of  Mount  Holyoke  College,  and 
Sandage.  A  long  series  of  blue  and  yellow 
plates  of  M  15  were  obtained  with  the 
100-inch  telescope  during  the  report  year. 
A  two-color  photoelectric  sequence  ex- 
tending from  V  =  11.5  to  V  =  17.2  was  also 



obtained  in  M  15  with  the  100-inch.  The 
plates  are  being  measured,  and  a  check 
on  the  prediction  of  the  theory  will  be 

Work  on  NGC  5897  by  Sandage  and 
Schmidt,  on  NGC  6712  by  Sandage  and 
Norton,  and  on  NGC  6356  by  Sandage  and 
Wallerstein,  reported  last  year,  is  nearly 
complete.  Additional  stars  have  been 
measured  photoelectrically  in  each  of  these 
clusters  with  the  100-inch  telescope.  The 
three-color  UBV  photographic  photometry 
of  NGC  5897  by  Schmidt  shows  that  this 
cluster  has  a  normal  color-magnitude  dia- 
gram of  the  M  13,  M  92  type.  There  is  a 
strong  ultraviolet  excess,  which  is  normal 
for  all  globular  clusters  so  far  studied. 
Over  30  new  RR  Lyrae  stars  have  been 
found  in  NGC  6712  by  Norton  from  a 
series  of  200-inch  plates  taken  of  this 
cluster  in  1956.  Norton  is  obtaining  two- 
color  light-curves  of  the  variables  and  a 
color-magnitude  diagram  for  the  cluster. 
By  far  the  most  interesting  cluster  on  the 
current  program  is  NGC  6356.  It  is  in  a 
group  that  differs  from  halo  clusters  be- 
cause the  strength  of  the  metal  lines  in  the 
integrated  spectrum  is  high.  The  group, 
first  isolated  by  Morgan,  contains  NGC 
6304,  6356,  6440,  6441,  6624,  6637,  6638, 
6652,  6712,  and  6838.  With  the  exception 
of  NGC  6838,  all  the  clusters  are  very 
close  to  the  direction  of  the  Galactic  nu- 
cleus. NGC  6356  has  the  very  late  spectral 
class  of  G5  and  is  the  most  readily  ac- 
cessible from  these  latitudes.  A  color- 
magnitude  diagram  by  Sandage  and  Wal- 
lerstein for  this  cluster  is  nearly  complete. 
They  find  no  horizontal  branch  whatso- 
ever to  4  magnitudes  below  the  top  of  the 
giant  branch.  In  this  respect  the  color- 
magnitude  diagram  is  similar  to  that  of 
NGC  6838,  which  is  another  member  of 
the  group.  The  color-magnitude  diagram 
for  NGC  6838  has  been  obtained  by  Dr. 
W.  Becker  of  Basel,  Switzerland,  and  by 
Dr.  J.  Cuffey  of  Indiana  University.  Spe- 
cial spectroscopic  and  UBV  photometric 
observations  of  the  field  around  NGC  6356 

are  being  made  by  Wallerstein  to  deter- 
mine the  interstellar  reddening  in  front  of 
NGC  6356  so  as  to  check  whether  the  in- 
trinsic colors  of  the  giant  branch  of  this 
cluster  are  normal. 

A  preliminary  survey  of  the  magnitudes 
and  colors  of  stars  in  M  13  was  made  by 
Baum  in  1953.  When  the  color-magnitude 
diagram  was  adjusted  to  place  the  cluster- 
type  variables  at  0.0  absolute  magnitude  it 
was  found  that  the  main  sequence  fell 
about  0.3  magnitude  at  the  blue  side  of  the 
standard  main  sequence.  Because  of  the 
importance  of  this  result  for  many  theo- 
retical problems  a  cooperative  program 
was  arranged  between  Dr.  W.  A.  Hiltner 
of  the  Yerkes  Observatory,  Dr.  Harold 
Johnson  of  the  Lowell  Observatory,  and 
Baum  and  Sandage  to  extend  the  measure- 
ments to  many  additional  stars  of  the 
cluster  and  thereby  to  reduce  the  statistical 
uncertainty.  Many  additional  photoelectric 
measurements  of  B  and  V  magnitudes 
have  been  made  with  the  200-inch  by 
Baum  and  Johnson,  and  with  the  42-inch 
of  the  Lowell  Observatory  and  the  82-inch 
of  the  McDonald  Observatory  by  Hiltner 
and  Johnson.  Sandage  has  measured  and 
reduced  B  and  V  magnitudes  of  250  faint 
stars  in  M  13  to  locate  the  main  sequence. 
Six  plates  in  each  color  were  measured 
and  reduced  by  means  of  the  extensive 
photoelectric  sequence  provided  by  the 
three  photoelectric  observers.  The  final  re- 
sults of  the  entire  program  are  not  yet 
definitive,  but  the  main  sequence  does  not 
appear  to  be  as  far  from  the  standard 
main-sequence  stars  as  the  early  results  of 
1953  indicated. 

Thirty-nine  stars  in  M  5  have  been  meas- 
ured to  fainter  than  V  =  20  by  Arp  photo- 
electrically in  three  colors  in  this  near-by 
globular  cluster.  The  photoelectric  meas- 
ures clearly  define  the  complete  color- 
magnitude  diagram  of  M  5  from  the 
brightest  stars,  down  to  the  main  sequence 
and  including  2  magnitudes  of  that  main 
sequence.  As  time  permits,  a  few  fainter 
stars  will  be  measured;  the  observational 


phase  of  this  problem  is  nearly  complete, 
however.  Enough  inner  stars  were  meas- 
ured so  that  all  the  previous  measures  on 
the  bright  part  of  the  color-magnitude  dia- 
gram can  be  converted  to  the  BV  system. 
A  supplementary  sample  of  fainter  stars 
will  be  measured  on  photographic  plates, 
and  the  ultraviolet  colors  will  be  analyzed. 

The  cluster  NGC  2158  has  been  sur- 
veyed photoelectrical^  by  Arp;  despite  its 
very  sparse  number  of  faint  stars,  it  ap- 
pears, from  preliminary  results,  to  have  a 
globular-cluster-like  color-magnitude  dia- 
gram. The  analysis  of  the  globular  cluster 
M  22  by  Arp  and  Melbourne  and  the  ad- 
joining nuclear  star  field  by  Arp  has  been 

The  results  of  a  cooperative  study  of  the 
Galactic  cluster  NGC  7789  by  E.  M.  Bur- 
bidge,  of  Yerkes  Observatory,  and  San- 
dage,  reported  last  year,  have  been  com- 
pleted. NGC  7789  has  a  color-magnitude 
diagram  like  that  of  NGC  752  but  much 
richer.  The  main-sequence  break  point 
occurs  at  Af„«  +  2,  (B-V)o^  +  0.35. 
There  is  a  Hertzsprung  gap  0.4  magnitude 
wide  extending  from  (B  — V)o  =  0.4  to 
(B  —  V)  o  =  0.8.  From  there,  the  giant  se- 
quence begins  and  extends  to  Mv=  — 1.5, 
(B  —  V)  o  =  1.6.  The  important  point  is 
that  the  giant  branch  intersects  the  giant 
sequence  of  M  11,  which  is  a  considerably 
younger  cluster  whose  main-sequence 
break  point  occurs  at  Mv= —1.  This  re- 
sult illustrates  the  funnel  effect  described 
earlier  and  is  the  strongest  observational 
evidence  suggesting  that  the  masses  of 
luminosity  class  III  giant  stars  range  from 
1.2  O  to  3.0  O. 

The  color-magnitude  diagram  and  lu- 
minosity function  for  the  Galactic  cluster 
NGC  188  have  been  completed  by  Dr.  S. 
van  den  Bergh,  of  Perkins  Observatory, 
and  Sandage.  A  UBV  photoelectric  se- 
quence of  54  stars  extending  from  V  =  8.0 
to  V=  16.7  was  set  up  with  the  60-inch  tele- 
scope. A  few  additional  critical  observa- 
tions of  the  five  faintest  stars  were  made 
by  Arp  with  the  200-inch.    Photographic 

plates  for  the  color-magnitude  diagram 
were  obtained  with  the  60-inch.  Photo- 
graphic plates  for  the  luminosity  function 
were  obtained  with  the  48-inch  schmidt. 
The  color-magnitude  diagram  has  the 
same  general  shape  as  that  of  M  67.  There 
is  no  Hertzsprung  gap,  and  there  is  a 
3-magnitude  rise  in  the  giant  branch  from 
the  main-sequence  termination  point  to 
the  top  of  the  giant  sequence.  But  the 
important  difference  between  NGC  188 
and  M  67  is  the  absolute  magnitude  of  the 
main-sequence  break  point.  The  main  se- 
quence in  NGC  188  terminates  at  least 
0.5  magnitude  fainter  than  M  67.  On  cur- 
rent evolutionary  theories  this  means  that 
NGC  188  is  about  1.5  billion  years  older 
than  M  67.  But,  to  date,  M  67  has  been 
the  oldest  Galactic  cluster  known;  its  age 
has  been  taken  as  5  to  6  X  10°  years.  If  the 
present  data  for  NGC  188  prove  to  be 
correct,  we  have  a  cluster  whose  age  is 
6.5  to  7.5  X  109  years.  With  the  preliminary 
modulus  which  these  first  results  indicate, 
the  subgiant  branch  of  the  color-magni- 
tude diagram  for  NGC  188  passes  through 
the  position  of  the  field  stars  h  Eridani  and 
u  Herculis  in  the  H-R  diagram.  These 
stars  have  always  been  considered  candi- 
dates for  stars  older  than  M  67  because  in 
the  diagram  they  are  fainter  than  M  67. 
Additional  observations  on  NGC  188  are 
being  made  to  check  the  results. 

Further  photoelectric  observations  in  the 
Galactic  clusters  NGC  2269,  2309,  2401, 
and  2453  were  carried  out  by  Schmidt. 
Owing  to  weather  conditions  this  program 
is  still  unfinished.  Photoelectric  observa- 
tions in  the  Galactic  cluster  NGC  6939 
were  completed  and  are  being  reduced. 

In  a  joint  program,  Arp,  Sandage,  and 
Schmidt  have  measured  photoelectrically 
about  100  stars  in  NGC  6940.  This  cluster 
is  similar  to  the  Hyades  but  richer.  Work 
is  continuing  on  fainter  stars  in  this  clus- 
ter, and  Wallerstein  is  taking  spectra  in 
NGC  6940.  Arp  has  measured  about  40 
stars  photoelectrically  in  M  37,  a  Galactic 
cluster  similar  to  the  Hyades  or  NGC  752 


but  with  a  strong  giant  branch.  Photo- 
graphic plates  of  the  cluster  remain  to  be 
measured  to  complete  the  diagram. 

Further  photoelectric  observations  on 
the  UBV  system  were  made  by  Matthews 
of  an  unnamed  cluster  of  early-type  stars 
possibly  connected  with  the  Association  I 
Camelopardalis.  The  cluster  is  situated  in 
a  region  of  very  heavy  obscuration  which 
is  responsible  for  a  pronounced  increase  of 
21-cm  radiation. 

White  Dwarfs  in  M  67 

Because  of  its  great  age  M  67  should 
contain  a  considerable  number  of  white 
dwarfs.  Moreover,  they  should  be  within 
reach  of  the  100-inch  and  200-inch  tele- 
scopes, since  the  distance  modulus  of  M  67 
is  only  m—M  —  9.6.  Therefore  in  the  win- 
ter of  1956-1957  Baade  started  at  the  200- 
inch  a  search  for  white  dwarfs  in  M  67 
which  was  continued  during  the  past  sea- 
son. Only  the  central  area  with  a  radius 
of  8  minutes  of  arc  has  been  searched. 
Two  sets  of  photographic,  photovisual, 
and  ultraviolet  exposures  were  available. 
Similar  sets  of  Selected  Area  51  provided 
the  necessary  standardization. 

Altogether  32  stars  were  found  which 
have  color  indices  — +  0?20.  They  range 
in  magnitude  from  about  19.5  photo- 
graphic to  the  limit  of  the  plates,  which  is 
close  to  mPg:=22.7.  There  seems  to  be  little 
doubt  that  these  very  faint  blue  stars  are 
white  dwarfs  of  M  67,  since  they  stand 
out  by  their  blueness  among  the  stars  of 
similar  brightness.  Actually  the  number 
of  white  dwarfs  in  the  center  area  of  M  67 
may  be  considerably  larger,  since  the  color 
limit  mentioned  above  is  rather  arbitrary. 
However,  color  alone  becomes  an  increas- 
ingly dangerous  criterion  as  one  goes  to 
larger  color  indices,  where  only  the  spec- 
trum can  decide  whether  one  is  dealing 
with  a  white  dwarf  or  not. 

Absolute  Magnitudes  of  Cepheids 

The  discovery  by  Irwin  (1955)  followed 
by  Kraft  (1957)  and  van  den  Bergh  (1957) 

of  cepheids  in  Galactic  clusters  has  made 
it  possible  for  the  first  time  to  derive  fun- 
damental parameters  for  at  least  some 
cepheids  in  our  own  Galaxy.  By  accurate 
three-color  photometry  of  the  cepheids  and 
the  clusters  containing  them  it  is  possible 
to  derive  accurate  absolute  magnitudes  and 
intrinsic  colors  of  the  cepheids.  One  of 
the  primary  purposes,  of  course,  is  to  de- 
termine an  absolute  zero  point  for  the 
period-luminosity  relation,  but  normal  yel- 
low-blue and  ultraviolet  color  indices  will 
also  be  very  important.  It  is  even  possible, 
using  evolutionary  theory,  to  derive  the 
age  of  these  cepheids,  and  also  a  better 
estimate  of  their  masses.  In  order  to  ob- 
tain definitive  results  Arp  and  Sandage 
have  undertaken  to  measure,  with  the  60- 
inch  and  100-inch  telescopes,  color-magni- 
tude diagrams  of  all  clusters  containing 
cepheids  and  the  light-curves  of  the  cephe- 
ids in  them  that  can  be  reached  from  the 
North.  Dr.  R.  P.  Kraft  of  Indiana  Uni- 
versity has  cooperated  in  an  important  part 
of  this  program  by  obtaining  spectra  in 
these  clusters.  Sandage  has  finished  the 
results  for  the  cepheid  CF  Cassiopeiae  in 
the  Galactic  cluster  NGC  7790.  He  is  com- 
pleting material  for  two  more,  CE  Cassi- 
opeiae A  and  B  in  the  same  cluster.  Arp 
has  completed  the  results  for  EV  Scuti 
in  NGC  6664.  Arp,  Sandage,  and  Miss 
Stephens  have  made  measurements  of  DL 
Cassiopeiae  in  NGC  129.  A  few  check 
measures  are  still  needed.  To  complete  the 
known  group,  the  following  cepheids,  lo- 
cated in  anonymous  clusters,  are  being 
analyzed  by  Arp  and  Sandage:  CV 
Monocerotis,  XZ  Canis  Majoris,  AO  Canis 

Dynamics  of  Clusters 

Oort  made  an  extensive  dynamical  study 
of  the  globular  cluster  M  3  which  he  had 
started  in  Leiden  in  collaboration  with 
Dr.  van  Herk.  This  work  was  based  on 
star  counts  made  by  Sandage.  Some  plates 
were  taken  with  the  48-inch  schmidt  tele- 
scope  in  order  to  extend  the  star-count 


data  to  large  distances  from  the  center. 
The  investigations  have  resulted  in  a  fairly 
satisfactory   theory   of  the  cluster's   struc- 

ture. This  analysis  indicates  that  in  the 
outer  regions  there  must  be  a  strong  pref- 
erence for  radial  motions. 


The  Local  Group 

Miss  Swope  finished  the  work  on  the 
color-magnitude  diagram  of  the  Draco  sys- 
tem using  plates  taken  by  Baade.  In  agree- 
ment with  expectation  the  diagram  is  of 
the  same  type  as  that  of  the  globular  clus- 
ters of  the  Galactic  halo.  It  also  conforms 
to  the  general  rule  that  large  numbers  of 
cluster-type  variables  (the  Draco  system 
contains  about  300)  can  be  expected  only 
if  the  horizontal  branch  on  the  red  side 
of  the  variable  gap  is  strongly  populated. 

Four  photoelectric  measures  have  been 
made  by  Arp  in  Draco  for  the  calibration 
of  the  above  program.  They  indicate  that 
the  original  photographically  transferred 
magnitude  scale  is  quite  good,  but  that 
the  color  indices  should  be  moved  about 
0.2  magnitude  toward  the  red,  making  the 
Draco  system  fall  more  into  line  with 
globular-cluster  color-magnitude  diagrams. 

During  the  past  year  the  observations 
of  the  variable  stars  in  the  Ursa  Minor 
system  and  the  Leo  II  system  were  con- 
cluded by  Baade.  In  the  near  future  de- 
tailed data  should  be  available  about  the 
variables  and  the  color-magnitude  dia- 
grams of  three  dwarf  E  galaxies,  to  which 
may  be  added  the  Sculptor  system  now  un- 
der investigation  at  the  RadclifTe  Observa- 
tory. It  is  also  hoped  that  the  series  of 
observations  of  NGC  185  which  Baade  has 
carried  out  at  the  200-inch  since  1950  will 
be  sufficient  to  determine  the  types  of  the 
variables  in  this  much  larger  E  galaxy. 
NGC  185  is  too  far  away  to  reach  the 
cluster-type  variables,  but  it  contains  long- 
period  red  variables  in  large  numbers.  The 
investigation  of  these  variables  should  pro- 
vide much-needed  data  about  the  long- 
period  variables  of  the  Population  II. 

In  a  start  on  the  reduction  of  the  photo- 
graphic plates  of  NGC  6822  taken  over 
the  years  by  many  observers,  Arp  has  ob- 

tained thirteen  photoelectric  calibration 
stars  bracketing  the  magnitude  level  of  the 

A  photoelectric  sequence  of  about  twelve 
stars  going  almost  to  V  =  21  magnitude 
has  been  established  by  Arp  in  the  near-by 
dwarf  galaxy  in  Sextans.  Measurements 
on  photographic  plates  of  about  fifty  mem- 
bers enable  a  preliminary  color-magnitude 
diagram  to  be  plotted.  The  photoelectric 
sequence  will  be  strengthened  and  carried 
to  fainter  magnitudes;  more  stars  will  be 
measured  in  the  color-magnitude  diagram, 
and  photographic  plates  taken  by  Baade 
will  be  measured  to  obtain  light-curves  of 
the  dozen  or  so  cepheids  contained  in  the 

An  investigation  was  continued  by  Code 
and  Dr.  T.  E.  Houck  of  Washburn  Ob- 
servatory on  the  luminosities  of  the  bright- 
est OB  stars  in  near-by  galaxies.  Houck 
observed  photoelectrically  on  a  three-color 
system  a  number  of  bright  blue  stars  in 
NGC  6822  and  M  33,  and  spectra  were 
obtained  by  Code  of  some  of  these  objects 
with  the  prime-focus  spectrograph  on  the 
200-inch  reflector.  In  general  these  ob- 
jects resemble  the  most  luminous  early- 
type  stars  in  our  own  Galaxy,  although 
they  all  show  strong  broad  H  and  K  lines 
presumably  of  interstellar  origin.  The  bolo- 
metric  magnitudes  of  the  brightest  stars 
are  in  excess  of  —  llm.  From  spectra  pre- 
viously obtained  by  Code  and  Houck  of 
objects  in  the  Large  Magellanic  Cloud  a 
search  was  made  for  analogous  stars  in 
our  own  Galaxy.  A  comparison  of  the 
Large  Magellanic  Cloud  star  HDE  269700 
with  the  star  £  Scorpii  was  made.  The 
spectra  of  these  two  objects  are  very  similar 
and  indicative  of  an  extremely  high  lu- 
minosity. From  the  presently  accepted 
distance  moduli  of  the  Large  Cloud  and 
the   I    Scorpii   association,   respectively,   a 



visual  absolute  magnitude  of  —  9m  is  found 
for  these  two  objects  and  a  bolometric 
absolute  magnitude  of  the  order  of  — 11™8. 
These  stars  show  some  evidence  of  incipi- 
ent instability  in  the  form  of  light-varia- 
tion and  P-Cygni  emission.  The  stability 
to  radiation  pressure  was  discussed. 

Although  the  spectra  and  colors  of  the 
early-type  stars  in  the  Large  Magellanic 
Cloud  suggest  that  these  objects  are  sub- 
stantially similar  to  those  in  our  own  Gal- 
axy, the  Small  Magellanic  Cloud  data  do 
not  show  these  similarities.  In  general  all 
the  spectra  obtained  by  Code  and  Houck 
in  the  Small  Cloud  possess  weaker  lines 
than  corresponding  Large  Cloud  objects, 
and  the  interpretation  of  these  results  as 
reflecting  a  lower  metal  abundance  for 
the  Small  Magellanic  Cloud  is  difficult  to 
avoid.  This  important  result  appears  to 
be  in  contradiction  to  the  results  of 
Thackeray  and  Feast. 

Rotation  and  Dispersion  of  Stellar 

Spectroscopic  observations  with  the  Palo- 
mar  prime-focus  spectrograph  of  the  H  II 
regions  in  M  81  were  continued  by 
G.  Munch  in  order  to  improve  the  rota- 
tional-velocity-curve previously  reported. 
In  a  spectrum  of  the  nucleus  of  M  81  ob- 
tained in  the  yellow  region  strong  nebular 
lines  of  [O  III]  were  discovered.  Another 
plate  in  the  ultraviolet  (dispersion  66 
A/mm)  showed  also  the  [O  II]  doublet 
and  the  [Ne  III]  lines  at  A3868  in  emission. 
These  lines  are  narrow  and  resemble  those 
observed  in  H  II  regions,  extending  out  to 
a  distance  of  4  seconds  of  arc  from  the 
nucleus.  With  Sandage's  distance  modulus 
m—M  —  XIX  the  diameter  of  the  emitting 
region  would  be  100  parsecs.  The  relative 
intensities  of  the  [O  II]  doublet  correspond 
to  an  electron  density  around  103,  decreas- 
ing slightly  outward.  The  emission  lines 
are  inclined  at  an  angle  of  2?5±0?4  with 
respect  to  the  direction  of  dispersion,  indi- 
cating an  angular  velocity  of  560  ±  40  km/ 
sec  per  kpsc  This  value  is  quite  close  to 

the  angular  velocity  of  the  Galactic  system 
at  0.11  kpsc  from  the  nucleus  (445  km/sec 
per  kpsc,  according  to  Kwee,  Muller,  and 
Westerhout).  Tentatively  it  is  believed 
that  this  emission  nebula  in  the  center  of 
M  81  is  of  the  same  nature  as  the  radio 
source  observed  in  the  center  of  the  Galac- 
tic system. 

Oort  has  cooperated  with  Minkowski 
in  the  investigation  of  velocity  dispersions 
and  rotations  of  elliptical  and  SO  galaxies, 
and  of  the  amorphous  parts  of  spirals, 
including  in  particular  an  extension  of  the 
dynamic  study  of  the  SO  galaxy  NGC  3115 
made  by  Oort  in  1939.  Such  observations 
are  essential  for  understanding  the  dy- 
namics of  elliptical  galaxies  and  the  initial 
conditions  under  which  they  were  formed. 
They  are  also  fundamental  for  finding  the 
mass  density  in  the  universe. 

Poor  weather  restricted  the  observations. 
Spectra  of  NGC  3115  indicate  that  the  ro- 
tational velocity  does  not  increase  linearly 
with  the  distance  from  the  nucleus  as  older 
results  had  implied.  The  new  results  sug- 
gest that  up  to  about  50  seconds  from  the 
center  the  mass  density  may  be  approxi- 
mately proportional  to  that  of  the  light- 
emission,  but  that  at  greater  distances  the 
distribution  of  mass  becomes  practically 
homogeneous  while  the  light-density  con- 
tinues to  diminish  rapidly.  The  inference 
rests  on  only  three  spectra  and  needs  to 
be  confirmed. 

Magnitudes  and  Energy  Curves  of 

Elliptical  galaxies  provide  the  indicators 
by  which  the  distances  of  clusters  of  gal- 
axies are  fixed.  In  order  that  proper  correc- 
tion may  be  made  for  evolutionary  changes 
it  is  important  that  information  about 
their  stellar  content  be  available. 

During  the  past  three  years  new  six- 
and  eight-color  observations  of  galaxies  of 
all  types  as  well  as  a  representative  sam- 
pling of  various  kinds  of  stars  have  been 
obtained  by  Baum.  The  best  synthesis 
thus  far  achieved  has  led  Baum  to  the  con- 



elusion  that  22  per  cent  of  the  visual  light 
and  5  per  cent  of  the  total  light  (bolo- 
metric)  of  the  large  elliptical  galaxies  in 
Virgo  come  from  Population  II  stars  and 
the  remainder  from  old  Population  I  stars. 
The  contribution  of  Population  II  in  M  32 
is  somewhat  larger.  Three  dwarf  ellip- 
tical galaxies  have  now  been  measured 
photoelectrically  with  enough  precision  to 
show  that  they  are  not  similar  to  the  large 
ellipticals  but  have  color  indices  like  those 
of  the  globular  clusters. 

From  data  now  in  hand,  the  integrated 
color  indices  of  the  old  stellar  systems 
ranging  from  globular  clusters  at  one  ex- 
treme to  large  elliptical  galaxies  at  the 
other  have  been  plotted  against  absolute 
magnitude,  and  the  results  indicate  a  clear 
dichotomy.  All  old  systems  (that  is,  those 
that  no  longer  produce  new  stars)  brighter 
than  —16  absolute  magnitude  are  in  the 
class  dominated  by  old  Population  I, 
whereas  those  fainter  than  —14  absolute 
magnitude  are  in  the  class  dominated  by 
Population  II.  The  transition  from  one 
class  to  the  other  occurs  between  — 14  and 
—  16  absolute  magnitude.  Roughly  speak- 
ing, M  32  represents  the  smallest  size  in 
the  old  Population  I  class. 

Tifft  has  completed  the  study  with  mul- 
ticolor photoelectric  photometry  of  57  gal- 
axies; 8  galaxies  were  also  observed  in  the 
infrared.  The  color  system  has  been  placed 
on  an  absolute  energy  basis.  Dependence 
of  color  on  nebular  type  is  well  marked. 
The  color  of  the  nuclei  depends  on  the  in- 
clination of  the  galaxy  in  a  different  way 
in  spirals  and  in  ellipticals.  The  variation 
of  color  with  radius  also  differs  between 
FG-type  and  F-type  galaxies.  Synthetic 
models  of  five  different  types  of  nuclei 
have  been  computed.  K  nuclei  can  be  rep- 
resented by  a  mixture  of  old  Population  I 
and  II  stars;  F,  FG,  and  probably  AF 
nuclei  can  be  produced  by  adding  various 
numbers  of  young  blue  stars  to  K  galaxies. 
A-type  galaxies  appear  to  be  completely 
dominated  by  young  stars.  Some  polariza- 
tion measurements  were  made  in  dark 
lanes  of  M  31  and  M  81. 

Observations  of  the  energy  distribution 
of  galaxies  were  continued  by  Code  with 
the  photoelectric  scanning  spectrograph. 
In  addition  to  the  photoelectric  scans, 
widened  spectra  of  the  nuclei  have  been 
obtained.  The  combined  data  of  line  pro- 
files and  continuum  are  being  used  to 
form  synthetic  stellar  populations.  A  dip 
has  been  found  in  the  continuum  of  ellip- 
ticals in  the  region  of  the  Mg  I  A5167-84 
triplet  that  is  characteristic  of  giant-type 
spectra  in  contradistinction  to  dwarfs. 

Distance  Scale 

A  reanalysis  of  all  steps  required  to  es- 
tablish the  extragalactic  distance  scale  has 
led  Sandage  to  the  conclusion  that  the 
Hubble  expansion  parameter,  which  re- 
lates redshifts  and  distances,  is  probably  as 
low  as  H  =  75  km/sec  per  106  parsecs.  This 
value  represents  a  total  change  of  a  factor 
of  7  in  Hubble's  1936  scale  of  distances. 
The  correction  is  composed  of  two  parts. 
A  correction  of  about  a  factor  of  3  for 
galaxies  within  the  local  group  comes  pri- 
marily from  the  change  in  zero  point  of 
the  cepheid  period-luminosity  relation 
(Baade,  Blaauw  and  H.  R.  Morgan,  and 
Mineur) .  The  second  factor  of  about  2  for 
distances  beyond  the  range  of  cepheid  vari- 
ables (i.e.,  beyond  the  local  group  and  the 
near-by  M  81,  M  101  groups)  comes  from 
the  misidentification  in  1936  of  the  H  II 
regions  for  brightest  stars.  By  filter  pho- 
tography, it  has  become  clear  that  all  the 
bright  "resolved"  knots  in  galaxies  with 
velocities  in  the  range  from  600  to  2000 
km/sec  are  H  II  regions  like  the  Orion 
nebula  rather  than  stars.  The  erroneous 
identification  in  1936  gave  distances  that 
were  too  small.  It  is  only  because  of  the 
great  technical  advances  of  Ha  plates  since 
1936  that  this  situation  has  now  been 

As  part  of  this  examination  of  the  dis- 
tance scale,  Sandage  concluded  that  the 
period-luminosity  relation  of  the  cepheid 
variables  is  not  as  simple  as  previously  be- 
lieved.   The    physical    relation    for    these 


oscillating  stars  appears  to  be  an  equation 
between  period,  luminosity,  and  intrinsic 
color  (or  temperature),  and  the  approxi- 
mation of  a  three-parameter  relation 
/  (P,  Mv,  B  —  V)  by  the  two  parameters  of 
period  and  luminosity  introduces  intrinsic 
scatter.  The  hypothetical  treatment  pre- 
dicts that  the  period-luminosity  relation 
for  cepheids  should  have  an  intrinsic 
spread  of  1.2  magnitudes  if  read  at  a  given 
period.  Arp's  two-color  light-curves  for  a 
number  of  cepheid  variables  in  the  Small 
Magellanic  Cloud  confirm  the  theory  in 
its  general  form,  but  more  data  are  needed 
before  the  full  implications  of  the  results 
for  the  distance  scale  are  known. 

In  a  program  to  provide  photoelectric 
scales  in  near-by  galaxies  that  have  exten- 
sive photographic  coverage  of  the  cepheid 
variables,  bright  local  standards  have  been 
established  by  Arp  and  Sandage  in  NGC 
2403,  M  81,  and  M  101. 

Velocities  of  Galaxies 

In  last  year's  Annual  Report  a  pro- 
cedure was  briefly  described  by  Baum  by 
which  both  the  redshifts  and  the  magni- 
tudes of  remote  galaxies  can  be  determined 
by  photoelectric  photometry.  Each  galaxy 
is  observed  in  eight  colors,  and  the  data 
are  reduced  to  a  plot  of  radiated  energy 
as  a  function  of  wavelengths.  If  a  galaxy 
is  redshifted,  this  whole  energy-distribu- 
tion-curve is  shifted  toward  longer  wave- 
lengths. In  this  manner,  redshifts  can  be 
determined  for  galaxies  considerably  be- 
yond the  range  that  can  be  reached  with 
the  spectrograph.  The  photoelectric  ob- 
servations also  provide  a  direct  measure  of 
bolometric  magnitude. 

A  preliminary  analysis  has  been  made  of 
the  results  thus  far  in  hand.  These  include 
observations  in  six  clusters  of  galaxies 
ranging  from  the  near-by  cluster  in  Virgo 
to  one  of  the  remotest  known  clusters 
with  a  redshift  of  0.4A.  When  the  loga- 
rithms of  the  redshifts  are  plotted  against 
bolometric  magnitudes,  the  points  (one  for 

each  cluster  of  galaxies)  all  fall  remarkably 
close  to  a  straight  line  of  slope  5. 


The  search  for  supernovae  has  been 
continued  by  Zwicky  in  cooperation  with 
the  Steward  Observatory,  the  Lick  Ob- 
servatory, and  the  observatory  of  the  Uni- 
versity of  Berne  in  Switzerland.  Humason 
joined  the  group,  after  his  retirement, 
searching  in  particular  for  supernovae  in 
near  and  medium-distant  clusters  of  gal- 
axies (as  far  as  the  Corona  Borealis  clus- 
ter). Mr.  H.  S.  Gates,  continuing  his 
search  with  the  18-inch  schmidt  telescope, 
found  three  supernovae  in  NGC  4374 
(  =  Messier  84)  in  the  Virgo  cluster,  NGC 
1365  in  the  Fornax  cluster,  and  NGC  5236 
(  =  Messier  83).  The  search  at  Palomar 
was  in  part  supported  by  funds  made  avail- 
able by  the  National  Science  Foundation. 

The  light-curves,  in  different  colors,  of 
recent  supernovae  were  followed  as  far  as 
possible  by  Zwicky  with  the  200-inch 
prime-focus  spectrograph;  miscellaneous 
observers  contributed  plates  obtained  with 
the  48-inch  schmidt  telescope  and  the  re- 
flectors on  Mount  Wilson.  Arp  and  San- 
dage have  started  a  program  to  set  up 
photoelectric  scales  near  all  galaxies  in 
which  supernovae  have  occurred.  Thus 
far  photoelectric  sequences  have  been  set 
up  in  NGC  3992  and  NGC  4214.  It  is  in- 
tended to  continue  the  supernovae  search 
vigorously  for  several  years  and  to  look 
also  for  supernovae  in  a  few  of  the  most 
distant  clusters  of  galaxies  for  the  purpose 
of  a  more  reliable  distance  determination 
than  is  available  at  the  present  time. 

Density  of  Gas  Clouds 

Minkowski  and  Osterbrock  attempted 
to  determine  the  electron  density  of  the 
interstellar  matter  that  emits  the  [Oil] 
A3727  doublet  in  the  elliptical  galaxy  NGC 
1052.  Though  the  two  components  are 
completely  blended  by  high-velocity  dis- 
persion of  the  system,  the  mean  wave- 
length  of  the  blend   was   measured   and 



compared  with  the  mean  wavelength  cor- 
responding to  various  relative  intensities 
of  the  components,  and  therefore  to  vari- 
ous electron  densities.  The  result  is  that 
the  intensity  ratio  is  near  the  asymptotic 
low-density  limit,  and  the  density  is  there- 
fore of  the  order  of  or  less  than  200/cm3 — 
a  rather  low  value;  from  the  strength  of 
the  line  it  can  therefore  be  deduced  that 
there  is  a  large  amount  of  ionized  inter- 
stellar gas  in  this  galaxy.  NGC  1052  is  an 
elliptical  galaxy  with  a  very  strong  A3727 
line,  and  it  is  probable  that  in  other  ellip- 
ticals with  weaker  emission  lines  the  den- 
sity of  ionized  interstellar  matter  is  even 
lower.  Osterbrock  continued  observing 
other  elliptical  galaxies  with  emission  lines, 
in  particular  NGC  4278,  in  which  the  lines 
are  very  strong.  The  plates  are  now  being 

Catalogue  of  Galaxies  and  Clusters  of 

Herzog  and  Zwicky  have  continued 
their  work  on  the  positions  and  apparent 
photographic  magnitudes  of  some  35,000 
of  the  brightest  galaxies  as  well  as  on  the 
positions,  populations,  diameters,  and  es- 
timated distances  of  about  10,000  clusters 
of  galaxies.  This  work  was  partly  sup- 
ported by  a  grant  from  the  Office  of  Naval 

Multiple  Galaxies 

The  direct  photography  of  multiple  gal- 
axies has  been  continued  by  Zwicky  for 
the  purpose  of  investigating  the  association 
of  types  and  the  relative  luminosities. 
This  project  has  included  a  study  of  the 
spectral  types  of  members  of  multiple  gal- 
axies as  well  as  their  dispersion  in  radial 
velocities.  A  group  of  about  100  elliptical 
galaxies  occupying  an  area  near  R.  A. 
lh  23m,  Decl.  —1°  34',  and  including  ob- 
jects NGC  535,  NGC  538,  NGC  541, 
NGC  543,  NGC  545,  and  others,  has  been 
under  special  study  for  the  determination 
of  spectral  types,  of  velocity  dispersion, 
and  of  the  possible  rotation  of  the  group 

of  galaxies  as  a  whole.   Humason  has  co- 
operated in  the  reduction  of  the  data. 

Spectra  of  Intergalactic  Luminous  Matter 

Two  successful  attempts  were  made  by 
Zwicky  in  January  and  in  April  to  obtain 
the  spectra  of  the  very  faint  (23rd  magni- 
tude per  square  second,  or  less)  extended 
luminous  bridges  and  clouds  of  intergalac- 
tic matter  that  connect  many  widely  sepa- 
rated galaxies. 

The  spectrum  of  a  pair  of  galaxies  with 
a  faint  bridge  between  them  and  a  very 
long  plume-like  extension  was  obtained 
first.  It  proved  to  be  a  pure  absorption 
spectrum  with  pronounced  H  and  K  lines 
as  well  as  the  G  band,  all  in  absorption, 
connecting  the  two  galaxies  and  extending 
along  the  plume.  It  is  expected  that  most 
of  the  luminous  intergalactic  bridges  will 
prove  to  be  of  this  type,  that  is,  that  they 
are  composed  of  stars,  presumably  old 
stars.  The  apparent  velocities  of  recession 
are  about  6280  km/sec  for  the  fainter 
galaxy  and  6380  km/sec  for  the  brighter 

A  special  effort  was  then  made  to  choose 
a  luminous  bridge  that  promised  to  show 
emission  lines.  The  group  at  R.  A.  llh 
8m  5s,  Decl.  +29°  2'  24",  seemed  particu- 
larly promising  for  the  purpose.  In  con- 
nection with  this  object  V.  A.  Ambar- 
tsumian,  on  red  and  blue  prints  of  the  Na- 
tional Geographic  Society-Palomar  Ob- 
servatory Sky  Survey,  had  discovered  that 
the  southernmost  tiny  knot  of  the  multiple 
galaxy  (at  the  end  of  the  stringlike  lumi- 
nous bridge)  is  extraordinarily  blue.  Ex- 
cellent blue  spectra  were  obtained  with  the 
slit  covering  the  central  irregular  barred 
spiral,  the  elliptical  or  SO  galaxy  south  of 
it,  and  the  thin  bridge  with  the  blue  knot. 
Whereas  the  spiral  proved  to  be  quite  blue, 
the  elliptical  galaxy  gives  a  much  redder 
continuous  spectrum,  but  both  galaxies 
show  strong  A3727  emission  lines.  The 
little  knot,  in  spite  of  being  much  fainter 
than  the  two  main  galaxies,  has  a  A3727 
emission  line  almost  as  strong  as  those  of 



the  main  galaxies,  which  easily  accounts 
for  its  extreme  blueness.  As  a  matter  of 
fact,  the  A3727  line  appears  in  emission 
all  along  the  exceedingly  thin  and  faint 
bridge  from  the  bright  elliptical  galaxy  to 
the  small  blue  knot  south  of  it.  The  ap- 
parent velocity  of  recession  of  the  whole 
group  is  about  8700  km/sec. 
In   spite  of  the  elevated  brightness  of 

the  night-sky  glow,  due  to  the  present  high 
solar  activity,  the  spectra  so  far  obtained 
of  the  luminous  intergalactic  matter  are 
surprisingly  distinct.  It  can  therefore  be 
anticipated  that  at  low  sky  brightness,  a 
few  years  hence,  it  will  be  possible  to  ob- 
tain significant  information  on  the  spectral 
types  of  many  luminous  intergalactic 


Minkowski  has  continued  the  investiga- 
tion of  objects  that  might  be  associated 
with  radio  sources.  Poor  observing  condi- 
tions have  hampered  the  progress  of  this 
work.  It  is  becoming  increasingly  certain 
that  only  a  small  fraction  of  sources,  even 
of  those  with  reliable  positions,  can  be 
identified  with  optically  observable  objects. 
The  conclusion  is  now  quite  definite  that 
there  are  not  many  galaxies  whose  radio 
emission  is  intermediate  between  the  weak 
emission  shown  by  normal  galaxies  and 
the  very  much — perhaps  1500  times — 
stronger  emission  of  a  few.  At  present  it 
is  not  possible  to  decide  whether  this 
means  that  most  radio  sources  are  exceed- 
ingly strong  emitters,  such  as  the  colliding 
galaxies  Cygnus  A,  or  whether  the  gen- 
erally accepted  assumption  that  the  ma- 
jority of  the  radio  sources  are  distant 
galaxies  is  wrong. 

A  large  amount  of  time  was  spent  by 
Matthews  in  completing  the  observations 
for  a  21-cm  survey  from  Galactic  longi- 
tudes 190°  to  270°.  The  reduction  of  the 
records  is  almost  complete. 

The  well  known  controversy  about  the 
distance  of  the  radio  source  Cassiopeia  A — 
from  the  optical  data  a  distance  of  about 
500  parsecs  had  been  inferred  whereas  the 
radio  observations  demanded  a  distance 
—3000  parsecs — was  finally  settled  when 
Baade  showed  that  the  optical  data  lead 
in  a  straightforward  way  to  the  picture  of 
an  expanding  shell  of  finite  thickness.  The 
outer  surface  of  the  shell  has  a  velocity  of 
7440  km/sec,  the  inner  a  velocity  of  6200 
km/sec.    Combining   these    figures    with 

the  observed  transverse  motions  of  the 
shell  one  obtains  for  the  distance  of  the 
shell  D  =  3.42  ±0.27  kiloparsecs,  in  agree- 
ment with  the  radio  determination.  It  is 
also  certain  now  that  because  of  its  high 
velocity  of  expansion  the  Cassiopeia  A 
nebulosity  must  be  regarded  as  the  remnant 
of  a  type  II  supernova,  the  first  supernova 
of  this  kind  identified  in  our  own  Galaxy. 
From  the  transverse  motions  of  the  con- 
densations in  the  shell  the  date  of  the  out- 
burst can  be  computed.  It  took  place  A.D. 
1697  ±14  years.  Why  the  event  was  not 
noticed  is  thoroughly  explained  by  the  dis- 
tance modulus  of  the  supernova  (m—M— 
12.6)  and  the  large  absorption  in  front 
of  it,  which  amounts  to  at  least  5  magni- 
tudes. It  is  therefore  highly  improbable 
that  the  supernova  reached  the  third  mag- 
nitude at  its  maximum.  An  interesting 
feature  of  the  shell  is  the  deceleration  of 
the  hemisphere  turned  toward  us,  which 
reaches  amounts  up  to  16  per  cent  of  the 
original  velocities.  In  the  direction  toward 
us  the  shell  is  therefore  moving  into  some 
interstellar  gas.  The  gas  must  contain 
patches  of  very  high  densities,  for  inter- 
mingled with  the  moderately  decelerated 
condensations  are  some  which  are  prac- 
tically completely  decelerated  and  which 
are  optically  characterized  by  extreme  red- 
dening. They  are  the  red  bits  of  the  nebu- 
losity that  can  be  photographed  only  in 
red  light.  They  must  have  been  deceler- 
ated extremely  fast,  since  they  have 
reached  distances  from  the  center  of  the 
shell  that  differ  in  no  way  from  those  of 
the    undecelerated    condensations.    More- 



over,  in  the  present  velocity  diagram  of  the 
shell  there  is  not  a  single  condensation  in 
such  a  state  of  fast  deceleration.  The  Cas- 
siopeia A  nebulosity  represents  the  first 
clear-cut  example  of  the  collision  of  a  high- 

velocity  gas  shell  with  the  interstellar  me- 
dium. It  is  probably  a  safe  guess  that  the 
radio  emission  is  a  result  of  this  collision 
involving  very  high  velocities  and  masses 
comparable  to  that  of  the  sun,  if  not  larger. 


A  theoretical  investigation  was  con- 
ducted by  Code  of  the  properties  of  the 
very  luminous  OB  stars  observed  by  Code 
and  Houck.  The  masses  of  these  stars  are 
of  the  order  of  100  O,  and  they  are  char- 
acterized by  electron  scattering  as  the 
main  source  of  stellar  opacity  and  by  the 
dominance  of  radiation  pressure.  A  limit 
theorem  was  derived  for  the  amount  of 
mass  in  the  convective  core.  It  is  found 
that  the  bulk  of  the  mass  is  in  convection 
and  well  mixed.  As  the  star  consumes 
hydrogen  in  the  convective  core  the  core 
shrinks,  the  luminosity  increases,  and,  since 
the  chemical  composition  of  the  outer  ra- 
dioactive envelope  does  not  change,  the 
force  due  to  radiation  increases,  expanding 
the  envelope.  It  is  suggested  that  later-type 
supergiants  are  produced  by  this  mecha- 
nism, and  an  investigation  of  the  stellar  at- 
mosphere of  such  structures  shows  it  to 
agree  well  with  the  observed  properties  of 
luminous  supergiants.  These  stars  are 
found  to  be  unstable  to  radial  pulsations 
with  a  secular  change  due  to  the  expan- 
sion. The  pulsation  may  be  correlated  with 
the  velocity  variations  observed  by  Abt. 

An  investigation  on  the  rate  of  star 
formation    as    a    function    of    time    was 

started  by  Schmidt.  A  comparison  be- 
tween the  distribution  of  gas  perpendicular 
to  the  Galactic  plane  and  that  of  young 
stars  suggests  that  the  rate  of  star  forma- 
tion varies  with  the  square  of  the  gas 
density.  It  is  then  possible  to  compute  the 
rate  of  star  formation  as  a  function  of  time 
by  taking  into  account  the  depletion  of  the 
gas  due  to  star  formation  and  the  ejec- 
tion of  gas  into  interstellar  space  during 
the  transition  from  the  giant  to  the  white- 
dwarf  stage.  The  present  rate  of  star 
formation  is  found  to  be  one-fifth  of  the 
average  rate.  The  initial  luminosity  func- 
tion, which  is  determined  from  the  rate 
of  star  formation  and  the  observed  main- 
sequence  luminosity  function  in  the  field, 
fits  the  observed  distribution  of  luminosi- 
ties in  open  clusters  within  the  uncertainty. 
The  helium  abundance  of  the  sun  and  of 
the  interstellar  gas  at  present  can  be  ex- 
plained if  the  average  efficiency  with 
which  hydrogen  is  converted  into  helium 
and  ejected  at  the  end  of  the  giant  stage 
is  40  per  cent.  The  theory  gives  at  least  a 
partial  explanation  for  the  nearly  constant 
gas  density  observed  in  the  plane  over  a 
large  range  of  distance  from  the  Galactic 


The  coude  spectrograph  of  the  100-inch 
telescope  has  been  undergoing  extensive 
revision  during  the  past  few  years.  A  de- 
vice similar  to  a  record  changer  was  first 
installed  to  make  the  interchange  of  grat- 
ings very  rapid  and  safe;  it  permits  the 
selection  of  the  grating  with  the  optimum 
dispersion  and  blaze  angle  for  each  object 
under  study.  At  the  same  time  the  slit 
was  replaced  with  a  new,  longer  model. 

During  the  current  year  two  new  cam- 
eras of  8-inch  and  16-inch  focal  lengths 
were  installed.  Both  are  of  the  schmidt 
type  with  twice-through  corrector  plate 
and  field  flattener.  The  corrector  plate  of 
the  16-inch  is  of  fused  quartz  to  permit  ob- 
servations in  the  ultraviolet  to  the  limit 
of  atmospheric  transmission.  With  the 
gratings  normally  used  these  new  cameras 
provide  dispersions  of  40  A/mm  and  20 


A/mm  in  the  blue,  in  addition  to  the  dis- 
persions of  10  A/mm,  4.5  A/mm,  and  2.9 
A/mm  already  available  with  the  older 

A  preliminary  model  of  a  photoelectric 
spectrum-scanning  device  was  tried  out 
by  Code  and  Wilson  on  the  coude  spec- 
trograph of  the  100-inch  telescope  this  year. 
It  scans  by  rotating  the  grating,  the  photo- 
multiplier  being  placed  behind  a  slit  lo- 
cated in  the  focal  plane  of  the  9-foot  cam- 
era. With  this  linear  dispersion  of  2.9 
A/mm  a  resolution  of  0.05  A  was  prac- 
tical. An  image  slicer  replacing  the  en- 
trance slit  proved  to  be  highly  desirable, 
substantially  reducing  the  fluctuations  due 
to  seeing;  it  also  increased  the  efficiency 
considerably.  Approximately  20  per  cent 
of  the  flux  in  the  seeing  tremor  disk  was 
transmitted,  representing  an  increase  of 
approximately  a  factor  of  10  over  the  use 
of  a  simple  slit  at  this  resolution  for  av- 
erage seeing.  Seeing  compensation  was 
employed  using  a  monitoring  photocell 
intercepting  a  fraction  of  the  light  passing 
through  the  first  slit.  The  spectral  region 
was  isolated  with  glass  and  interference 
filters,  and  the  electronic  circuitry  was 
similar  to  that  described  by  Hiltner  and 
Code.  On  the  basis  of  a  test  program  ex- 
tending over  20  nights,  changes  were  sug- 
gested, and  a  modified  instrument  is  now 
being  constructed.  In  particular  it  was 
found  that  accuracy  of  scanning  could  not 
be  achieved  by  rotating  the  grating  in  its 
present  bearings,  and  the  scanning  will  be 
done  by  a  small  optical  offset  mechanism 
in  front  of  the  second  slit.  The  spectral 
isolation  for  the  monitoring  cell  will  be 
obtained  by  picking  of?  light  in  the  focal 
plane  of  the  spectrograph.  Finally,  the 
electronic  circuitry  is  being  modified  to 
utilize  the  band-pass  characteristics  of  a 
single  amplifier. 

In  September  1957  a  procedure  was  set 
up  by  Arp  that  enables  all  photoelectric 

observations  to  be  automatically  reduced 
by  an  electronic  computer.  A  program  con- 
sisting of  about  1000  instruction  com- 
mands is  read  into  the  Datatron  Digital 
Computer  (Model  204)  at  the  California 
Institute.  The  computer,  starting  with 
given  values  of  atmospheric  extinction, 
then  transforms  all  observed  deflections 
into  magnitudes  and  colors  outside  the 
earth's  atmosphere.  A  least-squares  solu- 
tion is  made,  using  all  standard  stars  pres- 
ent, to  obtain  the  best  transformation  be- 
tween the  telescope-photometer  system  and 
the  standard-star  system.  The  extinction 
is  then  computed  on  the  requirement  that 
the  residuals  between  observed  and  given 
standard  magnitudes  be  a  minimum.  A 
least-squares  solution  for  the  extinction  is 
performed,  and  all  magnitudes  and  colors 
are  recomputed  using  the  new  extinction 

A  second  program,  prepared  in  April 
1958,  introduced  a  weighting  function  for 
observations  at  low  altitudes  and  is  in- 
tended to  give  a  more  sensitive  computa- 
tion of  the  extinction.  Program  II  is  pres- 
ently being  used.  It  is  able  to  handle  99 
stars  as  a  group,  and  requires  about  10 
minutes  of  actual  computation  time  to 
make  the  solution  for  that  many  stars.  The 
final  magnitudes  and  colors  are  auto- 
matically typed  out  on  permanent  record 
sheets  in  three  columns,  giving:  (a)  stand- 
ard magnitudes,  if  any;  (b)  first  computed 
values;  (c)  second  computed  values.  The 
sheets  are  headed  by  the  given  extinction 
and  first  computed  transformation  coef- 
ficients, and  ended  by  the  computed  ex- 
tinction and  second  set  of  transformation 
coefficients.  In  general  the  electronic  com- 
puter permits  more  efficient  tabular  han- 
dling of  data,  higher  accuracy,  and  a  data 
processing  about  10  times  faster  than  that 
previously  obtainable.  About  1500  stars 
have  been  processed  so  far  in  this  program. 




The  following  programs  have  been  car- 
ried out  by  guest  investigators  during  the 
current  year. 

Dr.  George  O.  Abell  of  the  Department 
of  Astronomy,  University  of  California  at 
Los  Angeles,  has  continued  his  investiga- 
tion of  rich  clusters  of  galaxies.  He  had 
determined  the  bright  end  of  the  lumi- 
nosity function  of  galaxies  in  rich  clusters 
by  means  of  extrafocal  photographic  pho- 
tometry with  the  48-inch  schmidt  tele- 
scope. Complete  sets  of  yellow  and  blue 
plates  are  now  available  for  33  of  the  36 
rich  clusters  selected  for  the  study.  The 
photometric  reductions  are  in  progress. 

In  a  recent  statistical  investigation  of  the 
distribution  of  rich  clusters,  Abell  found 
highly  significant  departures  from  a  ran- 
dom cluster  distribution.  The  results 
strongly  suggest  the  existence  of  second- 
order  galaxian  clusters  with  a  mean  di- 
ameter (for  the  Hubble  constant,  H  =  75 
km/sec  per  106  parsecs)  of  58  X 106  parsecs. 
Eight  individual  groupings  that  appear  to 
be  second-order  clusters  have  a  mean  di- 
ameter of  50  X  106  parsecs,  and  a  mean  pop- 
ulation of  14  rich  clusters.  It  is  estimated 
that  1016  solar  masses  is  an  upper  limit  to 
the  mass  of  a  typical  second-order  cluster. 
The  corresponding  maximum  value  to  be 
expected  for  the  root-mean-square  velocity 
dispersion  is  of  the  order  of  103  km/sec. 

Dr.  Dinsmore  Alter  of  the  Griffith  Ob- 
servatory continued  his  program  for  the 
study  of  various  features  of  the  moon's 
surface.  Eighty  direct  photographs  were 
obtained  at  the  Cassegrain  focus  of  the  60- 
inch  reflector.  About  half  of  them  were 
on  II-O  plates  and  the  others  on  1-N 
plates  using  a  Pyrex  filter  with  a  10  per 
cent  cutoff  near  7400  A. 

Dr.  Jan  Borgman  of  the  Kapteyn  As- 
tronomical Laboratory  at  Groningen, 
Netherlands,  carried  out  an  investigation 
for  the  purpose  of  detecting  faint,  possibly 
intergalactic,  variable  stars  in  a  field  near 
the  Galactic  pole.  A  total  of  60  plates  was 
taken  with  the  48-inch  schmidt  during  the 

months  of  February,  March,  and  April. 
Using  the  60-inch,  UBV  photometry  was 
carried  out  for  a  number  of  stars  in  SA  4 
and  5,  classified  by  Dr.  Elvius,  in  order  to 
investigate  the  possibility  of  using  the  ul- 
traviolet excess  as  a  "population  index"  as 
suggested  by  Dr.  Nancy  Roman. 

An  investigation  of  the  diameters  of 
meteor  trails  was  carried  out  by  Dr.  A.  F. 
Cook  of  the  Harvard  College  Observatory. 
Eighteen  trailed  exposures,  centered  on 
the  radiant  of  the  Geminid  meteors,  were 
obtained  with  the  48-inch  schmidt  camera 
on  the  nights  of  December  11  and  12.  The 
telescope  was  focused  alternately  on  the 
stars  and  on  the  meteors.  In  three  hours 
of  exposure  33  meteors  were  photographed, 
of  which  one  was  sporadic  and  another 
possibly  so.  The  seeing  on  the  second 
night  was  essentially  perfect  and  con- 
tributed nothing  to  the  widths  of  the  trails. 
Three  meteors,  of  which  two  were  in 
focus,  were  bright  enough  to  allow  densi- 
tometry of  their  trails.  By  visual  inspection 
both  in-focus  trails  show  a  width  greater 
than  the  star  trails. 

Dr.  T.  E.  Houck  of  the  Washburn  Ob- 
servatory carried  out,  in  collaboration  with 
Code,  an  investigation  of  the  properties  of 
OB  stars  in  near-by  galaxies.  Dr.  Houck 
made  three-color  photoelectric  measure- 
ments of  these  stars  while  Code  investi- 
gated their  spectra. 

A  search  for  flare  stars  was  made  by 
Dr.  Hugh  M.  Johnson  of  the  University 
of  Iowa.  On  24  plates  taken  in  rapid  suc- 
cession with  the  48-inch  schmidt  camera 
of  the  region  of  SA  57  near  the  north 
Galactic  pole,  of  an  Ophiuchus-Scorpius 
dark-cloud  field,  and  of  an  intermediate 
field  around  SA  89,  were  found  141  sam- 
ple flare-  and  "flicker"-stars.  The  term 
"flicker"  is  introduced  because  of  the 
rather  low-grade  activity  of  the  samples, 
which  were  at  first  blinked  and  then  iris 
photometered.  About  1  high-latitude  field 
star  in  2000,  brighter  than  photographic 


magnitude  20,  detectably  flickers  at  least 
once  in  88  minutes. 

A  study  of  Galactic  symmetric  nebulae 
was  also  made  by  Dr.  Johnson  using  ex- 
isting 48-inch  material. 

Between  July  1,  1957,  and  June  30,  1958, 
Dr.  William  Livingston  of  the  University 
of  California  obtained  a  total  of  144  mag- 
netograms  of  the  solar  disk  with  the 
equipment  at  the  150-foot  tower.  One 
magnetogram  was  made  each  day  that  the 
weather  and  instrumentation  permitted. 
The  sensitivity  was  adjusted  at  4  to  15 
gauss  per  interval  of  the  raster  in  order  to 
obtain  optimum  sensitivity  without  caus- 
ing confusion  in  the  sunspot  zone.  Copies 
of  the  magnetograms  have  been  sent  to 
cooperating  organizations  as  part  of  the 
International  Geophysical  Year  program. 

In  addition  to  the  above,  161  magneto- 
grams  were  obtained  of  the  polar  regions 
with  a  higher  sensitivity  of  about  1  gauss 
per  interval  of  the  trace.  Systematic  errors 
have  been  reduced  as  far  as  possible  in  an 
attempt  to  measure  the  magnitude  and  ex- 
tent of  these  fields.  The  fall  and  winter 
observations  have  indicated  that  the  south 
polar  field  has  a  reversed  direction  to  that 
found  in  the  preceding  sunspot  minimum. 
The  north  polar  field  has  been  more  or 
less  intermediate,  suggesting  the  need  for 
more  observations  before  definite  conclu- 
sions can  be  drawn.  These  findings  agree 
with  the  observations  made  by  H.  D.  Bab- 
cock  at  the  Hale  Solar  Laboratory  in 

From  September  through  December 
1957  photoelectric  observations  were  made 
with  the  20-inch  reflector  on  Palomar 
Mountain  by  Dr.  R.  Lynds  of  the  Univer- 
sity of  California  at  Berkeley.  These  ob- 
servations consisted  of  intercomparisons  of 
the  early  B  giants  in  h  and  x  Persei  with 
the  aim  of  discovering  variability.  Of 
twenty-two  stars  observed,  seven  were 
found  to  be  definitely  variable.  Those  for 
which  the  type  is  fairly  certain  are  Ooster- 
hoff  612,  a  probable  3  Canis  Majoris  star, 
and  Oosterhoff  1586,  an  eclipsing  variable. 

In  addition,  five  of  the  remaining  fifteen 
stars  showed  evidence  of  variability. 

During  April  and  May  1958  eight  early 
B  giants  were  observed  for  variability  with 
the  Palomar  20-inch.  These  stars  are  from 
a  list  of  45  B  stars  north  of  §=-30°, 
brighter  than  V  =  7.0,  and  having  MK 
classifications  similar  to  those  of  the  known 
3  Canis  Majoris  stars.  Six  of  the  stars  ob- 
served show  variability;  it  is  too  early  to 
ascertain  the  type. 

Dr.  John  Mathis  of  the  Michigan  State 
University  studied  the  hydrogen/helium 
abundance  ratio  in  the  diffuse  nebula 
NGC  604  in  the  spiral  Messier  33.  In 
August  he  obtained  spectrograms  of  both 
NGC  604  and  the  Orion  nebula  using  the 
60-inch  telescope.  The  spectrograms  indi- 
cated that  the  helium/hydrogen  ratio  in 
NGC  604  is  0.13  by  number  of  atoms,  al- 
most exactly  that  found  in  the  Orion 
nebula.  NGC  604  showed  an  anomalous 
intensity  ratio  of  H3  and  Hy,  however, 
indicating  that  the  nebula  might  be  self- 
absorbed  in  the  hydrogen  lines.  If  NGC 
604  is  self-absorbed,  the  helium  abundance 
is  less  than  that  quoted  above. 

The  cooperative  program  with  the 
McMath-Hulbert  Observatory  has  contin- 
ued throughout  the  year.  Observations 
with  the  Snow  telescope  were  carried  out 
by  Mr.  Thomas  K.  Jones. 

The  programs  with  the  Snow  telescope 
were  closely  integrated  with  the  programs 
of  the  McGregor  Tower  Telescope  of  the 
McMath-Hulbert  Observatory.  Because  of 
an  incredibly  poor  observing  season  at 
Mount  Wilson  during  1957-1958,  the  Snow 
observations  have  lagged  seriously  behind 
the  Lake  Angelus  observations.  The  Snow 
telescope  was  used  on  only  85  days. 

The  observing  programs  remain,  because 
of  the  poor  season,  much  the  same  as  last 
year:  (1)  Systematic  observation  of  He, 
10830  A,  and  K,  3934  A,  lines,  at  the  limbs 
of  the  sun  and  in  plage  regions.  (2)  Sys- 
tematic observation  of  Ha  on  the  driest 
days  to  eliminate  the  effects  of  water  vapor. 
(3)    Abundance   observations   in   support 



of  the  McGregor  Tower  program.  (4) 
Absorption-coefficient  observations  in  sup- 
port of  the  McGregor  Tower  program.  (5) 
Continuation  of  wavelength  measurements 
in  the  5  u  region.  The  results  from  the 
Snow  telescope  have  been  used  entirely  for 
the  support  of  the  detailed  programs  of 
the  McGregor  Tower. 

During  the  current  year  Dr.  D.  H. 
McNamara  of  Brigham  Young  University 
continued  his  investigation  of  RZ  Scuti. 
He  finds  the  hydrogen  lines  split  into  two 
components,  the  stronger  of  which  origi- 
nates in  a  shell  of  gas  surrounding  the 
system,  whereas  the  weaker  component  is 
the  normal  absorption  line.  He  finds,  also, 
that  the  individual  helium  lines  give  dif- 
ferent rotational  velocities,  the  weakest 
lines  giving  the  largest.  Both  the  hydro- 
gen and  helium  lines  are  consistent  with 
a  hypothesis  of  a  shell  surrounding  the 
star  whose  rotational  velocity  decreases 
with  the  height  above  the  star. 

Dr.  McNamara  has  also  obtained  several 
plates  of  U  Cephei  that  show  incipient 
emission  filling  in  the  Ha  absorption  line 
during  phases  near  0.25  and  0.75  P.  With 
the  60-inch  telescope  he  has  also  obtained 
spectrograms  of  faint  B  stars  (magnitude 
10-12)  in  the  h  and  x  cluster  in  Perseus 
for  the  purpose  of  measuring  the  equiva- 
lent widths  of  the  Hoc  and  H£  absorption 
lines.  Plates  have  also  been  obtained  of 
a  number  of  sharp-line  Bi  and  B2  stars  in 
order  to  test  for  small  radial-velocity  varia- 
tions. For  0  Ophiuchi  the  radial-velocity 
observations  indicate  a  small  period  varia- 
tion in  velocity  of  3h  24m. 

A  revised  system  of  classification  of 
forms  of  galaxies  has  been  devised  by  Dr. 
W.  W.  Morgan  of  Yerkes  Observatory. 
This  system,  based  on  the  relative  im- 
portance of  the  nuclear  concentration  of 
light,  has  a  fairly  close  correlation  with 
the  integrated  spectral  types,  and  permits 
interpretation  in  terms  of  the  stellar  popu- 
lation of  the  galaxies.  It  complements  that 
of  Hubble. 

A  total  of  608  galaxies  was  classified  on 
the  revised  system  from  direct  photographs 

in  the  Hubble-Sandage  collection.  A  dis- 
cussion of  this  material  suggests  that  the 
stellar  populations  differ  systematically  in 
various  nearer  regions  of  the  universe. 
For  example,  the  percentage  of  "young-star 
rich"  galaxies  is  high  in  the  Ursa  Major 
Cloud;  it  is  considerably  lower  in  the 
Virgo  Cloud.  In  the  nuclear  region  of 
the  latter,  there  is  a  concentration  of  eight 
galaxies  brighter  than  the  12th  magnitude 
(Shapley-Ames),  all  members  of  which 
are  in  the  "young-star  deficient"  category. 

It  now  appears  feasible  to  investigate 
stellar  populations  for  all  clusters  of  gal- 
axies whose  members  are  bright  enough  to 
permit  widened  spectra  to  be  obtained  hav- 
ing dispersions  of  the  order  of  300  A/mm. 

Mr.  Luis  Munch  of  the  Tonantzintla 
Observatory  obtained  about  100  additional 
spectrograms  of  the  OB  stars  in  the  asso- 
ciations I,  II,  and  III  Cassiopeiae.  The 
plates  are  being  measured  for  radial  veloci- 
ties, which,  with  the  previously  determined 
spectroscopic  parallaxes,  will  be  used  for 
a  study  of  the  spatial  relationship  of  these 

Dr.  G.  }.  Odgers  of  the  Dominion  Astro- 
physical  Observatory  made  an  extensive 
series  of  observations  of  four  (3  Cephei 
stars.  Forty-six  spectrograms  of  (3  Cephei, 
48  of  12  Lacertae,  106  of  HD  199140,  and 
30  of  HD  202253  were  obtained  in  rapid 
succession  for  studies  of  the  short-period 
changes  in  the  velocity  displacement  and 
intensities  of  the  lines.  The  star  (3  Cephei 
showed  clear  evidence  of  variable  Hot 
emission  throughout  the  radial-velocity 
cycle.  Strong  Hoc  emission  was  observed 
both  in  HD  199140  and  in  12  Lacertae. 
Large  radial-velocity  discontinuities  appear 
in  the  velocity-curves  of  both  these  stars 
of  an  amount  equal  to  the  semiamplitude. 
Evidence  of  velocity  differences  with  re- 
spect to  H  and  He  on  the  one  hand,  and 
Si,  C,  and  N  on  the  other,  especially  near 
the  phase  of  radial-velocity  discontinuity, 
was  obtained  in  HD  199140. 

Dr.  Daniel  Popper  of  the  University  of 
California  at  Los  Angeles  continued  his 
observations  and  analyses  of  the  orbits  of 



eclipsing  binaries.  Attention  was  given  to 
problems  of  bolometric  corrections  and  the 
stellar  temperature  scale,  both  of  which 
are  needed  for  a  complete  analysis  of 
eclipsing  systems.  The  usual  system  of 
bolometric  corrections  and  of  effective  tem- 
peratures was  improved  considerably  by 
the  use  of  photoelectric  colors  such  as 
B  — V  instead  of  spectral  types  as  argu- 
ment. Numerous  observations  of  photo- 
electric colors  and  magnitudes  of  stars 
were  made  with  the  20-inch  reflector  on 
Palomar  Mountain  in  connection  with 
these  problems. 

Infrared  spectra  of  Venus  taken  by  John 
Strong  and  William  M.  Sinton  at  Palo- 
mar in  1953  exhibited  a  strong  absorption 
band  at  11.1  u.  This  band  has  been  tenta- 
tively attributed  to  polymerized  carbon 
suboxide.  The  existence  of  the  gas,  which 
has  the  formula  C3O2,  and  its  polymeriza- 
tion products,  has  also  been  suggested  by 
G.  P.  Kuiper  to  explain  the  cloud-layer 
color  and  polarization  data  of  Venus.  To 
ascertain  the  presence  of  carbon  suboxide 
through  its  absorption  bands  between  3000 
and  3300  A,  Dr.  Sinton  of  the  Lowell  Ob- 
servatory secured  three  plates  containing 
seven  spectra  of  Venus  on  April  5-6,  1958, 
with  the  60-inch  Cassegrain  spectrograph. 
Visual  inspection  of  these  plates  does  not 
reveal  the  presence  of  C3O2.  Its  presence 
or  absence  will  be  made  more  definite 
when  the  plates  have  been  densitometered 
and  compared  with  plates  made  of  the  sky 
spectrum  on  the  same  date. 

An  investigation  of  stars  of  spectral  type 
F2  and  earlier  in  a  north  Galactic  pole 
region  was  made  by  Dr.  Arne  Slettebak 
of  the  Perkins  Observatory  in  an  effort  to 
discuss  the  relative  space  densities  of  Popu- 
lation I  disk  stars  and  Population  II  halo 
stars  as  a  function  of  the  distance  from 
the  Galactic  plane.  This  program  origi- 
nated from  an  earlier  one  carried  out  at 
the  Hamburg  Observatory  in  which  a 
finding  list  of  601  such  stars  to  about  the 
14th  magnitude  was  obtained  by  means  of 
objective-prism  spectra. 

On  19  nights  at  the  60-inch  with  the 
X-spectrograph  and  the  4-inch  camera  and 
on  3  nights  at  the  100-inch  with  the  New- 
tonian spectrograph  and  the  3-inch  cam- 
era, slit  spectrograms  were  obtained  for 
125  Galactic  pole  and  standard  stars.  A 
preliminary  inspection  of  the  spectrograms 
showed  a  number  of  Population  II  hori- 
zontal-branch stars  and  subdwarfs  as  well 
as  a  number  of  apparently  normal  Popu- 
lation I  stars  at  very  great  distances  from 
the  Galactic  plane.  Accurate  spectral  clas- 
sifications and  radial  velocities  for  all 
these  stars  will  eventually  be  obtained. 

Drs.  Otto  Struve  and  Jorge  Sahade  of 
the  Berkeley  Astronomical  Department  of 
the  University  of  California  carried  out 
extensive  observations  of  close  early-type 
spectroscopic  binaries.  The  main  purpose 
of  the  investigation  was  to  trace  the  evolu- 
tion of  the  binaries,  and  to  discover  to 
what  extent  the  interaction  between  the 
component  stars  and  their  loss  of  mass 
influences  their  evolutionary  trends.  Beta 
Lyrae  was  observed  in  the  red  and  the  near 
infrared  to  ascertain  whether  lines  of  a 
late-type  component  appear  in  these  re- 
gions; the  results  were  negative;  the  large 
atlas  of  the  spectra  of  (3  Lyrae  has  been 
completed  during  the  year.  The  emission 
at  Ha,  arising  from  a  common  envelope, 
was  discovered  in  the  system  AO  Cassio- 
peiae.  Rapid  changes  in  the  structure  of 
the  He  I  line  at  A6678  were  discovered 
on  the  spectrograms  of  the  primary  com- 
ponent of  a  Virginis.  The  observations  of 
the  eclipsing  variable  29  UW  Canis  Ma- 
joris  have  disclosed  emission  at  Ha  and 
remarkable  changes  in  intensity  and  pro- 
file in  the  emission  lines  of  N  III  and 
He  II.  Several  observations  were  made  of 
the  eclipsing  variable  Y  Cygni.  Beta  Per- 
sei  (Algol)  was  observed  at  principal 
eclipse  in  the  red  and  near-infrared  regions 
in  an  attempt  to  detect  lines  of  the  sec- 
ondary component  of  the  eclipsing  pair, 
but  with  negative  results. 

Measurements  of  the  radial  velocities  of 
c  Aurigae  on  all  Mount  Wilson  coude 



spectrograms  obtained  since  the  1928-1929 
eclipse  were  completed  by  Dr.  Struve.  An 
extensive  spectrophotometric  study  of 
some  of  the  best  recent  spectrograms  has 
been  made  by  Dr.  Margherita  Hack. 

An  investigation  of  the  luminosity  func- 
tion and  radial  density  distribution  in  the 
Galactic  clusters  M  67,  NGC  7789,  NGC 
188,  and  Praesepe  was  undertaken  by  Dr. 
Sidney  van  den  Bergh  of  the  Perkins  Ob- 
servatory. During  eight  nights  in  Janu- 
ary and  February,  170  plates  were  obtained 
with  the  48-inch  schmidt.  A  series  of 
plates  of  each  cluster  were  taken,  twelve 
in  the  blue  and  one  in  the  red  plus  two 
transfer  exposures  of  Selected  Areas.  The 
blue  exposures  ranged  from  4  seconds  to 
10  minutes. 

A  magnitude  sequence  was  determined 
in  each  cluster  with  the  aid  of  an  Eichner 
photometer.  It  was  then  used  to  find  the 
limiting  magnitude  of  each  plate.  The 
luminosity  function  and  the  radial  density 
distribution  were  next  determined  by 
counting  the  number  of  stars  in  concentric 
rings  about  the  nucleus. 

The  luminosity  function  of  M  67  was 
found  to  be  deficient  in  faint  stars,  and  the 
faint  stars  are  located  predominantly  in 
the  outer  regions  of  the  cluster.  NGC 
7789  is  probably  the  richest  Galactic  cluster 
known.  The  faint  stars  are  again  located 
predominantly  in  the  outer  part  of  the 
cluster.  The  main-sequence  population  of 
the  cluster  is  deficient  in  faint  stars  in 
comparison  with  the  Salpeter  luminosity 
function,  although  not  to  the  same  extent 
as  M  67.  The  Praesepe  was  found  to  con- 
tain some  hundreds  of  stars  with  absolute 
visual    magnitudes    between     +7.0    and 

+  11.5.  The  cluster  radius  for  the  faint 
stars  is  larger  than  for  the  bright  ones. 

From  August  21  to  August  31,  inclu- 
sive, Dr.  Peter  van  de  Kamp  of  the  Sproul 
Observatory  of  Swarthmore  College  made 
a  visual  and  photographic  search  with  the 
100-inch  reflector  for  close  faint  compan- 
ions of  a  number  of  near-by  stars.  The  ob- 
serving conditions  were  adverse.  A  num- 
ber of  persistent  elongations  were  noted 
visually,  using  magnifying  power  up  to 
1000  X  .  None  were  considered  sufficiently 
reliable  to  be  regarded  as  positive  evidence 
for  companion  stars.  For  future  explora- 
tion twenty  plates  were  taken  with  ex- 
posures ranging  from  1  second  to  15 

Dr.  Merle  Walker  of  the  Lick  Observa- 
tory made  photoelectric  observations  of 
AE  Aquarii  in  cooperation  with  Code  on 
two  nights  during  August  1957.  Dr. 
Walker  observed  the  star  at  the  60-inch, 
obtaining  a  continuous  record  of  the  light- 
variations  in  the  ultraviolet.  Simultane- 
ously observations  in  the  Balmer  con- 
tinuum, in  H  emission  lines,  and  in  the 
continuous  spectrum  between  lines  were 
obtained  by  Code  with  the  photoelectric 
scanner  at  the  100-inch.  The  reduction  of 
the  60-inch  observations  is  complete,  and 
correlations  are  being  made  with  the  con- 
tinuous UV  traces  to  determine  how  the 
various  features  of  the  spectrum  change 
during  the  rapid  light-variations  of  the 
star.  This  program  is  a  continuation  of 
the  program  of  simultaneous  photoelec- 
tric and  spectroscopic  observation  of  the 
star  reported  earlier  by  Deutsch  and  Dr. 


Dr.  Walter  Baade  retired  from  the  staff 
of  the  Observatories  on  June  30,  1958.  On 
coming  to  the  Mount  Wilson  Observatory 
in  October  1931,  Baade  attacked  a  series 
of  photometric  problems.  They  included 
the  establishment  of  photometric  stand- 
ards in  Selected  Areas,  investigations  of 

the  light-curves  of  supernovae,  and  studies 
of  the  magnitudes  of  variables  in  globular 
clusters  and  in  galaxies  and  the  use  of 
these  magnitudes  for  distance  determina- 

During  World  War  II  Baade  took  ad- 
vantage of  the  darker  skies  due  to   the 



black-out  of  the  Los  Angeles  area  and  the 
development  of  faster  red-sensitive  plates 
to  make  a  critical  investigation  in  the  red 
of  M  32,  NGC  205,  NGC  147,  NGC  185, 
and  the  nucleus  of  the  Andromeda  galaxy. 
His  plates  achieved  resolution  of  these 
objects  for  the  first  time  and  showed  that 
the  brightest  stars  in  them  are  the  red 
giants.  This  finding  led  Baade  to  intro- 
duce the  concept  of  population  types. 
Population  I,  whose  most  prominent  fea- 
tures are  the  blue  giant  stars,  is  found  in 
the  neighborhood  of  our  sun  and  in  the 
spiral  arms  of  the  Andromeda  and  other 
spiral  galaxies;  Population  II,  which  is 
characterized  by  the  red  giants,  is  observed 
in  the  elliptical  galaxies  and  the  nuclei  of 
the  spirals  and  in  globular  clusters.  De- 
tailed studies  were  then  made  of  the  color- 
magnitude  relations  in  representative  sam- 
ples of  the  two  populations  by  Baade  and 
his  collaborators.  The  results  of  these  in- 
vestigations provide  the  observational  basis 
for  present  theories  of  stellar  evolution 
which  suggest  that  the  difference  between 
the  population  types  is  primarily  one  of 
age.  ^ 

With  the  completion  of  the  200-inch 
Hale  telescope  Baade  continued  his  studies 
of  the  stellar  content  of  the  Andromeda 
galaxy  and  the  other  members  of  the  local 
group.  Special  attention  was  given  to 
cepheid  variables  and  other  stellar  types 
that  might  be  used  as  distance  indicators 
for  fixing  the  distances  of  these  objects. 
These  studies  represent  the  first  step  in  the 
precise  determination  of  the  distances  of  all 
objects  outside  the  Milky  Way.  Baade's 
observations  provided  much  of  the  evi- 
dence for  the  revision  of  the  absolute  mag- 
nitudes of  the  cepheid  variables  and  the 
resultant  increase  in  the  distance  scale  of 
all  objects  outside  our  own  Galaxy  by  a 
factor  of  nearly  3. 

Baade  also  collaborated  with  Minkowski 
in  the  identification  of  many  of  the  radio 
sources  with  optically  observed  objects  and 
in  the  physical  interpretation  of  the  nature 
of  these  sources. 

Dr.  Baade  served  on  the  Observatory 

Committee  from  1953  until  the  time  of  his 

Dr.  }.  Beverly  Oke  was  appointed  to  the 
staff  at  the  end  of  the  report  year. 

Research  Division 

Staff  Members 

Hal  ton  C.  Arp 

Walter  Baade  1 

Horace  W.  Babcock,  Assistant  Director 

William  A.  Baum 

Ira  S.  Bowen,  Director 

Arthur  D.  Code  2 

Armin  J.  Deutsch 

Jesse  L.  Greenstein 

Fred  Hoyle 

Rudolph  L.  Minkowski 

Guido  Munch 

}.  Beverly  Oke 

Donald  E.  Osterbrock  3 

Robert  S.  Richardson  3 

Allan  R.  Sandage 

Olin  C.  Wilson 

Fritz  Zwicky 

Research  Associates 

Paul  W.  Merrill 

Jan  H.  Oort 

Richard  v.  d.  R.  Woolley 

Carnegie  Research  Fellows 

Robert  F.  Howard 
Thomas  A.  Matthews 
Maarten  Schmidt 

Research  Fellows 

David  R.  Dewhirst 
H.  Lawrence  Heifer 
Leona  Marshall 
Neal  Tanner 
George  Wallerstein 
Volker  Weidemann 

Research  Assistants 

Sylvia  Burd 
Mary  F.  Coffeen 
Thomas  A.  Cragg 
Dorothy  S.  Deutsch 
Edith  Flather 
Emil  R.  Herzog 
Joseph  O.  Hickox 

1  Retired  June  30,  1958. 

2  Resigned  August  31,  1958. 

3  Resigned  June  30,  1958. 



A.  Louise  Lowen 
Mildred  Matthews 
Cynthia  Stephens 
Henrietta  H.  Swope 
Carol  Tifft 3 

Student  Observers 

Walter  K.  Bonsack 
Daniel  E.  Harris 
William  G.  Melbourne 
Robert  G.  Norton 
William  G.  TifTt 
Robert  L.  Wildey 

Editor  and  Librarian 
Alexander  Pogo 


William  C.  Miller 

Instrument  Design  and  Construction 

Lawrence  E.  Blakee,  Electronic  Technician 

Floyd  E.  Day,  Optician 

Kenneth  E.  DeHufl,  Machinist 

Robert  D.  Georgen,  Machinist 

Don  O.  Hendrix,  Superintendent,  Optical 

Melvin  W.  Johnson,  Optician 
Bruce  Rule,  Project  Engineer 
Oscar  Swanson,  Instrument  Maker 
Russell   R.    Van   Devender,    Jr.,   Designer 

and  Superintendent,  Instrument  Shop 

Maintenance  and  Operation 

Mount  Wilson  Observatory  and  Offices 

Audrey  A.  Acrea,  Stewardess 
Paul  F.  Barnhart,  Truck  Driver 

Ashel  N.   Beebe,   Superintendent  of  Con- 
Wilma  J.  Berkebile,  Secretary 
Ernest  V.  Cherry,1  Janitor 
Hugh  T.  Couch,  Carpenter 
Stewart  F.  Frederick,  Janitor 
Eugene  L.  Hancock,  Night  Assistant 
Mark  D.  Henderson,  Gardener 
Anne  McConnell,  Administrative  Assistant 
Leah  M.  Mutschler,  Stenographer  and  Tele- 
phone Operator 
Bula  H.  Nation,  Stewardess 
Alfred  H.  Olmstead,  Night  Assistant 
Arnold  T.  RatzlafT,  Night  Assistant 
Clyde  Sanger,1  Gardener 
John  E.  Shirey,  Janitor  and  Relief  Engineer 
Benjamin  B.  Traxler,  Superintendent 

Palomar  Observatory  and  Robinson  Labora- 

Fred  Anderson,  Machinist 

Dorothea  Davis,  Secretary 

Eleanor  G.  Ellison,  Secretary  and  Librarian 

Ferd  Feryan,  Mechanic  4 

Arlis  R.  Grant,  Stewardess 

Leslie  S.  Grant,  Relief  Night  Assistant  and 

Byron  Hill,  Superintendent 

Charles  E.  Kearns,  Night  Assistant 

Harley  C.  Marshall,  Office  Manager 

George  W.  Pettit,  Janitor 

Robert  E.  Sears,  Night  Assistant 

William  C.  Van  Hook,  Electrician  and  As- 
sistant Superintendent 

Gus  Weber,  Assistant  Mechanic 

4  Died  June  11,  1958. 


Abell,  George  O.  The  distribution  of  rich  clus- 
ters of  galaxies.  Astrophys.  ].,  Suppl.  Ser., 
3,  211-288  (1958). 

Arp,  Halton  C.  Cepheid  variables  in  the  Small 
Magellanic  Cloud.  Astron.  ].,  63,  45  (1958). 

Baade,  Walter.  Large-scale  structure  of  spiral 
nebulae.  In  Intern.  Astron.  Union  Sympo- 
sium 5,  ed.  by  N.  G.  Roman,  pp.  1-3,  Cam- 
bridge, University  Press,  1958. 

Babcock,  Harold  D.  Arthur  Scott  King,  1876- 
1957.  Pubs.  Astron.  Soc.  Pacific,  69,  333-335 

Babcock,  Harold  D.,  and  Wm.  C.  Livingston. 
Changes  in  the  sun's  polar  magnetic  field. 
(Abstract.)    Science,  127,  1058  (1958). 

Babcock,  Horace  W.  The  sun's  general  magnetic 
field.  Nature,  178,  533  (1956). 

Babcock,  Horace  W.  Identification  of  solar  "M" 
regions  with  unipolar  magnetic  regions. 
Astrophys.  /.,  126,  224  (1957). 

Babcock,  Horace  W.  A  catalog  of  magnetic  stars. 
Astrophys.  J.,  Suppl.  Ser.,  3,  141-210  (1958). 

Babcock,  Horace  W.  Stars  as  sources  of  cosmic 
rays.  Phys.  Rev.,  109,  2210-2211  (1958). 

Baum,  William  A.  Review  of  Scientific  Uses  of 
Earth  Satellites,  ed.  by  J.  A.  Van  Allen,  Pubs. 
Astron.  Soc.  Pacific,  69,  368-369  (1957). 

Baum,  William  A.  Rapport  de  la  Sous-commis- 
sion des  convertisseurs  d'images.  In  Draft 
Reports,  Moscow  meeting,  August  1958,  In- 
tern. Astron.  Union,  pp.  60-69,  Cambridge, 
University  Press,  1958. 

Baum,  William  A.,  W.  K.  Ford,  Jr.,  and  J.  S. 
Hall.  Recent  astronomical  tests  of  thin-film 


image  converters.  (Abstract.)  Astron.  ].,  63, 
47-48  (1958). 

Bonsack,  Walter  K.  Radial-velocity  variations  in 
four  short-period  variable  stars.  Astrophys. 
].,  126,  291-301  (1957). 

Bonsack,  Walter  K.  Wavelength  variation  in  the 
spectrum  of  56  Arietis.  Pubs.  Astron.  Soc. 
Pacific,  70,  90-97  (1958). 

Bowen,  Ira  S.  Instrumentation  at  the  Mount 
Wilson  and  Palomar  Observatories.  Pubs. 
Astron.  Soc.  Pacific,  69,  377-384  (1957). 

Bowen,  Ira  S.  The  universe  from  Palomar. 
Griffith  Observer,  22,  69-78  (1958). 

Bowen,  Ira  S.  Astronomy  in  a  changing  world. 
In  Frontiers  in  Science,  ed.  by  E.  Hutchings, 
Jr.,  pp.  289-294,  New  York,  Basic  Books, 

Burbidge,  E.  Margaret,  and  G.  R.  Burbidge. 
Chemical  composition  of  the  Ba  II  star  HD 
46407  and  its  bearing  on  element  synthesis 
in  stars.  Astrophys.  J.,  126,  357-385  (1957). 

Burbidge,  E.  Margaret,  and  G.  R.  Burbidge. 
Cosmical  alchemy.  In  Frontiers  in  Science, 
ed.  by  E.  Hutchings,  Jr.,  pp.  180-187,  New 
York,  Basic  Books,  1958. 

Burbidge,  E.  Margaret,  G.  R.  Burbidge,  W.  A. 
Fowler,  and  F.  Hoyle.  Synthesis  of  the  ele- 
ments in  stars.  Revs.  Modern  Phys.,  29,  547- 
650  (1957). 

Burbidge,  E.  Margaret,  and  A.  Sandage.  Prop- 
erties of  two  intergalactic  globular  clusters. 
Astrophys.  J.,  127,  527-538  (1958). 

Burbidge,  Geoffrey  R.  Acceleration  of  cosmic- 
ray  particles  among  extragalactic  nebulae. 
Phys.  Rev.,  107,  269-271  (1957). 

Burbidge,  Geoffrey  R.  Particle  energies  and  mag- 
netic energy  in  the  Crab  Nebula.  Astrophys. 
J.,  127,  48-53  (1958). 

Burbidge,  Geoffrey  R.,  and  F.  Hoyle.  Anti-mat- 
ter. Set.  American,  198  (no.  4),  34-39 

Burbidge,  Geoffrey  R.  See  also  Burbidge,  E. 

Code,  A.  D.,  and  T.  E.  Houck,  A  superluminous 
B-type  star  in  the  Large  Magellanic  Cloud, 
and  its  galactic  counterpart.  Pubs.  Astron. 
Soc.  Pacific,  70,  261-266  (1958). 

Cragg,  Thomas.  Solar  activity  in  1957.  Pubs. 
Astron.  Soc.  Pacific,  70,  299-302  (1958). 

Cragg,  Thomas.  See  also  Nicholson,  Seth  B. 

Deutsch,  Armin  J.  A  method  for  mapping  the 
surfaces  of  some  magnetic  stars.  (Abstract.) 
Astron.  ].,  62,  139  (1957). 

Deutsch,  Armin  J.  Recent  observations  of  the 
companion  of  Mira.  (Abstract.)  Astron.  J., 
63,49  (1958). 

Ford,  W.  K.,  Jr.  See  Baum,  William  A. 

Fowler,  William  A.   See  Burbidge,  E.  Margaret. 

Greenstein,  Jesse  L.  Non-thermal  emission  in 
the  cometary   nebula  NGC   2261.    In   The 

Non-stable  Stars,  pp.  86-87,  Erevan,  Acad. 
Sci.  Armenian  S.  S.  R.,  1957. 

Greenstein,  Jesse  L.  Possible  energy  sources  of 
T  Tauri  type  stars.  In  The  Non-stable  Stars, 
pp.  104-112,  Erevan,  Acad.  Sci.  Armenian 
S.  S.  R.,  1957. 

Greenstein,  Jesse  L.  Theoretical  problems  of  dis- 
crete radio  sources.  In  Radio  Astronomy: 
Intern.  Astron.  Union  Symposium  4,  ed.  by 
H.  C.  van  de  Hulst,  pp.  179-191,  Cambridge, 
University  Press,  1957. 

Greenstein,  Jesse  L.  Studies  of  the  white  dwarfs. 
III.  The  recurrent  nova  WZ  Sge  as  a  prob- 
able white  dwarf.  Astrophys.  J.,  126,  23-27 

Greenstein,  Jesse  L.  The  spectra  of  the  white 
dwarfs.  In  Handbuch  der  Physi\,  ed.  by 
S.  Fliigge,  50,  pp.  161-186,  Berlin-Gottingen- 
Heidelberg,  Springer- Verlag,  1958. 

Greenstein,  Jesse  L.  Draft  report  of  the  President 
of  the  IAU  Commission  29,  Stellar  Spectra. 
In  Draft  Reports,  Moscow  meeting,  August 
1958,  Intern.  Astron.  Union,  pp.  247-263, 
Cambridge,  University  Press,  1958. 

Greenstein,  Jesse  L.,  and  M.  S.  Matthews.  Studies 
of  the  white  dwarfs.  I.  Broad  features  in 
white  dwarf  spectra.  Astrophys.  J.,  126,  14- 
18  (1957). 

Greenstein,  Jesse  L.,  and  W.  J.  Luyten.  Studies 
of  the  white  dwarfs.  II.  Some  new  white 
dwarfs.  Astrophys.  ].,  126,  19-22  (1957). 

Greenstein,  Jesse  L.,  M.  Hack,  and  O.  Struve. 
The  spectroscopic  binary  BD  +  74°493.  As- 
trophys. ].,  126,  281-290  (1957). 

Greenstein,  Jesse  L.,  R.  F.  Sanford,  and  F. 
Zwicky.  On  the  absolute  magnitude  of  Nova 
Puppis  1942.  Pubs.  Astron.  Soc.  Pacific,  69, 
352-353  (1957). 

Greenstein,  Jesse  L.,  and  P.  C.  Keenan.  Abun- 
dances of  metals,  CN,  and  CH  in  giant  stars. 
Astrophys.  J.,  127,  172-190  (1958). 

Greenstein,  Jesse  L.,  and  G.  Wallerstein.  The 
helium-rich  star,  Sigma  Orionis  E.  Astro- 
phys. J.,  127,  237-240  (1958). 

Greenstein,  Jesse  L.  See  also  Merrill,  Paul  W.; 
Swings,  Pol;  Tifft,  William  G. 

Hack,  Margherita.    See  Greenstein,  Jesse  L. 

Hall,  J.  S.  See  Baum,  William  A. 

Herzog,  E.,  P.  Wild,  and  F.  Zwicky.  A  new 
catalogue  of  bright  galaxies  and  clusters  of 
galaxies.  Pubs.  Astron.  Soc.  Pacific,  69,  409- 
420  (1957). 

Houck,  T.  E.  See  Code,  A.  D. 

Hoyle,  Fred.  Forecasting  the  future.  In  Fron- 
tiers in  Science,  ed.  by  E.  Hutchings,  Jr., 
pp.  117-121,  New  York,  Basic  Books,  1958. 

Hoyle,  Fred.  The  place  of  technology  in  civili- 
zation. In  Frontiers  in  Science,  ed.  by  E. 
Hutchings,  Jr.,  pp.  268-277,  New  York,  Basic 
Books,  1958. 



Hoyle,  Fred.  See  also  Burbidge,  E.  Margaret; 
Burbidge,  Geoffrey  R. 

Joy,  Alfred  H.  Flare  spectra  in  dwarf  stars.  In 
Non-stable  stars:  Intern.  Astron.  Union  Sym- 
posium 3,  ed.  by  G.  H.  Herbig,  pp.  31-34, 
Cambridge,  University  Press,  1957. 

Joy,  Alfred  H.  SS  Cygni  type  variables  as  spec- 
troscopic binaries.  In  The  Non-stable  Stars, 
pp.  137-140,  Erevan,  Acad.  Sci.  Armenian 
S.  S.  R.,  1957. 

Joy,  Alfred  H.  Henry  Norris  Russell.  Griffith 
Observer,  21,  86-89   (1957). 

Joy,  Alfred  H.  Walter  Sydney  Adams.  Bio- 
graphical Memoirs,  Natl.  Acad.  Sci.,  31,  1-31 

Keenan,  Philip  C.   See  Greenstein,  Jesse  L. 

Livingston,  William  C.   See  Babcock,  Harold  D. 

Luyten,  Willem  J.   See  Greenstein,  Jesse  L. 

Matthews,  Mildred  S.  See  Greenstein,  Jesse  L. 

Merrill,  Paul  W.  Walter  S.  Adams,  observer  of 
sun  and  stars.   Science,  124,  67  (1956). 

Merrill,  Paul  W.  Walter  Sydney  Adams  (1876- 
1956).  Year  Book,  Am.  Phil.  Soc,  103-105 
(1956).  See  also  Monthly  Notices  Roy. 
Astron.  Soc,  117,  243-244  (1957). 

Merrill,  Paul  W.  Symbiosis  in  astronomy.  Mem. 
soc.  roy.  sci.  Liege,  20,  436-448  (1958);  in 
Etoiles  a  raies  d'emission,  Les  congres  et 
colloques  de  PUniversite  de  Liege,  7 ,  436- 
448  (1958). 

Merrill,  Paul  W.  From  atoms  to  galaxies. 
Astron.  Soc.  Pacific  Leaflet  349,  8  pp.  (1958). 

Merrill,  Paul  W.,  and  J.  L.  Greenstein.  Double 
absorption  lines  in  the  spectrum  of  R  An- 
dromedae.  Pubs.  Astron.  Soc.  Pacific,  70, 
98-101  (1958). 

Minkowski,  Rudolph.  Optical  investigations  of 
radio  sources.  Introductory  lecture,  1955  Jod- 
rell  Bank  symposium.  In  Radio  Astronomy: 
Intern.  Astron.  Union  Symposium  4,  ed.  by 
H.  C.  van  de  Hulst,  pp.  107-122,  Cambridge, 
University  Press,  1957. 

Minkowski,  Rudolph.  The  problem  of  the  identi- 
fication of  extragalactic  radio  sources.  Pubs. 
Astron.  Soc.  Pacific,  70,  143-151   (1958). 

Munch,  Guido.  Study  of  the  hydrogen  and 
helium  II  lines  in  a  high-temperature  sub- 
dwarf.    Astrophys.  J.,  127,  642-657  (1958). 

Munch,  Guido.  Review  of  Non-stable  Stars:  In- 
tern. Astron.  Union  Symposium  3,  ed.  by 
G.  H.  Herbig.  Pubs.  Astron.  Soc.  Pacific, 
70,  226-227  (1958). 

Newburn,  Ray  L.,  Jr.  The  RR  Lyrae  stars  in 
NGC  4147.   Astron.  J.,  62,  197-203  (1957). 

Nicholson,  Seth  B.  The  solar  cycle.  Astron.  Soc. 
Pacific  Leaflet  348,  8  pp.  (1958). 

Nicholson,  Seth  B.,  Th.  Cragg,  and  others.  Sum- 
mary of  Mount  Wilson  magnetic  observa- 
tions of  sunspots  for  March,  1957,  to  Febru- 
ary,   1958.    Pubs.   Astron.    Soc.   Pacific,   69, 

358-362,  469-473,  578-583  (1957);  70,  114- 
119,  212-217,  319-324  (1958). 

Osterbrock,  Donald  E.  Parallel  filamentary  struc- 
ture in  diffuse  nebulae.  (Abstract.)  Astron. 
/.,  62,  144-145  (1957). 

Osterbrock,  Donald  E.  Electron  densities  in  fila- 
mentary nebulae.  Pubs.  Astron.  Soc.  Pacific, 
70,  180-184  (1958). 

Osterbrock,  Donald  E.  Review  of  The  Restless 
Universe,  by  Max  Born,  transl.  by  W.  M. 
Deans.  Pubs.  Astron.  Soc.  Pacific,  69,  372 

Richardson,  Robert  S.  Spectroscopic  observations 
of  Venus  for  rotation  made  at  Mount  Wilson 
in  1956.  Pubs.  Astron.  Soc.  Pacific,  70,  251- 
260  (1958). 

Sahade,  Jorge,  and  G.  Wallerstein.  The  spectrum 
of  Algol  in  the  near  infrared  at  principal 
eclipse.  Pubs.  Astron.  Soc.  Pacific,  70,  207- 
208  (1958). 

Sandage,  Allan.  Observational  approach  to  evo- 
lution. III.  Semiempirical  evolution  tracks 
for  M  67  and  M  3.  Astrophys.  ].,  126,  326- 
340  (1957). 

Sandage,  Allan.  The  birth  and  death  of  a  star. 
Am.  J.  Phys.,  25,  525-531  (1957);  in  Fron- 
tiers in  Science,  ed.  by  E.  Hutchings,  Jr.,  pp. 
188-189,  New  York,  Basic  Books,  1958. 

Sandage,  Allan.  Current  problems  in  the  extra- 
galactic  distance  scale.  Astrophys.  ].,  127 , 
513-526  (1958). 

Sandage,  Allan.  See  also  Burbidge,  E.  Margaret. 

Sanford,  Roscoe  F.   See  Greenstein,  Jesse  L. 

Smith,  Harlan  J.  See  TifTt,  William  G. 

Struve,  Otto.    See  Greenstein,  Jesse  L. 

Swings,  Pol,  and  J.  L.  Greenstein.  Presence  of 
forbidden  lines  of  oxygen  in  cometary  spec- 
tra. Compt.  rend.  acad.  sci.,  Paris,  246,  511- 
513  (1958). 

Tifft,  William  G.,  and  J.  L.  Greenstein.  MWC 
603,  a  high-velocity  symbiotic  star.  Astro- 
phys. ].,  127,  160-171  (1958). 

TifTt,  William  G.,  and  H.  J.  Smith.  T  Sextantis, 
an  RR  Lyrae  star  of  type  "c."  (Abstract.) 
Astron.  J.,  63,  54  (1958). 

Walker,  Merle  F.  The  extremely  rapid  light- 
variations  of  old  novae  and  related  objects. 
In  Non-stable  Stars:  Intern.  Astron.  Union 
Symposium  3,  ed.  by  G.  H.  Herbig,  pp.  46- 
56,  Cambridge,  University  Press,  1957. 

Wallerstein,  George.  Note  on  the  behavior  of 
the  RV  Tauri-type  star  No.  11  in  Messier  2. 
Astron.  ].,  62,  168  (1957). 

Wallerstein,  George.  Movement  observations  on 
the  Greenland  Ice  Sheet.  /.  Glaciol,  3,  207 

Wallerstein,  George.  Radial  velocities  and  spec- 
tral characteristics  of  the  Population  II 
cepheids  M  5  No.  42,  M  5  No.  84,  and  TW 


Capricorni.  Astrophys.  ].,  127,  583-590 

Wallerstein,  George.  See  also  Greenstein,  Jesse 
L.;  Sahade,  Jorge. 

Wild,  P.  See  Herzog,  E. 

Wilson,  Albert  G.  Astronomy  and  eschatology. 
In  Frontiers  in  Science,  ed.  by  E.  Hutchings, 
Jr.,  pp.  205-213,  New  York,  Basic  Books, 

Wilson,  Olin  C.  He  emission  in  late-type  stars. 
Astrophys.  ].,  126,  46-50  (1957). 

Wilson,  Olin  C.  Limitations  on  the  physical 
theories  of  H  and  K  emission  lines.  Astro- 
phys. J. ,126,525-528  (1957). 

Wilson,  Olin  C.  Review  of  Astrophysics  I:  Stel- 
lar Surfaces — Binaries,  vol.  50  of  the  Hand- 
buch  der  Physi\,  ed.  by  S.  Fliigge.  Pubs. 
Astron.  Soc.  Pacific,  70,  332-334  (1958). 

Zwicky,  Fritz.  Non-uniformities  in  the  apparent 
distribution  of  clusters  of  galaxies.  Pubs. 
Astron.  Soc.  Pacific,  69,  518-529  (1957). 

Zwicky,  Fritz.  Propellants  for  tomorrow's  rock- 
ets. Astronautics,  Aug.  1957,  pp.  45-49,  95- 

Zwicky,  Fritz.  News  on  gravitation  and  nuclear 
goblins.   Griffith  Observer,  22,  49-63  (1958). 

Zwicky,  Fritz.  The  first  shots  into  interplane- 
tary space.  Eng.  and  Sci.,  21  (no.  4),  20-23 

Zwicky,  Fritz.  The  morphological  method  and 
vector  associations.  In  Les  methodes  de  la 
science  moderne,  pp.  311-317  (French  transl., 
pp.  319-326),  Paris,  Editions  Science  et  In- 
dustrie, 1958. 

Zwicky,  Fritz.  See  also  Greenstein,  Jesse  L.; 
Herzog,  E. 


Cooperative  Project  of  Mount  Wilson  and  Palomar  Observatories,  Department  of  Ter- 
restrial Magnetism,  National  Bureau  of  Standards,  and  United  States  Naval  Observatory 

W.  A.  BAUM,  Mount  Wilson  and  Palomar  Observatories 

JOHN  S.  HALL,*  United  States  Naval  Observatory 

L.  L.  MARTON,  National  Bureau  of  Standards 

M.  A.  TUVE  {Chairman) 

Department  of  Terrestrial  Magnetism 

*  After  September  1,  1958,  Director,  Lowell  Observatory,  Flagstaff,  Arizona. 

Carnegie  Institution   of  Washington  Year  BooJ{  57,  1957-1958 

The  Committee  on  Image  Tubes  for 
Telescopes  has  continued  during  the  past 
year  to  encourage  the  development  of  pho- 
toelectric image-intensifying  devices  for 
use  in  astronomy.  Working  toward  this 
goal  the  Committee  has  fostered  research 
in  promising  schemes  for  the  production  of 
image  intensification  and  tried  to  evaluate 
new  image-tube  hardware  produced  for 
the  Committee  by  various  industrial  re- 
search and  development  groups.  Atten- 
tion has  been  directed  both  to  methods  that 
record  the  photoelectrons  themselves  and 
to  methods  that  use  photoelectrons  to  pro- 
duce an  optical  output. 

During  the  report  year,  three  barrier- 
film  image  tubes  were  tested  with  some 
success  on  the  Naval  Observatory's  40-inch 
telescope  at  Flagstaff,  Arizona.  Electrons 
ejected  from  the  photocathode  of  a  tube 
of  this  type  are  directly  recorded  when 
they  impinge  upon  a  very  fine-grained 
emulsion;  the  cathode  is  protected  from 
contamination  by  a  barrier  film  approxi- 
mately 1000  A  thick  through  which  the 
photoelectrons  (after  being  accelerated 
through  10  kv)  can  pass.  Such  a  system 
derives  its  advantage  over  direct  photog- 
raphy from  the  photocathode's  higher 
quantum  efficiency  implemented  by  the 
high  storage  capacity  of  fine-grained  nu- 
clear-type emulsions.  This  principle  is  dis- 
cussed in  earlier  reports  (e.g.,  Year  Book 
54,  1954-1955,  pp.  39  ft".),  and  the  previous 
work  on  thin-film  image  tubes  is  described 
in  last  year's  report  (Year  Book  56,  p.  79). 

The  early  tests  of  this  type  of  tube  were 
complicated  by  the  task  of  removing  the 
glass  cap  (which  protects  the  thin  barrier 
film  from  rupturing  under  atmospheric 
pressure)  by  cracking  the  glass  with  an 
electrically  heated  wire  after  the  tube  is 
mounted  in  a  vacuum  chamber.  We  have 
circumvented  this  difficulty  by  having  the 
tube  equipped  with  a  metal  cap  and  by 
using  a  mechanical  "can  opener"  in  the 
vacuum  chamber  in  which  the  tube  is 
mounted.    This  system  has  worked  well 

on  the  one  dummy  tube  and  the  three  live 
tubes  tested  this  year. 

In  August  1957,  a  thin-film  tube  having 
a  very  poor  cathode  (a  photosensitive  area 
only  4  mm  wide  and  10  mm  long  and  a 
sensitivity  of  less  than  1  microampere  per 
lumen)  was  used  to  record  the  images  of 
some  bright  stars  and  some  laboratory 
test  patterns.  Estimates  were  made  of  the 
thermionic  background  introduced  by  the 
photocathode  and  of  the  resolution  capa- 
bilities of  the  tube.  The  tube  was  operated 
for  three  nights  with  little  loss  in  sensi- 
tivity before  the  vacuum  pumps  were 
turned  ofT  and  the  photocathode  allowed 
to  decay. 

In  November  a  thin-film  image  tube 
having  a  photosensitivity  of  about  5  micro- 
amperes per  lumen  was  tested.  Provision 
was  made  for  partial  cooling  of  the  cath- 
ode to  reduce  the  spurious  background  due 
to  thermionic  emission  and  for  placing  the 
fine-grained  emulsion  about  40  microns 
from  the  barrier  film  to  reduce  the  loss  of 
resolution  due  to  the  scattering  of  the  elec- 
trons by  the  thin  barrier  film.  The  com- 
ponents of  double  stars  5  seconds  of  arc 
apart  could  be  resolved  when  this  was 
done.  A  number  of  30-second  exposures 
were  made  on  an  open  cluster,  M  36,  and 
compared  with  direct  photographs  of  M  36 
taken  on  103a-D  emulsions.  Micropho- 
tometer  tracings  indicated  that  the  signal- 
to-noise  performance  of  the  image  tube 
was  roughly  on  a  par  with  that  obtained 
by  short-exposure  photography.  The  same 
image  density  would  have  been  obtained 
in  a  much  shorter  exposure  if  the  cathode 
of  the  image  tube  had  been  of  average 

The  useful  exposure  time  of  this  tube  for 
individual  photographs  was  limited  to  less 
than  1  minute  by  the  background  emission 
from  the  SI  (infrared-sensitive)  cathode. 
In  order  to  extend  the  useful  exposure 
time,  the  equipment  was  modified  so  that 
the  cathode  could  be  cooled  with  liquid 
nitrogen.  This  reduced  the  spurious  back- 



ground  in  a  tube  tested  in  February  to 
the  extent  that,  with  exposure  times  of  1 
minute,  no  background  exposure  could  be 

These  preliminary  tests  indicate  that  the 
barrier-film  tubes  are  potentially  useful  de- 
vices for  astronomical  work,  even  though 
the  method  is  somewhat  encumbered  by 
the  continuously  pumped  vacuum  system 
required  on  the  telescope.  The  Committee 
is  continuing  their  further  development. 

A  more  convenient  system  of  image 
intensification  would  be  one  in  which 
electrons  originating  from  a  photocathode 
impinge  upon  a  phosphor  screen  and  the 
resulting  optical  image  is  recorded  by  con- 
ventional photography.  To  do  this  with  a 
net  gain  over  direct  photography,  however, 
the  photoelectrons  must  be  multiplied  in- 
side the  image  tube,  thereby  increasing 
the  brightness  of  the  image  on  the  phos- 
phor screen  to  compensate  for  the  inef- 
ficient process  of  photographing  the  screen. 

Two  methods  of  producing  this  internal 
intensification  are  being  explored.  One  of 
them  involves  a  series  of  thin  membranes, 
each  having  a  phosphor  on  one  side  and 
a  photocathode  on  the  other.  At  present 
two  companies  are  trying  to  overcome  the 
technical  difficulties  of  making  this  type  of 
tube,  but  no  samples  suitable  for  telescope 
tests  were  yet  available  at  the  end  of  the 
report  year. 

The  second  method  involves  a  similar 
series  of  thin  membranes  that  function  as 
secondary-electron  emitters.  As  in  the  first 
method,  the  electrons  must  be  focused 
from  the  photocathode  to  the  first  mem- 
brane, from  each  membrane  to  the  next, 
and  from  the  last  membrane  to  the  phos- 
phor screen.  Development  of  this  type  of 
secondary-electron  image  intensifier  is  now 
being  carried  forward  by  another  company. 

Both  these  methods  appear  promising, 
but  the  technical  difficulties  that  still  exist 
in  the  fabrication  of  the  tubes  must  be 
overcome  before  this  type  of  image  in- 
tensifier can  be  evaluated  under  the  con- 
ditions encountered  in  astronomical  work. 
The  Committee  also  looks  forward  to  fu- 
ture tests  on  similar  tubes  with  internal 
cascade  intensification,  and  with  electrical 
output,  which  could  have  many  advantages 
if  the  signal-to-noise  ratio  is  satisfactory. 

The  National  Science  Foundation  made 
a  grant  to  the  Committee  to  support  more 
fully  the  industrial  costs  and  the  testing  pro- 
gram now  required  for  the  development  of 
image-intensifying  tubes  for  use  in  astron- 
omy. This  grant  will  supplement  earlier 
appropriations  for  the  task  of  extending  the 
range  of  telescopes  to  fainter  objects,  both 
in  direct  photography  and  in  spectro- 
graphic  measurement.  The  Committee 
acknowledges  the  vigorous  and  effective 
contributions  made  by  W.  Kent  Ford,  Jr. 


Washington,  District  of  Columbia  MERLE  A.  TUVE,  Director 



Introduction    95 

Experimental  Geophysics  97 

Radio  astronomy   97 

Precise  position  apparatus 97 

Radio  emission  from  the  sun 98 

Flux  measurements  of  radio  sources .  99 
Satellite  observations  at  40  and  108 

mc     100 

Moon  reflections 100 

Other  activities 101 

Radio  hydrogen   101 

The  earth's  crust 104 

Seismic  studies   106 

Mineral  age  measurements Ill 

Theoretical  and  Statistical  Geophysics .  .  .  117 

Equatorial  electrojet   117 

Cosmic-ray  investigations  118 

Laboratory  Physics    122 

Nuclear  physics   122 

Coulomb  excitation  of  xenon 122 

Level  structure  of  Na22 123 

Atomically     polarized     ion     source 

(APIS)    125 

Biophysics   127 

Introduction    127 

Fractionation  methods    129 

Properties  of  ribosomes 132 

Kinetics  of  incorporation  into  mac- 

romolecular  fractions    138 

Virus  purification    142 

Incorporation  of  amino  acid  analogs 

into  bacterial  proteins 143 

Cell-free  systems    148 

The  metabolic  pool  problem  in  E. 

coli   149 

Synthesis  in  mouse  tissue 155 

Hydra  and  Planaria 157 

Operations  and  Staff 162 

Cooperative  work  of  the  Department.  162 

Administration  and   operation 163 

Bibliography   163 

Personnel 165 

Carnegie  Institution  of  Washington  Year  BooJ{  57,  1957-1958 

Plate  1 

Department  of  Terrestrial  Magnetism 

•    •;:.. 

^    :  ^  ■■        \      •  •      .*•     • 

The  Carnegie  Andes  Expedition  for  the  International  Geophysical  Year  occupied  more  than  200 
observing  sites  during  1957  in  Peru,  Bolivia,  and  Chile.  By  observations  on  the  waves  from  large 
explosions  in  open-pit  copper  mines,  a  beginning  was  made  in  determining  the  nature  of  the  deep 
rock  structures  of  subnormal  density  which  cause  the  elevation  of  these  huge  mountain  masses. 
About  half  of  the  observing  locations  were  above  14,000  feet,  on  the  spectacular  high  plateau  of  the 


Basic  research  is  not  a  business  activity. 
It  is  a  highly  personal  activity,  a  quest, 
rooted  in  devoted  personal  interest.  It 
grows  out  of  individual  initiative,  and  its 
progress  is  guided  and  modified  by  the 
investigator  himself.  Thus,  the  work  of  a 
basic  research  laboratory  is  not  designed 
by  administrators  as  a  comprehensive  pro- 
gram, and  it  correspondingly  fails  to  ex- 
hibit a  clear  and  stepwise  progression 
toward  a  recognized  end  result.  There  is 
no  plan  or  expectation  of  an  end  result; 
there  are  to  be  recorded  only  incidents  of 
fresh  recognition  or  opportunity  as  the 
investigator  hews  out  a  new  path  in  a 
recognized  direction. 

Research  is  not  measured  by  the  attain- 
ment of  predicted  goals  but  is  observed  as 
a  process  of  individual  development  and 
fulfillment  in  a  man's  relationship  to 
ideas  and  his  tests  of  their  validity.  A  later 
process,  a  bit  to  one  side  of  the  central 
creative  effort,  concerns  the  wider  sharing 
of  his  own  testing  of  ideas  with  others, 
for  whatever  common  use  they  may  serve. 
But  most  investigators  do  not  find  it  satis- 
factory to  work  strictly  alone.  Scientific 
knowledge  and  current  research  activity 
are  today  so  broad  and  yet  so  intensive 
that  no  one  individual  finds  in  himself 
all  the  information  and  all  the  stimulus 
needed  to  go  ahead  effectively,  even  on 
one  narrow  front  of  interest,  and  it  is 
natural  for  a  loose  and  free  association  into 
small,  informal  groups  to  occur  among 
like-minded  colleagues.  This  has  hap- 
pened also  here  in  the  Department. 

Our  areas  for  research  activity  have  been 
selected  by  a  process  of  discussion  and  on 
the  basis  of  a  mutual  agreement  among 
several  individuals,  or  at  least  between  the 
investigator  and  one  other  critical  col- 
league, that  a  given  area,  related  to  physics 
in  some  intelligible  way,  offers  opportuni- 
ties for  fresh  query  and  new  illumination 
concerning  process  and  order  in  nature. 

This  report  is  then  the  record  for  one  year 
of  the  research  activities  of  a  modest  group 
of  men  whose  basic  allegiance  and  train- 
ing are  in  physics,  and  whose  enthusiastic 
interests  range  over  a  number  of  fields 
where  a  background  in  physics  provokes 
questions  that  can  be  formulated  in  clear 
terms.  The  detailed  research  topics  con- 
cern physical  and  chemical  processes  gov- 
erning the  behavior  of  matter  and  energy 
under  such  widely  differing  circumstances 
as  obtain  in  the  tenuous  gas  clouds  of 
interstellar  space  or  in  the  fantastically 
compact  and  orderly  cosmos  which  is  the 
nucleus  of  a  living  cell. 

The  International  Geophysical  Year  has 
naturally  been  a  great  stimulus  to  our 
staff.  The  IGY,  1957-1958,  was  conceived 
some  years  ago  in  discussions  among  mem- 
bers and  former  members  of  this  Depart- 
ment. Instead  of  undertaking  a  great  ex- 
pansion of  staff  and  expenditures  in  order 
to  operate  expeditions  to  the  Antarctic  or 
elsewhere  to  help  carry  out  the  United 
States  national  program,  and  thus  altering 
the  highly  personal  character  of  our  re- 
search activities  here,  our  efforts  have  been 
restricted  to  specific  enlargements  of  re- 
search activities  already  in  hand,  using  our 
present  staff  and  some  (foreign)  collabo- 
rators. Two  of  these  enlarged  projects 
were  given  additional  support  by  generous 
grants  from  the  National  Science  Founda- 
tion, on  recommendation  of  the  U.  S. 
National  Committee  for  the  IGY. 

An  expedition  to  the  high  Andes  of 
Peru,  Bolivia,  and  Chile,  comprising  ten 
men  with  six  specially  equipped  vehicles, 
undertook  to  measure  the  waves  from 
large  explosions  in  open-pit  copper  mines. 
There,  in  an  effort  to  measure  some  of  the 
major  characteristics  of  the  structure  of  the 
earth's  crust  in  those  spectacularly  elevated 
regions,  in  several  months  of  work  the 
expedition    made    valuable    observations 



along  the  flank  of  the  Andes,  but  found 
unusually  drastic  attenuation  of  the  waves 
under  the  high  Andes  plateau.  The  recon- 
naissance was  therefore  only  partly  suc- 
cessful. Indications  were  obtained  that  the 
volcanic  and  metamorphosed  crust  of  the 
earth  is  thicker  near  and  under  the  Andes 
than  under  the  Rocky  Mountains,  for  ex- 
ample, but  these  explosion-wave  results 
in  South  America  must  be  augmented  by 
observations  on  local  earthquakes.  Prepa- 
rations for  a  collaborative  program  of 
earthquake  studies  are  under  way  with 
university  colleagues  in  Peru  and  Chile. 

The  program  of  isotope  age  measure- 
ments of  igneous  rocks,  especially  those  of 
Precambrian  age,  has  been  given  extra 
emphasis  and  geographical  coverage  as  a 
part  of  our  IGY  contribution.  Field  trips 
for  rock  samples  were  made  in  South 
America  and  in  western  and  northern 
Europe,  and  several  special  field  trips 
through  eastern,  northern,  and  western 
United  States,  to  collect  samples  from  se- 
lected areas.  It  was  also  arranged  to  ex- 
tend and  enlarge  our  previous  collaborative 
studies  of  locations  in  Australia  and  Africa, 
and  for  visiting  investigators  to  work  with 
us  here. 

The  network  of  five  magnetic  stations 
set  up  in  Peru  to  observe  the  equatorial 
"electrojet"  in  the  high  atmosphere  for  the 
IGY,  as  described  in  the  report  for  last 
year,  has  been  in  operation  during  much 
of  this  report  year.  The  data  are  of  particu- 
lar interest  during  periods  of  magnetic 

The  nuclear  physics  work  here  has  re- 
cently started  off  in  a  fresh  direction.  It 
has  been  clear  for  some  time  that  great 
analytical  advantages  would  accrue  in  ex- 
periments which  utilized  a  "polarized" 
beam  of  protons,  a  beam  in  which  the 
proton  spins  are  lined  up.  During  this  year 
arrangements  have  been  made  with  col- 
leagues at  Yale  University  for  a  joint  effort 
to  produce  a  beam  of  polarized  protons 

using  our  large  Van  de  Graaff  generator 
here.  The  6-meter  diameter  of  the  high- 
voltage  electrode  gives  ample  space  for 
installing  a  magnetically  split  beam  of  the 
Rabi  type,  to  replace  the  customary  ion 
source  at  the  top  of  the  high-voltage  tube. 
The  required  apparatus  is  still  under  con- 
struction as  the  report  year  ends. 

Observations  of  the  hydrogen  clouds 
among  the  stars  of  our  galaxy  were  sys- 
tematically carried  forward  during  much 
of  this  year,  using  our  54-channel  recorder 
and  the  Wurzburg  parabola  for  observing 
the  radio  emission  at  1420  megacycles  per 
second;  preparations  also  went  ahead  for 
installing  the  60-foot  parabola  being  con- 
structed for  us  by  the  Blaw-Knox  Com- 
pany. A  comprehensive  survey  of  the 
hydrogen  clouds  nearest  the  sun  (large 
apparent  diameter)  was  carried  well  to- 
ward completion,  and  detailed  records  for 
about  half  of  the  equatorial  (Milky  Way) 
areas  were  obtained.  Solar  activity  was 
observed  with  special  radio  antenna  arrays, 
and  performance  tests  were  begun  of  the 
extended  arrays  constructed  and  installed 
this  year  for  precise  radio  positions,  with  a 
view  to  added  optical  identifications  of 
celestial  radio  sources. 

The  biophysics  group  has  been  con- 
cerned with  studies  of  the  sites  and  rates 
of  synthesis  within  the  living  cell  of  nucleic 
acids  and  of  certain  specialized  proteins. 
The  articulation  of  the  synthetic  mecha- 
nisms, and  the  distribution  of  these  activi- 
ties among  the  morphological  components 
of  the  bacterial  cell,  continue  to  be  the 
focus  of  interest  here,  following  naturally 
in  sequence  after  the  previous  compre- 
hensively analytical  studies  of  amino  acid 
synthesis  and  incorporation.  Observations 
of  the  detailed  but  finite  specificity  of  syn- 
thetic mechanisms  and  the  corresponding 
degree  of  variation  in  the  resulting  pro- 
teins, using  chemical  analogs  of  amino 
acids,  have  constituted  another  area  of 
special  interest  and  activity. 




B.  F.  Bur\e,  W.  C.  Ericsson,1  J.  W.  Fir  or, 

H.  L.  Heifer,2  H.  E.  Tatel?  M.  A.  Tuve, 

and  H.  W.  Wells 


Testing  and  construction  have  proceeded 
on  a  400  mc/sec  linear  array  suitable  for 
measuring  positions  of  radio  sources  to 
within  a  few  square  minutes  of  arc.  The 
previous  report  (1956-1957)  described  the 
preliminary  tests  that  led  to  the  present 
array,  which  consists  of  a  pair  of  V  reflec- 
tors, each  614  feet  long,  arranged  on  an 
east- west  baseline  with  a  spacing  of  1842 
feet  between  centers.  The  arrays  can  either 
be  used  separately,  giving  a  fan  beam 
%°X20°  to  half-power  points,  or  together 
as  an  interferometer  with  a  lobe  spacing 
of  4  minutes  of  arc.  The  calculated  gain 
indicates  that  the  two  arrays  should  have 
the  effective  collecting  area  of  two  90-foot 
paraboloids.  Since  it  is  contemplated  to 
make  measurements  at  several  different 
azimuths,  the  antennas  were  built  in  sec- 
tions to  facilitate  moving.  Each  section  is 
a  60°  V,  8  feet  on  a  side  and  9l/2  feet  long, 
fed  by  four  collinear  full-wave  dipoles.  All 
the  elements  are  connected  by  a  symmetri- 
cal branching  feeder  system  with  open- 
wire  transmission  line  used  throughout. 
At  the  close  of  the  year,  all  the  elements 
had  been  mounted  and  final  connections 
were  being  made. 

An  essential  test  was  performed  on  the 
phase  stability  of  open-wire  transmission 
lines,  a  vital  factor  in  the  accuracy  of  the 
final  measurements.  The  extent  to  which 
atmospheric  conditions,  such  as  tempera- 
ture, humidity,  frost,  and  dew,  would  limit 
the  use  of  open-wire  lines  was  unknown 
when  the  present  program  was  under- 
taken, although  it  was  expected  that  the 
lines  would  be  usable  for  a  considerable 

1  Now  with  Convair  Scientific  Research  Lab- 
oratory, San  Diego,  California. 

2  Now  with  Mount  Wilson  and  Palomar  Ob- 
servatories, Pasadena,  California. 

3  Deceased  November  15,  1957. 

fraction  of  the  time.  During  a  2-week 
period  in  November  1957,  measurements 
were  made  that  confirm  these  expectations 
and  provide  a  quantitative  estimate  of  the 
accuracy  to  be  expected. 

Absolute  electrical  line  length  was  not 
important,  since  balanced  feeder  systems 
were  planned  for  the  arrays,  but  it  was  vital 
that  the  distances  from  the  center  of  the 
line  to  either  end  remain  equal,  within  the 
desired  limits  of  tolerance.  Most  of  our 
measurements  were  made  on  a  line  1900 
feet  long,  consisting  of  two  parallel  no.  6 
copper  wires,  with  a  spacing  of  3  inches, 
and  supported  every  50  feet  by  nylon 
filaments.  Short  circuits  were  placed  at 
the  ends,  and  at  the  center  a  385  mc/sec 
signal  was  fed  into  each  half.  By  compar- 
ing the  relative  phases  of  the  two  reflected 
waves,  the  difference  in  electrical  lengths 
of  the  two  halves  could  be  measured,  as- 
suming that  all  the  reflected  wave  came 
from  the  short  circuit  at  the  end.  By  re- 
placing the  short  circuit  with  a  matched 
termination,  it  was  possible  to  measure  the 
reflected  wave  from  irregularities,  droplets 
of  water  on  the  line,  and  similar  spurious 
effects.  Using  this  system,  a  change  in  line 
balance  of  2  parts  per  million  could  be 
detected,  although  spurious  reflections,  de- 
termined by  measurements  with  the 
matched  load,  limited  the  final  accuracy 
to  1  part  in  50,000.  In  general,  the  balance 
was  either  very  good  or  very  bad,  a  differ- 
ence of  more  than  1/12,000  being  con- 
sidered bad.  During  the  228  hours  of 
measurement,  64  per  cent  of  the  time  the 
balance  was  better  than  1/45,000,  corre- 
sponding to  an  angular  accuracy  (when 
used  in  an  array  for  measuring  radio 
source  positions)  of  about  4  seconds  of 
arc.  Seventy-five  per  cent  of  the  time  the 
balance  was  better  than  our  desired  toler- 
ance of  1/12,000.  The  short-term  stability 
was  excellent,  most  of  the  periods  of 
1/45,000  stability  being  many  hours  long. 
Serious   unbalance,  that   is,  greater   than 



1/12,000,  occurred  only  at  times  of  heavy 
frost,  heavy  dew,  or  falling  rain. 

Consequently,  open-wire  lines  can  be 
expected  to  give  many  consecutive  hours 
during  which  their  phase  stability  should 
be  sufficiently  good  to  use  in  our  position 
measurements  of  radio  sources.  Perhaps 
one-quarter  of  the  time  the  phase  unbal- 
ance will  be  too  great,  but  the  unbalance 
will  be  shown  on  the  monitor,  and  at  such 
times  other  components  of  the  system,  such 
as  the  antennas  themselves,  would  prob- 
ably also  be  suspect.    (B.  F.  B„  ].  W.  F.) 


The  solar  fan  beam  array  described  in 
last  year's  report  has  been  operating  now 
for  over  a  year,  and  enough  records  have 
been  obtained  to  describe  the  types  of 
events  occurring  on  the  sun  at  a  wave- 
length of  88  cm  (340  mc). 

The  most  common  feature  of  the  sun 
at  this  wavelength  is  the  quiet  bright  spot 
which  was  illustrated  by  some  scans  in 
our  report  last  year.  Such  a  spot  produces 
a  flux  at  our  antenna  ranging  from  the 
smallest  that  can  be  recognized,  about 
5  X  10~23  watt/meter/cycle/sec,  up  to  about 
5  X  10"22.  The  positions  on  the  solar  disk 
usually  agree  with  that  of  large  plages, 
but  other,  equally  large,  plages  apparently 
have  no  radio  bright  spot  associated  with 
them.  The  radio  spots  that  do  occur  per- 
sist for  several  days  or  sometimes  for  a 
complete  disk  passage. 

There  seems  to  be  no  clear  distinction 
between  the  quiet  bright  spots  and  the 
other  common  feature  of  the  88-cm  sun, 
the  active  spots.  The  active  spots,  although 
persisting  for  several  days  like  the  quiet 
ones,  show  changes  in  intensity  and  pro- 
duce frequent  small  bursts,  lasting  a  second 
or  less.  It  is  not  uncommon  for  a  quiet 
spot  to  become  active  during  a  disk  pas- 
sage. (A  burst-producing  region  was  also 
illustrated  in  the  previous  report.)  The 
active  spots  can  be  much  more  intense  than 
the   quiet   ones;   the   largest   so   far   seen 

gave  a  steady  flux  of  60  X 10~22  with  in- 
stantaneous values  during  the  bursts  of 
perhaps  twice  this  value. 

Two  comparisons  have  been  made  of  the 
positions  of  active  and  quiet  spots  with 
those  seen  at  other  wavelengths.  Some 
scans  of  the  sun  made  at  21-cm  wavelength 
have  been  published  by  the  Australian 
group,  and  similarly,  the  results  of  scans 
made  at  177-cm  wavelength  have  been  pub- 
lished by  the  French  group.  Neither  of 
these  sets  of  scans  was  made  simultane- 
ously with  our  88-cm  scans,  but  for  spots 
persisting  for  several  days,  and  if  proper 
allowance  is  made  for  solar  rotation,  the 
comparisons  should  still  be  of  value. 

At  the  shorter  wavelength,  21  cm,  the 
features  of  the  sun  are  almost  entirely  quiet 
bright  spots  agreeing  in  position  with  prac- 
tically all  large  plages.  There  must  be, 
then,  21-cm  bright  spots  with  no  88-cm 
counterpart,  and  the  comparison  of  the 
records  quickly  verifies  this  conclusion. 
But  the  more  surprising  result  is  that 
the  active  spots  at  88  cm  frequently  agree 
in  position  with  21-cm  quiet  bright  spots. 
If  borne  out  by  more  extensive  compari- 
sons, this  fact  provides  the  first  link  be- 
tween the  regions  seen  at  21  cm,  which  all 
evidence  indicates  are  the  result  of  rela- 
tively hot,  dense  regions  in  the  lower 
corona,  and  the  noise  storm  bursts  of  the 
long  wavelengths,  which  are  almost  cer- 
tainly nonthermal  in  origin. 

The  comparison  with  the  177-cm  scans 
shows  that,  whereas  many  noisy  regions 
are  seen  in  common,  some  very  intense 
features  at  the  longer  wavelength  are  ab- 
sent at  88  cm.  The  conclusion  is  that  there 
are  disturbed  regions  high  in  the  solar 
atmosphere  which  do  not  extend  to  the 
lower  levels  sampled  by  the  88-cm  radia- 

At  present,  comparisons  are  in  progress 
with  groups  in  many  countries,  including 
measurements  at  wavelengths  of  177,  88, 
21,  9.8,  8,  and  3  cm,  and  at  optical  wave- 
lengths.   (/.  W.  F.) 


$7 -Megacycle  Christianson  Array 

A  survey  was  conducted  at  Derwood  to 
determine  the  optimum  recording  fre- 
quency for  a  Christianson  array  at  these 
longer  wavelengths.  Tests  over  periods  of 
several  weeks  showed  that  the  sun  (and 
other  radio  sources)  could  be  observed 
without  serious  interference  using  a  fre- 
quency of  87.3  mc  (345  cm). 

Preparations  are  being  made  for  the 
installation  of  this  array  at  the   Seneca, 

tensity  measurements  on  discrete  radio 
sources.  All  observations  have  continued 
to  apply  the  same  principles  of  maximum 
simplicity  to  achieve  minimum  uncertainty 
in  the  results. 

The  status  of  "color  index"  of  radio 
stars  is  somewhat  improved  since  last  year, 
partly  as  a  result  of  the  efforts  of  URSI 
Commission  V  at  the  Boulder  meetings 
in  August  1957.  The  high  level  of  solar 
activity,  however,  has  continued  to  keep 


18.5         500  — 
27          430 
40          345 
50          380 
87          200 
108          200 
207          110 







<>  ^ 


























■      \ 



40  60  80       100  150         200  300 


Fig.  1.    Flux  of  Cassiopeia  A,  Derwood,  Maryland,  CIW-DTM,  June  1958. 

Maryland,  site  using  a  3000-foot  east-west 
baseline.  Sixteen  antenna  elements  and 
branching  open-wire  feeder  systems  are 
being  fabricated.  Tests  have  been  made 
on  several  types  of  antennas,  and  a  simple 
90°  corner  reflector  has  been  found  to  be 
most  suitable.  The  instrument  will  have 
fruitful  application  to  studies  of  "halo"  or 
related  effects  of  radio  sources  in  addition 
to  the  solar  observations.    (H.  W.  W.) 


The  report  of  last  year  outlined  the  basic 
objections  and  techniques  of  absolute  in- 

investigators  from  returning  to  the  spec- 
trum below  30  mc,  which  has  been  the 
region  of  greatest  disagreement. 

Further  calibrations  of  the  diode  noise 
generators  reveal  excellent  agreement  with 
thermal  sources  up  to  frequencies  some- 
what greater  than  100  mc,  but  at  200  mc 
the  instruments  were  found  to  be  in  error 
by  factors  of  2  or  3. 

Our  experience  with  satellite  recording 
at  40  mc  demonstrated  the  feasibility  of 
observing  the  strong  radio  sources  using 
receivers  with  narrow  band  widths,  say 
1  to  5  kc.  The  result  is  an  ability  to  make 


flux  measurements  in  parts  of  the  spectrum 
(that  is,  40  mc)  that  were  completely 
masked  by  interference  with  the  former 
broad-band  (100  kc)  receivers.  Before 
obtaining  a  flux  value  for  Cassiopeia  A  at 
40  mc  the  series  of  measurements  extend- 
ing from  18.5  to  207  mc  strongly  suggested 
a  distribution  as  shown  by  the  dashed  line 
of  figure  1.  Its  characteristic  feature  is  a 
"break"  or  change  in  slope  at  about  50  mc. 
The  addition  of  our  40-mc  value,  however, 
would  suggest  a  potential  straight-line 
"color  index"  as  shown  by  the  solid  line 
of  the  figure. 

These  developments  clearly  indicate  a 
need  to  verify  our  earlier  results  at  50  mc. 
Intercomparisons  with  other  investigators 
show  our  50-mc  results  to  be  more  "in 
agreement"  than  our  40-mc  values.  Any 
change  in  slope  as  suggested  by  the  dashed 
line  of  figure  1  should  not  necessarily  be 
interpreted  as  a  basic  change  in  the  spec- 
trum as  generated.  More  probably  it  would 
result  from  modification  in  color  index 
imposed  by  dust  or  other  particles  between 
the  source  and  the  earth,  the  absorption 
becoming  more  pronounced  at  the  lower 
frequencies.  Another  possible  effect  of 
absorption  in  interstellar  space  may  be  a 
net  over-all  change  in  slope  of  the  spec- 
trum of  a  particular  source  with  no  ob- 
servable "breaks."  Either  type  of  spectrum 
has  interesting  possibilities  for  develop- 
ment of  a  scale  of  distance  (or  perhaps 
a  measure  of  galactic  dust)  but  must 
await  the  precise  establishment  of  color 
indices  for  several  discrete  radio  sources. 
(H.  W.  W.) 


Within  a  few  hours  of  the  launching 
of  the  first  Soviet  satellite  on  October  4, 
1957,  we  were  recording  its  signals  at  40 
mc.  An  interferometer  designed  for  38  mc 
was  quickly  converted  to  the  satellite  fre- 
quency, and  the  first  transit  was  recorded 
at  22h25m  EST,  October  4  (03h25ra  GMT, 
October  5).  The  continued  monitoring  of 

its  signals  through  October  25  led  to  dis- 
covery of  an  "image"  effect  and  to  other 
interesting  propagation  phenomena. 

In  addition  to  the  characteristic  inter- 
ference pattern  of  normal  satellite  transits 
there  were  a  few  distinct  recordings  at 
times  when  the  satellite  was  on  the  oppo- 
site side  of  the  earth  near  an  antipodal 
point.  This  effect  has  been  described  as  a 
radio  "image"  or  "ghost"  satellite. 

Although  "images"  are  well  established 
in  optical  and  electrical  theory,  it  is  difficult 
to  offer  a  very  satisfactory  explanation  of 
the  effect.  A  review  of  the  literature  re- 
vealed that  a  basic  theory  for  antipodal 
focusing  was  developed  by  Professor  B. 
van  der  Pol  of  Holland  in  1919.  Any  theo- 
retical or  physical  explanation  of  the  image 
effect  appears  to  require  a  stable,  ho- 
mogeneous atmosphere  with  considerable 
dependence  on  the  position  or  height  above 
earth  of  the  transmitter. 

Subsequent  recordings  of  the  second  and 
third  Soviet  satellites  at  40  mc  have  failed 
to  produce  any  additional  "images."  Con- 
tributing factors  are  believed  to  be  the 
greatly  reduced  transmitter  power  and 
substantially  different  orbits. 

At  108  mc  the  signals  from  United  States 
satellites  were  also  monitored  for  unusual 
propagation  effects.  Although  over  a 
period  of  several  months  we  have  not 
identified  any  radio  images  there  is  reason 
to  anticipate  substantial  data  on  atmos- 
pheric refraction  and  absorption  when  the 
recordings  are  analyzed.    (H.  W.  W.) 


A  short  series  of  experiments  were  con- 
ducted in  cooperation  with  Stanford  Uni- 
versity and  the  Evans  Signal  Laboratory. 
Transmitting  stations  at  Stanford  or  Evans 
Signal  Laboratory  beamed  signals  on  the 
moon  at  106  to  108  mc.  When  received 
with  a  conventional  phase-switching  inter- 
ferometer the  moon  signals  appeared  as 
an  artificial  radio  source  showing  the  char- 
acteristic lobe  structure.  In  addition  to 
the  moon  signals,  tests  with  the  Evans 



Signal  Laboratory  (distance:  150  mi) 
showed  a  number  of  spurious  interference 
patterns  thought  to  be  reflections  from 
high-flying  aircraft.  Other  applications  of 
these  techniques  have  been  discussed. 

Calculations  based  on  known  parameters 
of  the  experiments  show  the  effective  radar 
cross  section  of  the  moon  to  be  105  sq  km, 
which  is  about  1  per  cent  of  its  actual 
cross  section.    (H.  W.  W.) 


A  phase-switching  interferometer  at  38 
mc  has  been  installed  and  operated  at  the 
National  Radio  Astronomy  Observatory, 
Green  Bank,  West  Virginia.  The  instru- 
ment is  intended  to  provide  an  inde- 
pendent check  on  certain  radio  star  ob- 
servations. It  has  also  been  useful  for 
demonstration  purposes  and  as  a  means  of 
assessing  the  level  of  radio  interference  in 
the  area. 

A  search  has  been  made  for  Jupiter  sig- 
nals at  40  mc.  Several  months  of  records 
at  times  normally  favorable  to  the  recep- 
tion and  identification  of  Jupiter  signals 
have  been  examined.  The  results  are  en- 
tirely negative,  no  signals  having  been 
identified  at  this  frequency.    (H.  W .  W.) 


The  density  distribution  and  velocity 
distribution  of  atomic  hydrogen  gas  clouds 
in  our  galaxy  continue  to  be  of  major 
interest  to  our  radio-astronomy  group. 
Attempts  to  observe  atomic  hydrogen  (by 
the  radio  line  emission  at  1440  mc/sec) 
in  more  distant  objects,  whether  by  ab- 
sorption (as  for  Cygnus  A)  or  by  emission, 
has  been  postponed  until  we  can  have  the 
use  of  a  parabola  larger  than  the  8-meter 
Wiirzburg  of  our  present  equipment. 

For  several  years  two  or  three  of  us  have 
been  intimately  concerned  with  the  design 
of  large,  movable  parabolas  for  radio 
astronomy,  and  have  served  the  govern- 
ment and  other  groups  in  these  studies. 
Early  in  this  report  year  (July  1957)  an 
order    was    placed    with    the    Blaw-Knox 

Company  of  Pittsburgh  for  a  60-foot 
parabola  on  an  equatorial  mount,  but  con- 
struction has  been  delayed  by  government 
priorities,  and  erection  is  not  expected 
until  late  autumn  1958.  Somewhat  larger 
parabolas  (85  feet)  based  on  the  design 
developed  here  at  the  Department  are  be- 
ing built  by  Blaw-Knox  for  the  University 
of  Michigan  and  for  the  National  Radio 
Astronomy  Observatory  at  Green  Bank, 
West  Virginia.  During  this  year  most  of 
our  efforts  in  connection  with  radio  hydro- 
gen have  been  devoted  to  systematic 

The  54-channel  21-cm  radiometer  de- 
scribed in  last  year's  report  gave  gratify- 
ing performance  for  much  of  the  year. 
With  this  instrument  the  whole  spread  of 
Doppler  velocities  in  any  one  direction  in 
the  sky  (up  to  450  km/sec)  is  measured 
in  a  single  observation  lasting  4.8  minutes. 
No  modifications  were  made  of  the  elec- 
tronics, most  of  the  effort  this  year  being 
expended  on  survey  work  and  on  diag- 
nostic tests.  In  the  previous  report  it  was 
noted  that,  when  the  system  is  used  on 
the  sky,  zero  shifts  of  ±1°  K  can  occur 
in  a  period  of  several  hours.  It  appears 
that  this  effect  is  largely  due  to  variable 
mismatch  in  the  antenna  itself.  Empiri- 
cal antenna  adjustments  are  necessary  to 
minimize  the  shift  and  slope  of  the  zero 
line,  a  consequence  of  the  rather  large 
separation  (six  wavelengths)  between  an- 
tenna and  mixer.  In  practice,  the  best 
method  of  overcoming  the  problem  of 
zero  level  variability  has  proved  to  be 
the  taking  of  check  runs  on  parts  of  the 
sky  known  to  contain  little  hydrogen. 
Such  "cold  sky"  runs  taken  every  few 
hours  permit  the  determination  of  the 
zero  level  within  ±1°  K  unless  unusual 
atmospheric  conditions  such  as  snow  or 
rain  are  prevailing. 

The  precision  of  single  observations  of 
4.8  minutes'  duration  on  each  single  chan- 
nel of  the  54-channel  set  was  found  to  be 
±2.3°  K  for  each  separate  channel,  and 
averages  of  three  observations  on  the  same 



channel  showed  a  statistical  rms  fluctua- 
tion of  1.5°  K  (P.  E.  1.0°  K).  The  same 
result  was  obtained  for  runs  taken  2 
months  apart  on  cold  areas  of  the  sky. 
This  value  is  in  reasonable  agreement 
with  the  input  circuit  equivalent  noise 
temperature  of  around  1000°  K  indicated 
by  the  Bendix  diode.  Each  channel  is 
12  kc  wide  to  half-power  points,  and  chan- 
nels are  centered  on  points  18.9  kc  apart, 
corresponding  to  a  Doppler  shift  of  4 
km/sec  from  one  channel  to  the  next,  each 
channel  covering  a  hydrogen  velocity 
spread  of  about  2l/2  km/sec. 

stant  effectively  broadens  the  filter  band- 
width, thus  diminishing  the  frequency 
resolution  of  the  instrument,  and  long- 
period  frequency  variations  affect  the  posi- 
tion of  the  local  standard  of  rest  and  hence 
the  accuracy  of  relative  velocity  measure- 
ments. Both  effects  were  checked  simul- 
taneously by  radiating  an  independent 
crystal  oscillator  harmonic  close  to  1420 
mc/sec  from  the  laboratory  window  to- 
ward the  parabolic  dish.  When  the  local 
oscillators  are  properly  adjusted,  short-term 
frequency  fluctuations  appear  to  have 
negligible  effect  on  the  bandwidth,  but  the 




1420  MC 





27  MC 

27  MC 

A  '393  MC 
1391   M  C 


AT  463   CPS 



463    CPS 


Fig.  2.    Block  diagram  of  54-channel  H-line  receiver,  CIW-DTM,  1957. 

The  over-all  sensitivity  of  the  instrument 
was  measured  by  determining  the  peak 
temperature  at  several  points  in  the  sky  by 
means  of  a  Bendix  noise  diode,  accepting 
the  manufacturer's  correction  (0.70)  for 
transit  time  at  1400  mc/sec.  The  hydrogen 
maximum  at  /=147,  b  =  0  measured  con- 
sistently 7=102°  ±2°;  /=50,  b  =  0  gave 
T  =  99°±3°. 

A  source  of  concern  had  been  the  sta- 
bility of  the  two  local  oscillators  at  ap- 
proximately 1393  mc/sec,  and  particularly 
the  H-frequency  oscillator,  on  which  veloc- 
ity measurements  depend.  Both  the  short- 
term  and  the  absolute  frequency  stability 
are  of  importance,  for  frequency  variation 
in  times  short  compared  to  one  time  con- 

check  has  proved  valuable  as  a  routine 
measurement  of  absolute  frequency,  for 
the  servo  system  that  controls  the  H-fre- 
quency oscillator,  though  stable  over  peri- 
ods of  many  days,  does  drift  slowly  with 
time.  A  desirable  improvement,  now  un- 
der way,  is  the  replacement  of  these  self- 
excited,  servo-controlled  oscillators  by 
crystal-controlled  oscillators  having  the 
usual  multiplier  chains.  A  block  diagram 
of  the  present  equipment  is  shown  in 
figure  2. 

High-Latitude  Survey 

A  particularly  appropriate  task  for  a 
multichannel  receiver  is  the  survey  of 
large    areas    of   sky,   since    the   speed   of 


8=190      b  =  +  20 


5  =  190      b=  +  l5 

H h 

8  =  190     b=-l5 


8  =  190      b  =  +  IO 


5=190     b=+7 



8=190      b=  +  5 



8=190     b=  +  3 



— i '"'"i " 


8=190     b  =  -IO 



8=1^0     b=-7 

H H- 


8  =  190     b=-4 



8=190      b=-l    +     50         100  km/s 

8  =  190      b=  +  l   50         100  km/s 

Fig.  3.     H-line  profiles   taken  with  the  multichannel  recorder  at  a  galactic  longitude  of  190c 
The  vertical  line  on  each  curve  indicates  the  velocity  of  the  local  standard  of  rest. 



gathering  information  is  so  much  greater 
than  for  conventional  scanning  receivers. 
The  first  observing  program,  therefore, 
was  a  general  survey  of  the  entire  visible 
sky  at  galactic  longitudes  from  ±20°  to 
the  galactic  poles  in  order  to  study  the 
distribution  and  dynamics  of  the  local 
interstellar  gas.  The  observations  were 
taken  during  the  summer  of  1957  and  dur- 
ing the  month  of  January  1958,  using  the 
8-m  Wiirzburg  parabola.  Each  complete 
observation  consisted  of  two  sets  of  aver- 
ages of  one  to  three  complete  scans,  inter- 
woven in  frequency.  H-line  profiles  were 
observed  at  10°  longitude  intervals  along 
the  ±20°,  ±30°,  and  ±40°  parallels  of 
galactic  latitude;  at  20°  intervals  along  the 
±50°  and  ±60°  parallels;  at  40°  intervals 
along  the  ±70°  and  ±80°  parallels;  and 
at  the  galactic  poles.  In  addition,  the 
/=150°-330°  meridian  was  observed  at 
10°  latitude  intervals,  and  scattered  ob- 
servations were  taken  near  the  galactic 
plane  to  correlate  with  the  Leiden  and 
Sydney  surveys.  Reduction  of  the  profiles 
has  proceeded  (under  the  supervision  of 
Dr.  Heifer  and  Dr.  Erickson,  formerly 
Fellows  of  the  Carnegie  Institution  at  the 
Department)  at  the  California  Institute  of 
Technology  and  at  the  Convair  Scientific 
Research  Laboratory,  where  a  contour  map 
has  been  constructed  of  integrated  hydro- 
gen intensity  for  the  entire  sky  covered  by 
the  survey.  In  addition,  contour  maps 
have  been  constructed  for  each  latitude 
strip  showing  isophotes  of  antenna  tem- 
perature with  relative  velocity  and  galactic 
longitude  as  ordinate  and  abscissa. 

Low -Latitude  Galactic  Survey 

Since  January  1958,  most  of  the  observ- 
ing time  of  the  54-channel  H-line  receiver, 
in  conjunction  with  the  8-m  Wiirzburg, 
has  been  utilized  for  a  survey  of  a  strip 
40°  wide  along  the  Milky  Way.  Our  ob- 
served areas  are  at  010°,  013°,  017°,  and 
on  similar  points  for  every  10°  of  longitude 
from  1=317°  through  0°  to  /  =  227°  and 
have  centered  on  a  galactic  plane  slightly 

tilted  with  respect  to  the  old  Lund  equator. 
Thus,  in  the  region  to  each  side  of  /  =  360° 
we  have  centered  our  cross  section  on 
b=l°;  from  /  =  040°  to  /=110°  we  have 
centered  on  b  =  0;  and  for  /=113°  to 
1  —  227°  we  have  centered  on  b  =  \° .  Near 
the  plane  we  observe  at  2°  intervals,  in- 
creasing to  5°  intervals  beyond  £  =  10°. 
The  antenna  beam,  to  half-power  points, 
is  about  2°  X3°,  elliptical  in  shape.  Aver- 
ages, usually  of  two  sets  of  three  scans 
each,  interwoven  in  frequency,  were  taken 
as  in  the  high-latitude  survey,  frequent 
checks  of  radiometer  zero  being  made  on 
"cold"  sky.  At  the  close  of  the  report  year, 
approximately  half  the  survey  had  been 
completed,  and  a  preliminary  report,  cov- 
ering a  single  strip  along  the  galactic 
equator,  together  with  two  meridian  strips, 
had  been  prepared  for  the  Paris  Confer- 
ence on  Radio  Astronomy  (July  1958). 
A  sample  appears  as  figure  3,  which  shows 
the  curves  obtained  along  the  7=190° 
meridian.  A  long  vertical  line  on  each 
curve  indicates  the  frequency  correspond- 
ing to  zero  velocity  with  respect  to  the 
local  standard  of  rest.  The  usual  conven- 
tion is  adopted,  positive  velocity  corre- 
sponding to  motion  away  from  the  ob- 
server. It  is  expected  that  the  survey,  when 
completed,  will  be  useful  in  conjunction 
with  optical  observations  for  studies  of 
the  large-scale  structure  of  the  galaxy,  in 
addition  to  providing  a  complete  set  of 
check  points  for  detailed  studies  made 
with  instruments  of  greater  resolution. 
(B.  F.  B.,  W.  C.  E.,  /.  W.  F.,  H.  L.  H., 
H.  E.  T.,  M.  A.  T.) 


L.  T.  Aldrich,  H.  E.  Tatel*  M.  A.  Tuve,  and 
G.  W.  Wetherill 

The  International  Geophysical  Year  ac- 
tivities among  our  colleagues  and  friends 
the  world  over  have  served  to  stimulate 
and  encourage  extra  efforts  in  our  own 
areas  of  interest  during  1957  and  1958. 

Our  geological  age   measurement  pro- 

4  Deceased  November  15,  1957. 


cedures,  using  selected  minerals  from  spe- 
cial types  of  rock,  have  been  thoroughly 
tested  during  the  past  several  years.  Dur- 
ing these  two  years  of  the  IGY  particular 
efforts  are  being  made  to  obtain  samples 
from  a  wide  range  of  locations.  The  pre- 
vious age  measurements  from  rubidium- 
strontium  abundance  ratios  gave  agree- 
ment with  the  potassium-argon  age 
measurements  when  micas  from  granites 
and  pegmatites  were  used.  The  resulting 
age  scale  was  calibrated  against  concordant 
uranium-lead  ages  from  some  of  the  same 
pegmatites,  and  was  also  verified  by  care- 
ful laboratory  determinations  of  the  radio- 
active decay  constants  involved.  Our 
range  of  sample  rock  types  and  locations 
has  included  suites  of  rocks  from  eastern 
Canada  and  both  eastern  and  western 
United  States  as  well  as  from  Africa  and 

Other  laboratories  have  now  accepted 
our  calibrations  and  adopted  the  same  de- 
tailed procedures,  and  are  rapidly  and 
widely  extending  the  geological  provinces 
and  type  localities  under  study.  We  have 
likewise  expanded  our  selected  areas  to 
include  other  parts  of  the  United  States 
and  Canada,  and  we  have  also  collected 
samples  in  parts  of  South  America.  As 
this  report  year  closes  we  have  staff  mem- 
bers (summer  1958)  in  Europe  and  in  the 
western  United  States  collecting  additional 
pertinent  samples  for  age  measurements. 
Further  collaboration  with  colleagues  in 
Africa  and  Australia  is  actively  in  progress, 
but  some  delays  have  been  experienced  in 
connection  with  the  extension  of  our 
South  American  efforts. 

The  rock  magnetism  work,  which  has 
been  such  a  conspicuous  feature  of  our 
crustal  studies  for  more  than  a  decade,  was 
largely  discontinued  this  year  when  Dr. 
John  Graham  left  us  to  join  the  staff  of 
the  Woods  Hole  Oceanographic  Institu- 
tion. Before  he  left,  Dr.  Graham  brought 
to  completion  a  remarkable  series  of  field 
and  laboratory  studies.  His  direct  dem- 
onstrations of  the   large  effects   of  mag- 

netostriction on  the  remanent  magnetiza- 
tion of  rock  samples,  as  measured  after 
they  have  been  uncovered  by  erosion  and 
had  been  affected  by  various  other  geo- 
logical processes,  have  challenged  the 
validity  of  most  of  the  spectacular  claims 
made  in  the  literature,  especially  by  British 
workers,  and  have  placed  the  burden  of 
proof  on  those  who  claim  to  deduce  mag- 
netic directions  in  long-past  geological 
epochs  from  measurements  of  this  simple 
kind  on  rock  samples.  Clearly  most  rock 
magnetism  observations  are  the  fortuitous, 
although  partly  systematic,  result  of  a 
long  series  of  changes,  both  chemical  and 
magnetostrictive,  and  any  simple  pro- 
cedure for  drawing  "conclusions"  as  to  the 
direction  of  the  earth's  magnetic  field  at 
the  location  of  the  rock  sample  during 
some  period  of  the  distant  past  appears  to 
be  thoroughly  unjustified.  Complex  cor- 
rection procedures,  or  some  honest  and  un- 
equivocal criteria  for  the  retention  or 
elimination  of  samples,  may  some  day  re- 
trieve part  of  the  body  of  rock  magnetism 
data  for  discussion  of  such  problems  as 
the  wandering  of  the  geographic  pole 
(shift  of  the  crust  relative  to  the  axis  of 
rotation)  or  for  "continental  drift,"  but 
discussions  based  on  the  kinds  of  data 
published  to  date  by  workers  in  several 
countries  are,  in  our  collective  opinion 
here,  deceptively  romantic  and  misleading. 

Rock  magnetism  data  are  still  of  con- 
siderable interest  and  value  in  connection 
with  geological  problems  of  much  smaller 
scale,  however,  as  in  the  working  out  of 
local  or  regional  relationships.  We  have 
been  interested  in  certain  aspects  of  the 
magnetization,  plastic  flow,  and  subsequent 
chemical  alteration  and  remagnetization 
of  lava  flows,  for  example.  During  1957- 
1958  Mr.  Donald  Lindsley  has  been  work- 
ing on  these  problems  at  Johns  Hopkins 
on  a  Carnegie  scholarship  under  Professor 
Aaron  Waters,  and  as  the  report  year 
closes  he  is  in  the  John  Day  Basin  with 
our  magnetometer  equipment. 

Among  the  geographical  features  of  the 


world  there  are  relatively  few  elevated 
plateaus  with  large  areas  at  consistently 
high  elevations,  compared,  for  example, 
with  the  many  ranges  of  mountains.  In 
studying  the  various  large-scale  geotectonic 
mechanisms  that  can  be  devised  or  imag- 
ined to  account  for  continents  and  ocean 
basins,  and  in  our  thinking  about  the  rela- 
tive permanence  of  such  great  features,  to- 
gether with  the  long  succession  and  spe- 
cialized distribution  of  mountain-building 
activities,  we  have  been  attracted  to  the 
examination  of  these  several  high  plateaus. 
They  are  of  special  interest  in  relation  to 
our  studies  over  the  past  decade  using 
waves  from  large  explosions  for  estimating 
the  thickness  of  the  "crust"  of  the  earth. 
We  have  used  explosion  waves  extensively 
in  a  search  for  "roots"  of  mountain  chains. 
The  idea  that  a  mountain  chain  or  plateau 
may  be  held  at  its  higher  elevation  by  hav- 
ing directly  beneath  it  rocks  less  dense 
than  the  surrounding  "mantle"  rock  is 
similar  to  the  concept  of  the  submerged 
ice  "root"  of  an  iceberg.  The  unexpectedly 
shallow  crust  we  found  in  the  Colorado 
plateau  regions  in  1955  with  no  indication 
of  mountain  "roots"  made  us  very  much 
interested  in  the  high  plateau  of  the  Andes 
in  South  America. 

The  intense  IGY  activities  in  all  coun- 
tries stimulated  us  to  undertake,  as  a  major 
feature  of  our  IGY  participation,  a  seis- 
mic expedition  to  Peru,  Bolivia,  and  Chile 
during  the  last  half  of  1957.  It  was  carried 
out,  supported  in  part  by  a  generous  grant 
from  the  National  Science  Foundation,  as 
one  of  the  two  or  three  principal  ways  in 
which  the  Department  has  participated  di- 
rectly in  the  work  of  the  International 
Geophysical  Year.  Another  major  aspect 
of  our  participation  has  been  the  five-sta- 
tion network  across  the  magnetic  equator, 
described  elsewhere  in  this  report. 


Gravity  measurements  on  continents 
and  oceans,  and  over  elevated  plateaus  and 
mountainous  regions  as  well  as  sea-level 

areas,  show  that  the  earth's  crust  is  ap- 
proximately in  "floating  equilibrium,"  evi- 
dently like  rough  ice  on  a  pond.  The  high 
plateaus  and  mountainous  regions  show 
gravity  values  indicating  that  they  are 
held  up  by  the  strength  of  the  crustal  sur- 
face to  an  extent  that  would  support  not 
more  than  400  to  800  feet  of  rock.  The 
rest  of  the  elevation  is  due  to  the  presence, 
somewhere  under  the  plateaus,  of  a  belt 
or  thickness  of  material  less  dense,  on  the 
average,  than  the  corresponding  thickness 
of  rocks  beneath  the  low-lying  plains.  The 
traditional  explanation  for  several  decades 
has  been  that  under  mountain  chains  and 
high  plateaus  the  lighter  crustal  rocks 
bulge  downward  into  the  heavier  mantle 
rocks.  Earthquake  data  have  indicated  the 
crustal  rocks  to  be  30  to  35  km  thick  under 
continental  areas  of  low  elevation,  and 
also  have  given  some  evidence  of  the  ex- 
pected downward  bulge  under  the  Alps 
and  elsewhere.  A  rather  sharp  transition 
is  observed  in  many  localities  between  the 
bottom  of  the  lighter  "crustal  rocks"  (ve- 
locity 6  to  7  km/sec  for  compressional 
waves)  and  the  heavier  mantle  rocks  (ve- 
locity 8  km/sec);  this  transition  is  called 
the  M  discontinuity  (the  Mohorovicic  dis- 
continuity). Our  expedition  of  1955  to  the 
Colorado  plateau,  however,  showed  the 
M  discontinuity  at  about  the  same  depths 
(29  to  33  km)  under  Arizona,  New 
Mexico,  and  Utah,  with  average  elevations 
of  6600  feet  above  sea  level,  as  under  the 
Chesapeake  Bay  region  (31  to  34  km)  and 
elsewhere  at  approximately  sea-level  eleva- 
tions. This  finding  inclined  us  to  accept 
an  alternative  hypothesis,  namely,  that 
the  "mantle"  rocks  under  the  Colorado 
plateau,  perhaps  down  to  a  depth  of  150 
km  or  more,  might  be  of  slightly  (perhaps 
1  per  cent)  lower  density  than  the  "mantle" 
rocks  under  the  coastal  plain  and  similar 
low-lying  continental  areas.  This  might 
be  called  the  hypothesis  of  a  "diffusely 
extended  root."  A  situation  like  this,  in- 
volving slight  regional  differences  in  the 
"mantle,"  would  account  for  the  elevation 



of  the  mountainous  regions  just  as  satis- 
factorily as  the  older  and  highly  specific 
picture  of  a  downward-bulging  "root"  of 
lighter  crustal  rock.  The  M  discontinuity 
under  the  Colorado  plateau  failed  to  show 
any  such  downward  bulge  of  the  lighter 
"crustal"  rocks,  at  least  where  we  were  able 
to  make  explosion  observations. 

There  are  several  other  major  plateau 
areas  in  the  world,  two  of  the  most  con- 
spicuous being  in  the  high  Andes  and  in 
the  Himalayas  and  adjacent,  rather  inac- 
cessible, regions.  For  about  five  years  we 
had  been  considering  the  attractive  possi- 
bility of  making  observations  in  the  Andes, 
inasmuch  as  in  southern  Peru  and  in 
northern  Chile  there  are  several  sizable 
open-pit  copper  mines  in  which  charges  as 
large  as  30  or  50  tons  of  explosive  are  shot 
several  times  each  week.  During  1957  the 
shooting  schedule  reached  a  new  high  level 
of  activity.  We  undertook,  then,  on  our 
IGY  expedition  to  measure  the  waves  from 
these  explosions,  as  they  propagated  along 
lines  parallel  to  the  Andes  and  also  as 
they  passed  across  and  beneath  the  tower- 
ing ranges  of  the  Andes  and  the  high 
(14,000  feet)  plateau  of  Peru,  Chile,  and 

Our  group  of  eight  men  and  six  trucks 
left  Washington  in  July  1957  and  assem- 
bled in  Lima,  Peru,  in  mid-August.  Dur- 
ing the  next  three  months  we  made  ob- 
servations and  recordings  at  more  than  200 
sites  in  Peru,  Bolivia,  and  Chile,  each  site 
being  carefully  selected  for  low  level  of 
local  seismic  "noise"  or  ground  unrest. 
The  selection  of  a  specific  site  in  a  de- 
sired locality  often  required  many  miles 
of  difficult  travel  and  noise  level  tests  at 
many  possible  rock  outcrop  sites.  Radio 
timing  techniques  were  used  on  nearly  50 
large  explosions  while  our  observers  and 
their  trucks,  traveling  in  pairs,  were  deep 
in  the  mountains  80  to  500  km  away.  Most 
of  these  explosions  were  at  the  Toquepala 
Mine  of  the  Southern  Peru  Copper  Cor- 
poration, about  60  miles  northwest  of 
Tacna,  Peru.  During  the  latter  part  of  the 
expedition  the  observed  explosions  were 

at  the  Chuquicamata  Mine  of  the  Chile 
Exploration  Company,  about  125  miles 
northeast  of  Antofagasta,  Chile.  Both 
these  mines  are  on  the  western  edge  of  the 
plateau,  at  elevations  of  about  10,000  feet. 
In  Peru  we  were  greatly  assisted  by  Dr. 
J.  Broggi  and  staff  members  of  his  Insti- 
tuto  Geofisico  de  Huancayo  (offices  in 
Lima) ;  one  of  the  staff  members  partici- 
pated in  our  observations  in  the  Lake 
Titicaca  region.  In  Chile  we  were  joined 
by  Fr.  G.  Saa,  S.  J.,  of  San  Luis  College, 
Antofagasta,  and  Dr.  C.  Lomnitz,  Director 
of  the  Centro  de  Geofisica,  University  of 
Chile,  Santiago.  These  two  scientists 
helped  us  in  our  measurements  on  the  M 
discontinuity  in  the  regions  inland  from 

Although  our  records  apparently  show 
the  M  discontinuity  in  both  Peru  and 
Chile,  for  wave  paths  running  along  the 
flank  of  the  plateau,  at  medium  elevations, 
we  cannot  deny  our  disappointment  in 
finding  that  we  could  not  observe  the 
waves  much  beyond  200  km  in  directions 
across  the  mountains  and  across  the  pla- 
teau, and  that  we  could  find  no  evidence 
for  locating  the  depth  of  the  M  discon- 
tinuity in  these  directions. 

We  had  revised  our  seismic  equipment 
for  this  expedition,  building  new  seismom- 
eters (period  5  cycles  per  second)  with 
higher  output,  to  give  signals  well  above 
the  electronic  noise  level  of  the  amplifiers. 
Our  new  equipment  would  measure  verti- 
cal amplitudes  of  ground  motion  as  low 
as  %  or  even  %o  angstrom  unit  (10-9  cm), 
in  the  frequency  range  5  to  25  cycles  per 
second.  In  North  America  very  few  sites 
have  ever  been  found  with  so  low  a  level 
of  ground  unrest.  When  we  found  the 
attenuation  of  the  waves  across  the  "alti- 
plano"  so  extreme,  we  made  special  efforts 
to  select  quiet  sites,  and  at  many  of  our 
observing  stations  a  wave  arrival  as  small 
as  y2  angstrom  in  amplitude  would  have 
been  definitely  identified,  yet  in  the  range 
from  about  230  km  out  to  about  550  km, 
across  the  plateau,  no  wave  arrivals  were 
found.   In  Peru  and  Bolivia  we  occupied 



more  than  70  such  frustrating  sites,  from 
north  of  Cuzco  to  the  east  of  Oruro,  Bo- 
livia, a  few  miles  from  Cochabamba.  In 
our  extensive  studies  in  eastern,  central, 
and  western  United  States  and  in  the 
Yukon-Alaska  regions  we  have  never  en- 
countered anything  even  resembling  such 
extreme  attenuation  of  explosion  waves. 
The  Andes  are  clearly  different  from  other 
mountains  we  have  studied. 

Meanwhile,  in  the  course  of  numerous 
truck  repair  episodes  involving  trips  down 
from  the  plateau,  large-amplitude  wave 
returns  were  finally  discovered  for  paths 
in  Peru  to  the  northwest  of  Toquepala, 
along  the  flank  of  the  high  plateau,  and 
the  anomalous  attentuation  was  not  ob- 
served along  these  paths.  These  large- 
amplitude  waves  were  very  much  like 
those  recorded  in  the  Rocky  Mountains, 
with  normal  first  arrivals  out  to  200  km 
and  beyond,  and  strong  second  arrivals 
in  the  region  of  200  to  230  km.  In  the 
Rockies  the  strong  second  arrivals  are  ob- 
served at  100  to  150  km  from  the  shots. 
On  the  flank  of  the  Andes,  however,  the 
terrain  is  too  rough  to  place  observers  at 
each  of  the  desired  distances  and  thus  set 
precise  limits  on  the  range  of  distances 
in  which  the  "reflections"  are  observed. 
The  velocities  appeared  to  be  normal 
(about  6  and  about  8  km/sec)  for  both 
crust  and  mantle  regions,  and  the  M  dis- 
continuity was  thus  indicated  to  be  in  the 
range  of  perhaps  44  to  48  km  depth,  on 
the  basis  of  the  strong  second  arrivals  at 
200  km  and  the  first  arrivals  at  farther 
distances,  and  the  possible  existence  of  a 
"hidden"  layer  of  intermediate  velocity 
rock  (7  km/sec)  just  above  the  M  discon- 
tinuity. Our  first  impression  was  that  of 
a  normal  depth  around  34  (or  36)  km, 
based  on  the  two  usual  rock  types,  6  and 
8  km/sec;  this  is  revised  to  an  alternate 
estimate  of  about  46  km  depth  if  there  is 
an  intermediate  layer  above  the  M  discon- 
tinuity. These  sketchy  observations  can 
hardly  be  considered  a  satisfactory  meas- 
ure of  the  depth  of  the  crust,  because  the 
compressional  waves  through  the  mantle 

(Pn  arrivals)  could  not  be  observed  out  to 
the  necessary  distances  of  300  to  400  km  to 
give  a  good  measure  of  the  deeper  veloci- 
ties. Along  the  flank  of  the  Andes  this 
was  due  to  the  topography  of  the  country, 
which  limited  our  access  to  desired  loca- 

In  view  of  the  great  effort  expended  in 
the  Peruvian  mountains  in  obtaining  these 
rather  inconclusive  measurements,  the  very 
marked  attenuation  of  the  waves  across 
the  plateau  into  Bolivia,  and  the  shortness 
of  time,  the  expedition  moved  to  Chile  in 
early  October  1957.  Observing  sites  were 
again  established  on  the  high  plateau 
(14,000  feet)  in  Chile  and  Bolivia,  to  the 
north  and  east  of  Chuquicamata,  and  again 
the  same  very  marked  attenuation  was  ob- 
served as  in  Peru  and  Bolivia,  contrary  to 
our  experience  in  North  America.  Again, 
however,  search  for  large-amplitude  wave 
returns  from  a  deep  reflecting  zone,  as 
from  the  M  discontinuity,  was  successful 
along  the  flank  of  the  mountains,  this  time 
to  the  south.  The  intense  wave  returns 
(second  arrivals)  were  first  observed  here 
at  220  km  from  the  shot,  and  beyond,  indi- 
cating a  depth  of  about  46  km  to  the  M 
discontinuity,  if  only  the  two  rock  types 
are  present  (6  and  8  km/sec).  In  Chile 
there  was  a  faint  indication  of  a  second 
"layer"  of  intermediate  velocity  near  7 
km/sec,  and  if  this  is  actually  present  the 
depth  to  the  M  discontinuity  is  about 
56  km,  much  greater  than  any  we  have 
hitherto  observed. 

Neither  the  data  from  Peru-Bolivia  nor 
those  from  Chile-Bolivia  can  be  considered 
satisfactory  measures  of  the  crustal  struc- 
ture in  those  regions,  or  even  as  definitive 
measures  of  the  existence  and  depth  of 
the  M  discontinuity.  The  large-amplitude 
second  arrivals  observed  in  both  regions, 
however,  surely  were  similar  to  those  we 
have  observed  in  North  America,  although 
they  were  found  at  definitely  much  greater 
distances  from  the  explosion  points  than 
any  in  our  previous  experience.  The  ob- 
servations obtained  in  Peru,  Bolivia,  and 
Chile  are  shown  in  figures  4,  5,  and  6. 



Fig.  4.  Seismic  observations  on  the  waves  from  explosions  were  made  by  the  Carnegie-IGY 
Andes  Expedition,  1957,  at  intervals  out  to  several  hundred  kilometers  along  lines  radiating  out 
from  the  large  open-pit  copper  mines  at  Toquepala,  Peru,  and  Chuquicamata,  Chile,  approximately 
as  shown.  The  high  ranges  of  the  Andes  lie  just  to  the  east  and  north  of  Arequipa  and  the  two 
mining  centers,  and  the  high  plateau  extends  eastward  beyond  La  Paz. 


<  +4 





>H                 i> 

H  =  Altiplano 

\          „t 

L  =  Low 




T  - 

C  =  Coast 

O  +2 





I  T 




CO    +  | 

<  -1 


'H  -ft* 



6.0  km/sec 

*H    V»^H 


6.23  km/sec 

Q   -2 

O        , 




Toquepala,  Peru 

s^8  km/sec 

o    -3 


Sept-Oct   1957 






i               i               i               i 

1                1 

i\           I     .  ... I 


80     120    160    200   240    280    320 

Fig.  5.  Arrival  times  of  explosion  waves  from  the  mine  at  Toquepala,  Peru,  show  no  evidence 
of  any  layer  of  velocity  intermediate  between  the  crust  (6  km/sec)  and  the  mantle  (8  km/sec),  and 
give  only  a  rough  indication  of  the  depth  to  the  M  discontinuity  (36  km  if  no  intermediate  layer). 
Waves  in  and  under  the  high  Andes  were  too  strongly  attenuated  to  be  observed. 



Four  of  the  special  trucks  the  Depart- 
ment bought  and  equipped  for  the  Andes 
expedition  were  left  in  South  America,  in 
the  hope  that  arrangements  might  be  made 
by  friends  and  colleagues  in  South  Ameri- 
can institutions  to  continue  and  extend 
our  seismic  and  gravity  measurements. 
Reductions  of  the  IGY  appropriations  by 
the  government  of  Peru  have  restricted 
the  opportunities  for  collaboration  there, 
but  in  Chile  a  whole  series  of  measure- 
ments is  being  supported  by  the  University 
of  Chile.  Dr.  Cinna  Lomnitz,  Chief  of  the 

and  Chile  Exploration  Company) ;  E.  M. 
Tittman  and  C.  Pollock  (New  York  offi- 
cers of  the  American  Smelting  and  Re- 
fining Company  and  the  Southern  Peru 
Copper  Company);  Charles  McGraw, 
Lima  (Manager  of  the  Marcona  Mining 
Company) ;  and  O.  C.  Laird,  Caracas 
(officer  of  the  Orinoco  Mining  Company)  ; 
and  Warren  T.  Smith  and  Marion  Robin- 
son, local  officers  of  the  Southern  Peru 
Copper  Company  at  Toquepala.  In  addi- 
tion, the  officials  of  the  International  Geo- 
physical   Year    in    Peru,    Bolivia,    Chile, 

£  +4 



§  +  2 


2  - 


slope  7  km/sec    ■  n^ 

6  km/sec 

Chuquicamata,  Chile 
Oct  1957 

8  km/sec 



120  160         200        240 


280       320 

Fig.  6.  If  the  intermediate  layer  indicated  (7  km/sec)  is  actually  present,  the  M  discontinuity 
lies  at  56-km  depth  to  the  south  of  Chuquicamata  (tabs  on  points  show  direction  of  wave  travel) ; 
for  the  simpler  case  (6  km/sec  crust  overlying  8  km/sec  mantle)  the  calculated  depth  is  nearer  46  km. 
Again  the  waves  were  strongly  attenuated  across  the  high  Andes. 

new  Center  for  Geophysics  at  that  Uni- 
versity (in  Santiago),  has  already  made 
measurements  across  the  Andes  escarp- 
ment at  different  latitudes,  using  our 
Worden  gravity  meter,  and  as  this  report 
is  written  (July  1958)  he  and  his  colleagues 
are  beginning  a  further  series  of  observa- 
tions on  the  explosion  waves  from  Chuqui- 
camata using  our  seismic  equipment. 

The  Institution  acknowledges  with 
pleasure  the  cordial  participation  of  the 
officials  of  the  several  mining  companies 
concerned  in  the  arrangements  for  these 
explosion  wave  observations:  C.  E.  Weed, 
Charles  M.  Brinkerhoflf,  Thomas  A. 
Campbell,  and  D.  M.  Dunbar  (New  York 
officers  of  the  Anaconda  Copper  Company 

Argentina,  and  Venezuela  were  most  help- 
ful and  cordial  in  making  local  arrange- 
ments in  their  respective  countries.  We 
received  permission,  for  example,  to  use 
our  two-way  radios  in  each  of  these  coun- 

We  owe  the  local  officials  at  Chuqui- 
camata, especially  Robert  C.  Becker,  Dep- 
uty Manager,  W.  E.  Rudolph,  W.  H. 
Swayne,  and  John  Kent,  and  members  of 
their  geology  and  engineering  staffs,  a 
measure  of  gratitude  that  is  very  deeply 
and  personally  felt.  We  had  been  experi- 
encing difficulties  in  reaching  desired  dis- 
tances in  the  mountains,  along  selected 
directions  from  the  shots,  owing  to  various 
mechanical  failures  of  our  four-wheel-drive 



heavy-duty  trucks.  When  we  were  so  dis- 
tressed by  the  prospect  of  further  break- 
downs at  inaccessible  points,  far  from  any 
normal  roads  or  traffic,  that  we  were  ready 
to  abandon  an  important  sector,  these 
courageous  officials  offered  to  send  a  re- 
triever expedition  from  the  mine,  if  neces- 
sary, and  gave  us  the  courage  to  proceed. 
Part  of  our  resulting  efforts  were  success- 
ful, but  one  of  our  vehicles  lost  a  rear  axle 
deep  in  the  mountains  near  the  Salar  de 
Atacama,  and  it  required  two  successive 
four-man  expeditions  from  the  mine,  over 
a  period  of  two  weeks,  to  retrieve  our 
truck,  which  was  finally  brought  out  by 
a  southern  route,  several  hundred  miles 
from  the  mine.  The  magnificent  hospi- 
tality and  helpfulness  of  this  action,  com- 
ing as  it  did  when  the  members  of  our 
expedition  were  tired  out  and  due  to 
return,  will  never  be  forgotten. 


L.  T.  Aldrich,  G.  W.  Wetheritt,  G.  R.  Tilton? 
and  G.  L.  Davis  5 

In  previous  annual  reports  of  this  group, 
independent  mineral  ages  obtained  from 
cogenetic  minerals  from  the  same  rocks 
have  been  used  to  check  the  consistency 
and  agreement  of  ages  as  determined  from 
the  measured  ratio  of  daughter  to  parent 
isotopes  from  the  decay  of  isotopes  of 
uranium,  thorium,  rubidium,  and  potas- 

The  results  of  these  measurements  may 
be  summarized  briefly  as  follows:  (1)  the 
fact,  first  demonstrated  by  Wasserburg, 
Hayden,  and  Jensen,  that  the  two  urani- 
um-lead ages  of  pegmatitic  uraninites  al- 
most always  agree  (are  concordant)  has 
been  shown  for  several  additional  loca- 
tions; (2)  rubidium-strontium  ages  of 
micas  and  potassium  feldspars  agree  with 
the  concordant  uranium-lead  ages  of  co- 
genetic  uraninites  within  5  per  cent;  (3) 
potassium-argon  ages  of  mica  agree  with 
uraninite  ages  within  10  per  cent,  and, 
though  Wasserburg,  Hayden,  and  Jensen 

5  Geophysical  Laboratory,  C.  I.  W. 

have  shown  that  pegmatitic  feldspars  seem 
to  lose  an  amount  o£  argon  nearly  propor- 
tional to  the  time  since  they  were  formed, 
this  property  is  not  displayed  by  feldspars 
generally;  (4)  the  rubidium-strontium  and 
the  potassium-argon  ages  of  micas  from 
unaltered  rocks  usually  agree  with  each 
other,  and  when  they  do  not  the  disagree- 
ment is  always  in  the  direction  that  indi- 
cates some  leakage  of  argon  from  the 
mica;  (5)  micas  from  metamorphosed 
rocks  display  a  variable  and  as  yet  not  fully 
understood  pattern  of  ages;  (6)  it  has  been 
further  shown  that  the  uranium-lead  ages 
of  monazite,  zircon,  and  columbite-tanta- 
lite  are  not  predictably  concordant. 

It  could  have  been  predicted  from  last 
year's  report  that  this  year's  work  would 
be  primarily  concerned  with  two  problems. 
The  investigation  of  geographical  patterns 
of  the  ages  of  the  Precambrian  rocks  of 
North  America  and  the  investigation  of 
the  ages  of  minerals  involved  in  meta- 
morphic  processes  have  both  been  contin- 
ued. Progress  on  the  geographical  patterns 
has  not  been  limited  by  our  own  capacity 
for  measurements.  The  data  obtained  at 
the  Lamont  Geological  Observatory  and 
the  Geology  Departments  of  the  University 
of  Minnesota  and  the  Massachusetts  Insti- 
tute of  Technology,  examined  together 
with  our  own,  show  conclusively  the  broad 
outlines  of  major  periods  of  mineral  for- 
mation for  the  parts  of  the  continent  that 
have  now  been  measured.  These  patterns 
are  shown  on  the  map  of  figure  7.  The 
points  with  stubs  attached  are  data  ob- 
tained at  other  laboratories. 

A  large  area  with  rocks  of  ages  exceed- 
ing 2500  million  years  extends  from  west- 
ern Quebec  through  Ontario  to  eastern 
Saskatchewan  and  northern  Minnesota 
and  reappears  in  Montana  and  Wyoming. 
Another  large  area  with  rocks  of  ages 
close  to  1350  million  years  lies  in  western 
United  States,  and  rocks  of  this  age  have 
now  been  found  in  Missouri,  Wisconsin, 
Michigan,  and  Ontario,  so  that  this  period 
also  appears  to  have  been  one  in  which 


processes  of  mineral  formation  occurred 
over  very  widespread  areas  in  North 
America.  It  is  of  some  interest  to  note  that 
no  occurrence  of  rocks  of  this  age  has  yet 
been  found  outside  this  continent.  A  final 
large  region  with  rocks  of  similar  age 
extends  from  northern  Quebec  south  into 
Ontario,  New  York,  New  Jersey,  Virginia, 
and  North  Carolina.  All  these  rocks  were 
formed  close   to   1000  million   years  ago. 

mineral  formation  in  the  Baltimore  area 
during  the  period  300  to  350  million  years 
ago  had  been  found  by  Wasserburg,  Lip- 
son,  and  Pettijohn.  Because  the  Baltimore 
gneiss  is  a  metamorphic  rock  these  mica 
ages  probably  date  the  metamorphism,  but 
give  no  information  about  the  premeta- 
morphic  history  of  the  rock.  New  mineral 
age  measurements  indicate  the  possibility 
of  two  periods  of  mineral  formation.  Dur- 

Fig.  7.     Reliably  dated  localities  in  North  America. 

Regions  near  the  boundaries  of  these  areas 
often  contain  rocks  of  different  ages,  and 
the  patterns  of  the  ages  obtained  from 
different  minerals  in  the  same  rocks  indi- 
cate a  multiple  history  of  mineral  forma- 
tion. Such  areas  appear  notably  in  the  Sud- 
bury region  of  Ontario  and  the  upper 
peninsula  of  Michigan,  and  may  well  be 
found  elsewhere  as  additional  data  are 

The  major  part  of  the  study  of  meta- 
morphic rocks  has  been  the  investigation 
of  the  Precambrian  rocks  of  the  central 
and  southern  Appalachians.   Evidence  of 

ing  one  of  them  (1000  to  1100  million 
years  ago),  zircon  and  probably  potassium 
feldspar  were  formed;  in  the  second  (300 
to  350  million  years  ago),  the  mica  of  the 
gneiss  was  formed.  Measurements  of  other 
Appalachian  gneisses  give  mica  ages  agree- 
ing more  closely  with  the  zircon  ages,  a 
further  indication  of  the  Precambrian  ori- 
gin of  the  metamorphosed  rocks.  Again, 
the  power  of  the  attack  on  geologic  prob- 
lems by  all  methods  of  age  determination 
on  all  available  minerals  has  been  demon- 
In  another  study,  micas  from  many  of 


the  zones  of  the  regionally  metamorphosed 
area  in  Iron  County,  Michigan,  have  been 
measured.  This  area  has  been  studied  in- 
tensively by  Dr.  H.  James  and  his  col- 
leagues of  the  United  States  Geological 
Survey,  with  whose  guidance  the  suite  of 
rocks  for  our  study  was  obtained.  In  sum- 
mary, the  ages  of  the  mica  from  the 
sillimanite,  garnet,  and  biotite  zones  are 
all  the  same  within  the  error  of  measure- 
ment (5  per  cent).  Further,  this  age  is 
that  of  the  mica  from  fresh  granite  thought 
to  be  the  energy  source  for  the  regional 

The  Grenville  Orogeny  in  the 
Appalachian  Region 

The  existence  of  rocks  between  900  and 
1100  million  years  old  in  the  southern  por- 
tion of  the  Canadian  Shield  known  as 
the  Grenville  Subprovince  has  been  recog- 
nized for  many  years.  Until  recently  these 
rocks  have  been  found  only  in  a  relatively 
small  area  in  Quebec  and  Ontario.  It  had 
been  suspected  for  some  time  that  the 
igneous  and  metamorphic  rocks  of  the 
Adirondack  Mountains  in  New  York  rep- 
resented a  southern  extension  of  this  sub- 
province.  It  had  also  been  suspected  that 
the  entire  complex  of  granites,  pegmatites, 
gneisses,  and  metasediments  represented 
the  deeply  eroded  core  of  an  ancient  moun- 
tain system  trending  in  a  northeast-south- 
west direction  through  Ontario,  Quebec, 
and  New  York.  A  few  years  ago  Tilton 
measured  a  sample  of  zircon  from  Natural 
Bridge,  New  York,  in  the  Adirondacks 
and  found  an  age  very  similar  to  ages  in 
the  Grenville  of  Canada.  This  finding 
strengthened  belief  in  the  idea  that  the 
Adirondacks  are  a  southern  extension  of 
the  Grenville  rocks.  If  these  rocks  really 
were  formed  during  an  orogenic  (moun- 
tain-building) epoch,  however,  it  would 
be  expected  that  this  mountain  chain  was 
linear  and  extended  for  thousands  of  miles, 
as  do  more  modern  mountain  chains. 
Therefore,  it  should  be  expected  that  more 
1000-million-year-old    rocks    exist    farther 

along  the  axis  of  the  hypothetical  moun- 
tain system. 

About  two  years  ago  workers  at  Colum- 
bia University  showed  that  rocks  in  the 
New  Jersey  and  Hudson  Highlands,  near 
New  York  City,  were  Precambrian  in  age. 
Further  work  of  our  group  (reported  in 
Year  Book  56)  confirmed  this  discovery 
and  showed  that  their  age  was  approxi- 
mately the  same  as  that  of  the  Grenville 
rocks  farther  north.  During  the  past  year 
the  measurements  have  been  extended  to 
the  south,  and  the  existence  of  rocks  of 
"Grenville  age"  at  least  as  far  south  as 
Virginia  has  been  demonstrated. 

The  principal  difficulty  encountered  in 
measuring  the  ages  of  the  Precambrian 
rocks  in  the  Appalachian  region  is  that 
all  of  them  have  been,  to  a  greater  or 
lesser  extent,  involved  in  the  tectonic 
movements  and  metamorphic  events  of  the 
Paleozoic  Appalachian  orogeny.  In  so  far 
as  they  have  been  recrystallized  during  this 
more  recent  orogeny,  the  record  of  their 
original  age  has  been  destroyed.  Thus  the 
mica  in  most  of  the  metamorphic  rocks 
in  the  Appalachian  region  has  an  age  of 
around  300  million  years,  and  there  is  no 
way  of  knowing  whether  the  rock  is  only 
slightly  older  than  300  million  years  or  is 
actually  of  Precambrian  age.  As  is  shown 
in  tables  1  and  2,  however,  some  rocks 
have  been  found  that  still  bear  evidence 
of  their  Precambrian  age. 

The  group  of  rocks  listed  in  table  1  all 
contain  zircons  of  Grenville  age  together 
with  micas  probably  1000  million  years  old 
that  have  suffered  some  alteration  during 
the  Appalachian  orogeny.  All  these  micas, 
however,  are  definitely  pre-Appalachian  in 
age,  and  the  simplest  interpretation  of  the 
results  is  that  both  the  mica  and  the  zircon 
were  formed  during  the  Grenville  orogeny, 
and  that  the  gneisses  represent  a  further 
extension  of  the  Grenville  gneisses. 

Measurements  on  samples  of  the  Balti- 
more gneiss  are  given  in  table  3.  The  Bal- 
timore gneiss  underlies  all  the  metasedi- 
mentary  rocks  in  the  vicinity  of  Balti- 
more and  has  generally  been  regarded  as 


TABLE  1.   Mineral  Ages  of  Precambrian  Gneisses 

Age,  million  years 














Bear  Mountain,  N.  Y. 

Canada  Hill  gneiss 



Storm  King  granite 



Shenandoah  National  Park, 



Mary's  Rock  Tunnel  gneiss 


Hibernia,  N.  J. 

gneiss  (dark) 


gneiss  (light) 


1140        1150        1170        1030 
960         990        1060         850 


1100        1150        1110 






TABLE  2.    Mineral  Ages  from  the  Baltimore  Gneiss 


Age,  million  years 













Baltimore  area 

Towson  dome 








Phoenix  dome 







Woodstock  dome 



Philadelphia   area 

Spring  Mill,  Pa. 








TABLE  3.     Ages  of  Metamorphic  and  Igneous  Rocks  in  the  Washington-Baltimore  Area 













Kensington   gneiss 
Sample  A 







Sample  B 







Woodstock  granite 
Baltimore  gneiss 
Towson  dome 




Woodstock  dome 



Phoenix  dome 




of  Precambrian  age.  Wasserburg,  Lipson, 
and  Pettijohn  have  recently  shown,  how- 
ever, that  the  K-A  ages  of  the  metamorphic 
mica  in  these  overlying  sediments,  as  well 
as  that  in  the  Baltimore  gneiss  itself,  are 
Paleozoic.  Thus  the  metamorphosis  of  the 
gneiss  was  Paleozoic,  and  its  Precambrian 
age  cannot  be  inferred  from  its  degree  of 
metamorphism.  Consequently,  the  ques- 
tion of  the  age  of  the  gneiss  was  reopened. 
As  is  shown  in  table  2,  zircon  and  feldspar 
of  Grenville  age  have  now  been  found  in 
the  gneiss.  Petrographic  examination  of 
the  rock  indicates  that  it  is  improbable 
that  the  feldspar  is  of  detrital  origin.  In 
agreement  with  the  findings  of  Wasser- 
burg et  al.,  the  mica  appears  to  have  been 
recrystallized  during  the  metamorphism. 
The  Rb-Sr  ages  of  micas  from  the  Balti- 
more gneiss  near  Baltimore  are  compared 
with  those  of  the  intrusive  Woodstock 
granite  and  the  concordant  Kensington 
granite  gneiss  in  table  3.  All  the  ages  are 
in  agreement  within  their  experimental 
errors,  and  indicate  that  the  metamorphism 
in  this  area  took  place  310  ±10  million 
years  ago.  The  Baltimore  gneiss  at  Phila- 
delphia occupies  a  similar  stratigraphic 
position,  but  the  correlation  of  the  sedi- 
ments and  gneiss  between  the  two  areas  is 
obscure.  The  mica  from  the  Philadelphia 
sample  does  not  seem  to  have  been  so 
completely  recrystallized  as  that  in  the 
Baltimore  samples.  Again  the  zircons  in- 
dicate a  Grenville  age. 

It  is  interesting  to  note  that  the  K-A 
age  of  the  mica  from  the  Philadelphia 
sample  is  much  greater  than  its  Rb-Sr  age. 
The  same  phenomenon  is  observed  to  a 
lesser  degree  in  the  Kensington  granite 
gneiss  and  the  Towson  dome  samples. 
This  effect  has  previously  been  found  in 
metamorphic  rocks  near  Sudbury,  Ontario, 
as  was  reported  by  this  group  in  the  previ- 
ous annual  report  and  was  also  found  by 
the  geochronology  group  at  Massachusetts 
Institute  of  Technology.  It  seems  to  be 
peculiar  to  metamorphic  rocks,  since,  in 
the  fifty  or  more  cases  measured  in  this 
laboratory    where   Rb-Sr    and    K-A    ages 

have  been  compared  on  the  same  mica 
sample  from  igneous  rocks,  the  K-A  age  is 
invariably  younger.  It  may  be  that  during 
the  recrystallization  of  the  metamorphic 
rocks  the  old  radiogenic  argon  is  not  com- 
pletely excluded  from  the  mica. 

Therefore,  it  appears  that  the  gneisses 
listed  in  tables  1  and  2  are  to  be  grouped 
with  the  Grenville  rocks  farther  north. 
As  is  indicated  in  figure  7,  this  grouping 
extends  the  Grenville  orogeny  a  consider- 
able distance  to  the  south  of  the  original 


Recent  measurements  reported  by  the 
Massachusetts  Institute  of  Technology  and 
the  University  of  Minnesota  groups  now 
extend  the  1000-million-year-old  rocks  up 
into  northern  Quebec  and  Newfoundland, 
thus  fulfilling  the  expectation  that  a  long 
linear  belt  of  these  rocks  would  be  found. 

Age  Patterns  in  Zones  of  Regional 
Metam  orph  ism 

The  rocks  from  the  metamorphosed 
zones  in  Iron  County,  Michigan,  are  all 
part  of  the  Michigamme  Formation,  which 
is  a  slate  placed  in  the  middle  Precambrian 
section.  The  Mary  Lake  granite  intrudes 
the  Michigamme  slate,  and  its  relationship 
to  the  metamorphic  zones  of  the  slate  is 
such  that  it  is  indicated  as  the  source  of 
energy  for  the  regional  metamorphism. 
The  mineral  age  measurements  obtained 
on  mica  from  this  granite  and  the  various 
metamorphic  zones  are  given  in  table  4. 
All  the  Rb-Sr  ages  agree  within  the  error 
of  measurement.  The  K-A  ages  are  in 
general  agreement  with  the  Rb-Sr  ages  but 
consistently  lower.  The  muscovite  in  the 
sillimanite  zone  is  doubtless  secondary, 
and  its  K-A  age  is  indicative  of  this  fact. 

It  is  concluded  that  the  mineral  age 
measurements  are  not  in  disagreement 
with  the  field  evidence  of  the  regional 
metamorphic  pattern.  The  pegmatites  in 
neighboring  Dickinson  County  intrude  a 
pre-Huronian  granitic  series,  and  it  had 
previously  been  assumed  that  the  meta- 
morphism from  the  granitic  intrusion  and 


the  intrusion  of  the  pegmatites  were  simul- 
taneous. These  measurements  show  con- 
clusively that  at  least  300  million  years 
separated  the  two  events. 

obtained  at  our  laboratory  are  given  by 
age  in  table  5.  The  1000-million-year-old 
rocks  have  been  discussed  more  completely 
above.    Additional  locations  of  2600-mil- 

TABLE  4.    Ages  of  Micas  from  Northern  Michigan 

Age,  million  years 





Iron  County,  metamorphic  zones 

Mary  Lake  granite 




Sillimanite   Zone    (Peavy   Falls) 




Sillimanite   Zone    (Peavy   Falls) 



Garnet  Zone  (Horserace  Rapids) 




Biotite  Zone   (Cedar  Falls) 




Republic,  Marquette  County 





Felch,   Dickinson 




Felch,   Dickinson 





TABLE  5.    Ages  of  Micas 

Age,  million 






2600  m.y.  ages 

Kirkland  Lake,  Ont. 




Timmins,  Ont. 




Hearst,  Ont. 




East  of  Kenora,  Ont.* 



Winnipeg  River,  Man. 




International  Falls,  Minn.* 

Gneiss  inclusion 



Bonneville,  Wyo. 




1350  m.y.  ages  (1957-58) 

Mary  Lake,  Iron  Co.,  Mich. 




Menominee,   Wise. 




Frederickstown,  Mo. 




Decaturville,  Mo. 



Troy,  Okla. 



Miscellaneous    (1957-58) 

Sioux  Lookout,  Ont. 



Wichita  Mts.,  Okla. 




Death  Valley,  Calif.f 




Dayton  Bend,  N.  C. 



*  Analyzed  1957-1958. 

fin  collaboration  with  G.  J.  Wasserburg. 

This  year's  work  has  increased  the  in- 
formation available  on  the  geographic  ex- 
tent of  the  2600-million-,  1350-million-,  and 
1000-million-year  orogenies  by  almost  a 
factor  of  2.  The  data  for  the  map  of  figure 
7  for  the  first  two  groups  which  have  been 

lion-year  rocks  measured  this  year  are 
shown  with  our  previous  measurements  in 
the  table.  The  work  of  P.  W.  Gast  at  the 
Lamont  Geological  Observatory  has  shown 
the  general  occurrence  of  rocks  of  this  age 
both  in  the  Winnipeg  River  area  in  west- 


ern  Manitoba  and  in  the  Big  Horn  Moun- 
tains of  Wyoming.  Goldich,  Baadsgaard, 
and  Nier  at  the  University  of  Minnesota 
are  responsible  for  most  of  the  data  on 
Minnesota  rocks.  Hurley,  Fairbairn,  and 
Pinson  have  provided  most  of  the  ages  in 
Ontario  between  Sudbury  and  Sioux  Look- 
out. The  western  Quebec  data  are  from 
the  University  of  Toronto. 

Data  on  eleven  1350-million-year-old 
granitic  rocks  in  western  United  States 
were  summarized  last  year.  It  is  seen  in 
the  tables  that  orogenies  in  this  period  left 
their  imprint  in  Oklahoma,  Missouri,  Wis- 
consin, Michigan,  and  southern  Ontario. 
The  long-known  uranium  ore  of  this  age 
at  Great  Bear  Lake,  N.W.T.,  Canada,  is 
further  demonstration  of  the  common  oc- 
currence of  minerals  of  this  age.  The  older 
age  of  the  Death  Valley  sample,  measured 
in  collaboration  with  Wasserburg,  indi- 
cates that  the  area  included  in  this  orogeny 
is  bounded  by  older  Precambrian  rocks  to 
the  west. 

Additional  rocks  and  minerals  of  the 
Cutler  batholith  that  have  been  analyzed 
are  shown  in  table  6,  together  with  those 
previously  obtained.  That  there  were  two 
important  periods  in  the  history  of  this 
rock  unit  seems  obvious,  but  there  remain 
the  little-understood  discrepancies  in  the 
K-A  and  Rb-Sr  ages  of  the  same  mineral, 
which  will  preclude  any  complete  inter- 
pretation of  the  data  until  the  effects  of 
alteration  on  mineral  ages  are  better  under- 

The  Department  of  Terrestrial  Magnet- 
ism members  of  this  mineral  age  group 
took  part  in  the  Carnegie  Andes  Seismic 

Expedition  during  July-October  1957,  and 
collected  a  small  number  of  rocks  in  Peru. 
The  rocks  analyzed  had  been  mapped  as 
pre-Cretaceous,  and  from  the  data  of 
table  7   (a  middle-Cretaceous   age  of  100 

TABLE  6.     Summary  of  Age  Determination 
Work  on  the  Cutler  Batholith 



Age,  million 



Large  pegmatite 







Mica  schist  in  con- 

tact with  above 


Small  pegmatite 




Granite  1 





Granite  2 




TABLE  7. 

Ages  of  Peruvian  Micas 


Age,  million  years 
K-A          Rb-Sr 

Macchu  Picchu,  Dept. 

Cusco,   Peru 
Mollenda,  Dept.  Arequipa, 

Atica,  Dept.  Arequipa, 


410±30  400±40 
330  300±100 

million  years  was  measured  also  on  a  peg- 
matite from  Pala,  California)  one  would 
conclude  first  that  they  are  indeed  pre- 
Cretaceous  as  mapped  and  secondly  that 
they  are  not  Precambrian.  Other  samples 
obtained  from  Chile  and  Brazil  will  be 
analyzed  in  the  coming  year. 


S.  E.  Forbush 


As  was  stated  in  last  year's  report  the 
investigation  of  the  equatorial  electro  jet  is 
being  carried  out  as  part  of  the  United 
States  program  for  the  International  Geo- 
physical Year  (IGY)  through  grants  ap- 

proved by  the  IGY  Panel  on  Geomagnet- 
ism. The  four  Askania  variographs 
loaned  to  the  Department  of  Terrestrial 
Magnetism  through  the  cooperation  of  the 
U.  S.  Coast  and  Geodetic  Survey  are  being 
operated   on   the   west  coast   of  Peru   at 



Talara,  Chiclayo,  Chimbote,  and  Yauca, 
which  extend  from  geographic  latitude 
4°  N  to  15°  S. 

These  four  temporary  magnetic  observa- 
tories are  operated  and  maintained  through 
the  cooperation  of  the  John  A.  Fleming 
Observatory  of  the  Instituto  Geofisico  de 
Huancayo,  which  also  provides  magnetic 
variation  data  from  Huancayo.  The  De- 
partment of  Terrestrial  Magnetism  has  co- 
operated with  the  Instituto  Geofisico  de 
Huancayo  in  the  establishment  of  a  per- 
manent magnetic  observatory  for  the  Uni- 
versity of  Arequipa  (latitude  17°  S)  by 
providing  a  la  Cour  vertical  intensity  vari- 
ometer. This  observatory  is  expected  to 
be  in  operation  by  July  1958,  when  its 
records  will  become  available  for  investi- 
gation of  electrojet  effects. 

Owing  to  delay  in  procurement  and  to 
serious  defects  in  instruments  rather  few 
simultaneous  and  complete  daily  records 
have  been  obtained  at  the  four  variograph 
stations.  Consequently  the  few  magnetic- 
storm  sudden  commencements  which  oc- 
curred near  midday  were  too  inadequately 
recorded  to  indicate  whether  they  were 
influenced  by  electrojet  effects.  Most  of 
the  instrumental  defects  have  now  been 
corrected,  and  it  is  expected  that  by  July 
1958  satisfactory  records  will  be  available 
from  the  four  Askania  variographs.  These, 
with  records  from  Huancayo  and  the  new 
Arequipa  Observatory,  should  provide  rec- 
ords from  a  total  of  six  stations. 

From  results  obtained  on  the  prelimi- 
nary survey  mentioned  in  last  year's  report 
the  maximum  diurnal  variation  in  H  was 
found  within  1°  in  latitude  from  Huan- 
cayo (12°  S)  in  accord  with  results  ob- 
tained by  A.  A.  Giesecke,  Jr.,  Director  of 
the  Instituto  Geofisico  de  Huancayo,  with 
QHM's  in  1949.  This  result  is  also  con- 
firmed by  the  fact  that  the  diurnal  varia- 
tion in  Z  from  the  1957  survey  changes 
sign  at  a  latitude  within  1°  of  Huancayo. 
The  maximum  diurnal  variation  ampli- 
tude in  Z  occurred  at  about  latitude  16°  S 
and  the  minimum  near  8°  or  9°  S. 

The  latitudinal  distance  of  500  miles  be- 

tween these  extremes  provides  a  rough 
estimate  of  the  width  of  the  electrojet  band. 
The  variation  with  latitude  of  the  ampli- 
tude of  the  diurnal  variation  of  Z  indi- 
cates the  importance  of  the  return  currents 
(from  the  main  eastward  current  band) 
which  must  flow  westward,  north  and 
south  of  the  electrojet. 

To  effect  reliable  estimates  of  the  width 
and  height  of  the  electrojet  will  require 
derivation  of  the  electric  current  system 
for  the  electrojet  field,  after  deducting  the 
field  of  the  "normal"  Sq  diurnal  variation 
estimated  from  observatory  data  at  loca- 
tions free  from  the  influence  of  the  electro- 

A  preliminary  examination  of  records 
indicates  that  during  disturbed  periods 
many  of  the  fluctuations  in  Z  have  the 
largest  absolute  magnitude  at  the  latitude 
where  the  largest  amplitude  of  Sq  in  Z 
was  observed.  The  sign  of  these  fluctua- 
tions in  Z  is  opposite  at  these  two  latitudes, 
thus  indicating  electrojet  effects. 


Cosmic-ray  variations  and  solar  activity. 
During  the  period  1937  to  1957  two  en- 
tire cycles  of  solar  activity  were  com- 
pleted, and  in  1957  there  occurred  the 
largest  sunspot  numbers  on  record.  Fig- 
ure 8  shows  the  annual  means  of  sunspot 
numbers  and  those  for  cosmic-ray  in- 
tensity from  Huancayo  and  from  Chelten- 
ham (Fredericksburg)  for  this  interval. 
The  years  of  maximum  cosmic-ray  in- 
tensity occur  near  those  for  minimum 
sunspot  numbers.  Moreover,  the  minima 
of  cosmic-ray  intensity  in  1947  and  1957 
were  lower  than  in  1937,  a  finding  in  ac- 
cord with  the  fact  that  sunspot  numbers 
were  greater  in  1947  and  1957  than  in 
1937.  The  decrease  in  cosmic-ray  intensity 
between  1955  and  1957  evidently  lags  a 
year  or  so  behind  the  increase  in  sunspot 
numbers  in  that  period.  Figure  9  shows 
that  the  decrease  in  cosmic-ray  intensity  at 
Huancayo  betwen  1954  and  1957  also 
lagged  behind  the  decrease  recorded  by 
D.  C.  Rose  with  a  neutron  monitor   at 


Ottawa  and  behind  that  observed  by  H.  V. 
Neher  in  balloons  at  Thule.  Near  sunspot 
minimum  in  1954  Neher  found,  near  the 
top  of  the  atmosphere  at  Thule,  that  there 
was  no  evidence  for  the  exclusion  of  low- 
energy  primary  particles — at  least  not  for 
protons  with  energy  50  Mev  or  more. 
When  solar  activity  increased  after  1954 
the  primaries  near  the  low-energy  part  of 
the  cosmic-ray  spectrum  were  first  ex- 
cluded; later  as  the  solar  activity  cycle  pro- 

barrier  appears  to  be  created  which  ef- 
fectively prevents  primary  particles  from 
our  galaxy  from  reaching  the  earth.  This 
barrier  most  likely  comprises  plasma 
clouds  with  magnetic  fields  ejected  from 
the  sun.  Such  individual  plasma  clouds 
must  account  for  magnetic  storms  and, 
through  magnetohydrodynamic  effects,  for 
large  sudden  decreases  in  cosmic-ray  in- 
tensity that  sometimes  occur  during  mag- 
netic storms.  Analysis  of  IGY  cosmic-ray 

-10                                                      /           1 

,                     0- 

'1       x        1 

L'  \ 

o/ol       \  1 


/'         J              \\ 

o           f/o 

\               35- 

/ay        /                           \ 


-o                 /'     \°/                          I 

i  J 

\  X  7°- 

0      °///                                            \\ 

i\]     o 

i     \      SS 

/                                                     \  * 



/        o 

\    \        105- 

°                                                 1 

\             1        1 

»     1 

C-R                                                                        1 

J        J 

•     1 

IN  %                                                                         \ 

\         /      ,<* 

'    1 

-10                                                                        1 

\  y 

'.     1      140- 

SS                                                ' 

\  wi 

\    1 


\  \ 


\     /  o 


1                  O       C-R  AT  CHELTENHAM 

\     l75~ 


1937                                1942 




Fig.  8.    Annual  means  cosmic-ray  intensity  (C-R)  and  sunspot  numbers  (SS). 

gressed  toward  a  maximum,  in  1957,  pri- 
mary particles  with  greater  and  greater 
energy  appeared  to  be  excluded.  This  ex- 
clusion accounts  for  most  of  the  decrease 
in  high  latitudes,  as  in  figure  9  for  Thule 
and  Ottawa.  The  decrease  in  intensity  at 
Huancayo  which  started  in  1956  indicates 
a  reduction  in  the  number  of  primary  pro- 
tons even  for  energies  above  at  least  15 
bev.  Neher's  results  at  Thule  showed 
there  was,  in  1954,  no  "knee"  in  the  varia- 
tion of  intensity  with  latitude,  near  the  top 
of  the  atmosphere.  The  work  of  several 
investigators  shows  that  by  1957  the  "knee" 
was  near  latitude  48°  geomagnetic. 
Thus  with   increasing  solar   activity   a 

data  from  several  latitudes  should  pro- 
vide information  on  the  change  in  energy 
spectrum  of  primary  particles  during  the 
solar  cycle  and  also  during  transient  de- 
creases which  should  lead  to  a  better  un- 
derstanding of  the  mechanism  responsible 
for  these  changes  in  cosmic-ray  intensity. 
This  in  turn  is  certain  to  reveal  important 
changes  in  electromagnetic  conditions  in 
the  solar  system  that  could  not  be  revealed 
in  any  other  way. 

The  change,  with  solar  cycle,  of  the  vari- 
ability of  cosmic-ray  intensity  is  shown  in 
figure  10  by  the  yearly  pooled  standard 
deviation  of  daily  means  from  monthly 
means.  Near  the  sunspot  minima  in  1944 



and  1954  the  standard  deviations  were  also 
minimal.  These  minimal  values  are  at 
about  the  noise  level  of  the  instrument.  In 
figure  10  it  is  seen  that  the  standard  devia- 
tion is  decidedly  larger  in  1957  than  in 
1937,  in  accord  with  the  much  larger  sun- 
spot  maximum  in  1957  than  in  1937. 

sure  have  been  completed  for  Huancayo 
and  Fredericksburg  through  April  1958. 
From  the  records  scaled  at  Christchurch 
the  reduction  of  daily  means  has  been  ef- 
fected through  June  1957.  To  cooperate  in 
the  United  States  program  for  cosmic-ray 
research  in  the  International  Geophysical 

JAN  1955        JAN   1956 

JAN  1957 

5        o 


oo    o\ 




o    o 

D          0     \ 












o\              <&     O 
\            ° 





5  1- 


o  o\     /—\  1  \ 










0                     \ 









15    -1 

1  °      Ld 


37.5  S 


o  V «« 

o         o-° 




(ROSE)  MONTHLY           < 

A                         ^ 

MEANS                           ° 






\            ' — ' 







o        ®\       3~ 




Fig.  9.    Neutron  intensity  (N)  at  Ottawa,  ionization  (/)  at  Huancayo,  and  ionization  under  15 
g  cm-2  at  Thule  (T). 

Old  cosmic-ray  program.  Compton- 
Bennett  meters  were  satisfactorily  operated 
throughout  the  report  year  at  Godhavn 
(Greenland),  Climax  (Colorado,  U.  S.), 
Ciudad  Universitaria  (Mexico,  D.  F.), 
Huancayo  (Peru),  Christchurch  (New 
Zealand),  and  Fredericksburg  (Virginia, 

The  scaling  and  reduction  of  records  in- 
cluding the  tabulation  of  bihourly  means 
of  ionization  corrected  for  barometric  pres- 

Year,  tabulations  of  corrected  bihourly 
means  of  cosmic-ray  intensity  have  been 
forwarded  to  the  four  IGY  World  Data 

It  was  originally  planned  to  cooperate 
with  the  U.  S.  IGY  cosmic-ray  program  by 
making  available  to  World  Data  Centers 
(WDCs)  only  the  tabulations  of  corrected 
bihourly  values  of  cosmic-ray  intensity 
from  Huancayo  and  Fredericksburg.  The 
results  at  Christchurch  were  to  be  made 



available  to  WDCs  from  New  Zealand. 
The  stations  at  Godhavn,  Climax,  Mexico, 
and  Derwood  (large  ionization  chamber) 
were  to  be  operated  for  solar-flare  patrol 
observations  and  unusual  magnetic  storm 

The  U.  S.  IGY  Technical  Panel  for 
Cosmic  Rays  emphasized  the  importance 
of  results  from  the  program  for  continu- 

the  reduction  of  as  many  as  possible  of 
these  data. 

On  the  basis  of  the  recommendation  of 
the  Panel  and  the  approval  of  the  U.  S. 
IGY  National  Committee  a  modest  grant 
was  made,  to  provide  the  necessary  addi- 
tional assistance  for  the  reduction  of  the 

Large   ionization   chamber.    The   large 

II  0 



o         S  FOR  ALL  EXCEPT 


//       l5°" 


\  / 

"" ! | 


D     '\ 



l\                                                         J 



-0.6     1 


/    \ 


1         100- 

S      / 

o   \  \                                                           1 

0  \ 


IN   %/ 


i       SS 

\     v                                                     1  ' 


\                    vv 

/        NOS 

\       \                                            1  / 

0  \            / 

i        o\                               1   j 



\  \        ;/ 



.     '9,42   °       o 


I        i 

1952         V 

i.i        i     . 


Fig.  10.    Yearly  pooled  standard  deviation  (S)  of  daily  means  of  cosmic-ray  intensity  at  Huan- 
cayo,  and  yearly  mean  sunspot  numbers  (SS). 

ous  registration  of  cosmic-ray  intensity, 
begun  by  the  Carnegie  Institution  of  Wash- 
ington between  1936  and  1938,  and  noted 
the  fact  that  this  program  had  provided 
the  only  continuous  record  of  cosmic-ray 
variations  over  so  long  a  period. 

The  Panel  therefore  urged  that  the  data 
from  Godhavn  and  Ciudad  Universitaria 
(Mexico  City)  be  completely  reduced  for 
the  duration  of  the  International  Geo- 
physical Year  (July  1957-December  1958). 
The  data  from  Godhavn  for  the  period 
January  1951  to  July  1957  not  having  been 
reduced,  the  Cosmic  Ray  Panel  also  urged 

cosmic-ray  ionization  chamber  was  main- 
tained in  essentially  continuous  operation 
at  Derwood  during  the  report  year.  No 
solar-flare  effects  have  been  observed  since 
February  23,  1956. 

Cooperation  in  operation  of  cosmic-ray 
meters.  The  successful  operation  of 
Compton-Bennett  cosmic-ray  meters  over 
a  long  period  at  so  many  stations  has  been 
possible  only  through  the  wholehearted 
and  unselfish  cooperation  of  several  or- 
ganizations and  individuals.  We  wish  to 
express  our  appreciation  to  the  following 
organizations  for  the  operation  and  main- 



tenance  of  cosmic-ray  meters:  the  Danish 
Meteorological  Institute  and  the  staff  of 
its  Godhavn  Magnetic  Observatory  at 
Godhavn,  Greenland;  the  U.  S.  Coast  and 
Geodetic  Survey  and  the  staff  of  its  mag- 
netic observatory  at  Fredericksburg,  Vir- 
ginia; the  High  Altitude  Observatory  of 
the  University  of  Colorado  and  its  staff  at 
Climax,  Colorado;  the  Instituto  Nacional 

de  la  Investigacion  Cientifica  and  the  Uni- 
versidad  de  Mexico,  Mexico,  D.  F.;  the 
Government  of  Peru  and  the  staff  of  its 
Instituto  Geofisico  de  Huancayo  for  mak- 
ing available  the  Compton-Bennett  records 
from  Huancayo;  and  the  Department  of 
Scientific  and  Industrial  Research  and  the 
staff  of  its  Magnetic  Observatory  at  Christ- 
church,  New  Zealand. 


N.  P.  Heydenburg  and  G.  M.  Temmer 

During  the  past  year,  we  have  once 
again  returned  to  the  domain  of  the  lighter 
nuclei  and  "proper"  nuclear  reactions,  after 
a  most  unusual  and  rewarding  interlude 
of  about  three  years  spent  with  Cou- 
lomb excitation  of  most  elements  up 
through  uranium.  Although  a  few  groups 
still  continue  to  follow  up  more  detailed 
problems  raised  in  this  electric  excitation 
of  nuclei,  we  felt  that  within  the  means 
at  our  disposal,  and  within  the  limitations 
set  by  our  own  interests,  we  had  come  to 
a  definite  stopping  point.  Looking  back, 
we  have  measured  the  energies  and  abso- 
lute transition  probabilities  (lifetimes)  of 
about  200  electromagnetic  transitions  in 
some  120  nuclear  species,  employing  about 
75  enriched  isotopic  targets  for  definite 
assignments  and  greater  accuracy.  About 
one-third  of  these  transitions  represented 
previously  unknown  cases,  and  essentially 
none  of  the  transition  rates  had  been  meas- 
ured before.  It  is  clear,  therefore,  that  we 
gained  a  great  deal  of  insight  into  the 
systematic  behavior  of  these  transitions, 
especially  in  the  even-proton-even-neutron 
nuclei,  where  without  exception  the  first- 
excited  state  of  spin  parity  2+  was  excited. 
Very  clear  evidence  was  obtained  for  the 
existence  of  rotational  states  in  both  odd 
and  even  heavy  deformed  nuclei  beyond 
europium  as  well  as  for  the  vibrational 
states  in  the  lighter  nuclei  with  spherical 
equilibrium  shapes.  In  fact,  the  very  sharp 
line  of  demarcation  between  them  was  lo- 
cated within  two  neutron  numbers   (151 

and  153).  The  details  of  most  of  these 
developments  have  been  discussed  in  the 
last  three  annual  reports.  With  the  study 
of  one  of  the  last  (and  most  exotic)  ele- 
ments, to  be  discussed  below,  we  have 
essentially  completed  our  work  in  Cou- 
lomb excitation  of  nuclei  with  a  particles 
up  to  7  Mev.  Some  of  the  future  develop- 
ments in  this  field  seem  to  lie  with  the 
heavy-ion  accelerators  such  as  have  begun 
to  operate  at  Berkeley  and  at  Yale,  heavy 
ions,  because  of  their  higher  charge,  per- 
mitting the  bombardment  of  lighter  ele- 
ments without  danger  of  purely  nuclear 

Coulomb  Excitation  of  Xenon 

In  our  early  results  on  the  Coulomb 
excitation  of  normal  xenon,  the  isotopic 
assignment  of  the  observed  y  rays  was  not 
clear  on  account  of  the  many  isotopes  pres- 
ent in  normal  xenon.  Enriched  isotopic 
samples,  prepared  by  thermal  diffusion  at 
Yale,  were  made  available  to  us,  and  in  co- 
operation with  G.  F.  Pieper  and  C.  E. 
Anderson,  guest  investigators  from  Yale, 
the  y-ray  spectra  of  these  gas  samples  were 
observed  when  Coulomb-excited  by  6.6- 
Mev  a  particles.  The  samples  were  bom- 
barded in  a  small  chamber  separated  from 
the  main  accelerating  tube  by  a  thin  nickel 
window.  The  y  rays  were  observed  with  a 
sodium  iodide  detector  and  80-channel 
pulse  height  analyzer.6  By  comparing  rela- 

6  This  analyzer,  which  incorporates  a  quartz 
delay  line  memory  system,  was  constructed  by 
Mr.  Buynitzky  under  the  guidance  of  G.  F. 
Pieper.  It  has  operated  very  well  during  the  past 


tive  y-ray  intensities  in  the  two  enriched 
samples  and  normal  xenon,  the  observed  y 
rays  at  energies  670,  535,  and  445  kev 
could  be  assigned  to  the  even-proton,  even- 
neutron  isotopes  Xe132,  Xe130,  and  Xe128, 
respectively.  These  y  rays  appear  in  the 
de-excitation  of  the  first  T  levels  which 
are  excited  in  the  Coulomb  excitation  proc- 
ess. Their  energies  and  intensities  vary  in 
a  systematic  way,  characteristic  of  the  ap- 
proach to  a  closed  shell,  in  this  case  to  the 
closed  shell  at  neutron  number  82.  Be- 
cause of  a  slight  nitrogen  contamination  in 
the  samples  a  y  ray  appeared  at  875  kev, 
due  to  the  intense  reaction  N14(a,  py)  O17. 
The  presence  of  a  possible  y  ray  due  to 
Xe134  which  should  occur  in  this  energy 
region  was  therefore  difficult  to  determine. 
Gamma  rays  were  also  observed  at  290 
and  365  kev.  The  365-kev  y  was  assigned 
to  Xe131  and  is  due  to  the  de-excitation  of 
a  known  level  in  Xe131  at  365  kev.  A 
cascade  transition  is  also  known  to  occur 
from  this  level  through  a  level  at  80  kev. 
The  observed  peak  at  290  kev  is  due  in 
part  to  the  cascade  y  ray  at  295  kev.  Most 
of  the  intensity  at  290  kev,  however,  is  due 
to  Xe129.  A  y  ray  of  this  energy  has  not 
previously  been  reported  in  Xe129. 

Level  Structure  of  Na22 

Over  the  last  few  years  the  interest  in 
the  lighter  nuclei  has  gradually  shifted 
from  the  discovery  of  energy  levels  and 
their  spin  and  parity  assignments  to  at- 
tempts to  measure  their  transition  rates  or 
at  least  relative  decay  probabilities.  That 
almost  all  energy  levels  and  many  of  their 
characteristics  have  been  discovered  by 
now  is  not  too  surprising  in  view  of  the 
limited  number  of  light  nuclei  (about  20 
species)  and  the  large  number  of  Van  de 
Graaff  generators  in  the  world  (about  50) 
capable  of  exciting  them.  One  of  the  re- 
maining fruitful  frontiers  is  the  measure- 
ment of  level  widths  (transition  probabili- 

year,  and  has  made  possible  many  observations 
that  could  not  have  been  made  with  a  single- 
channel  analyzer. 

ties).  This  approach  is  complementary  to 
more  recent  theoretical  progress  in  obtain- 
ing nuclear  wave  functions  on  the  basis 
of  various  models  and  hence  making  it 
possible  to  calculate  transition  rates  be- 
tween identified  nuclear  states.  As  this 
particular  parameter  is  sensitive  to  the 
nuclear  models  used  it  gives  promise  of 
being  able  to  choose  between  them. 

Our  instrumentation  for  the  detection 
and  analysis  of  pulse-height  distributions 
had  been  greatly  expanded  during  the  pre- 
vious report  year.  In  addition,  we  built  a 
fast  coincidence  system  with  a  resolving 
time  of  about  20  millimicroseconds  capa- 
ble of  gating  the  multichannel  device  with 
a  movable  single  channel.  It  allows  us  to 
find  out,  for  instance,  what  parts  of  a  y-ray 
spectrum  excited  in  a  nuclear  reaction  are 
in  either  prompt  or  delayed  coincidence 
with  a  particular  transition  selected  in  the 
single  channel.  We  spent  considerable 
time  improving  the  over-all  circuitry,  es- 
pecially on  the  time  compensation  part, 
the  device  allowing  the  simultaneous 
prompt  coincidence  of  all  parts  of  the  ex- 
amined spectrum  regardless  of  pulse 
height.  The  usefulness  of  the  entire  set-up 
of  course  depends  upon  the  proper  func- 
tioning of  this  component. 

One  of  the  first  problems  we  attacked 
with  our  new  instrumentation  was  one  we 
had  encountered  some  four  years  ago  in 
the  nucleus  Na22.  The  Na22  nucleus  is  of 
the  "self-conjugate,"  odd-odd  type:  it  has 
equal  (odd)  numbers  of  protons  and  neu- 
trons. Such  a  nucleus  has  two  even-even 
isobaric  neighbors,  in  this  case  Ne22  and 
Mg22,  the  three  nuclei  forming  what  is 
known  as  an  "isobaric  triplet."  In  analogy 
with  the  ordinary  spin  in  spectroscopy,  an 
"isobaric  spin"  T  is  introduced,  which 
takes  the  value  0  if  only  the  central  mem- 
ber of  the  triplet  shows  a  certain  configura- 
tion, or  the  value  1  if  corresponding  con- 
figurations occur  in  all  three  members  of 
the  triplet,  the  isobaric  spin  "projection" 
taking  on  values  1,  0,  and  —1.  The  ex- 
istence of  the  T=l  variety  of  multiplets 


is  evidence  for  the  charge  independence  of 
nuclear  forces,  i.e.  the  basic  identity  of 
neutron-neutron,  proton-proton,  and  neu- 
tron-proton forces.  Particular  interest  at- 
taches to  the  lowest  excited  state  of  the 
central  member  having  T=l  (Na22),  since 
this  corresponds  to  the  ground  states  of  the 
two  even-even  neighbors.  It  must  there- 
fore necessarily  have  spin  parity  0\  Now 
the  ground  state  of  Na22  has  a  measured 
spin  of  3+;  therefore,  if  the  first-excited 
state  of  Na22  at  593  kev,  which  we  dis- 
covered several  years  ago,  were  the  analog 
state  in  question,  its  de-excitation  would 
involve  a  spin  change  of  three  units  with 
no  change  in  parity,  that  is,  a  magnetic 
octupole  transition.  Such  a  transition 
should  have  a  lifetime  of  about  0.1  second 

or  longer. 

Upon  re-examining  the  reaction  F19 
(a,  #y)Na22,  one  of  the  first  facts  we  dis- 
covered was  that  the  lifetime  of  that 
transition  was  only  0.266  microsecond,  thus 
ruling  out  spin  0+  and  hence  the  T=l 
character  for  the  593-kev  state.  We  further 
found  that  the  593-kev  radiation  was  in 
coincidence  with  a  low-energy  y  ray  of 
73  kev  which  had  to  come  from  a  state  at 
666  kev.  Figure  11  shows  both  the  coinci- 
dent pulse-height  distribution  when  the 
single  channel  was  placed  on  the  peak  cor- 
responding to  593  kev,  revealing  a  pro- 
nounced peak  at  73  kev,  and  the  coinci- 
dence rate  as  a  function  of  artificial  delay 
introduced  into  the  73-kev  detector,  show- 
ing the  exponential  decay  of  the  593-kev 
state  with  a  half-life  of  0.266  microsecond. 
Several  pieces  of  evidence  point  to  this 
state  at  666  kev  as  the  one  we  had  been 
looking  for,  not  the  least  important  of 
which  is  the  remarkable  agreement  of  its 
location  with  that  calculated  from  the 

In  the  familiar  series  of  mirror  nuclei 
known  as  Wigner  elements  the  energy 
difference  between  adjacent  isobars  can  be 
accounted  for  entirely  by  the  Coulomb 
energy  difference  due  to  the  one  addi- 
tional proton.  In  fact,  one  of  the  classic 
methods  for  the  determination  of  nuclear 

radii  is  based  on  the  observed  energy  dif- 
ferences. Now  for  a  self-mirrored  nucleus 
such  as  Na22  one  evidently  has  to  use  the 
difference  between  the  analogous  0+  states 
of  neighboring  isobars,  the  ground  state 
of  Ne22  and  the  newly  discovered  state 
of  Na22  at  666  kev  in  our  case.  In  figure  12 
the  systematic  plot  of  Coulomb  energy 
differences  as  a  function  of  A2/s  for  these 





0.2         0.3         0.4         0.5         06 



i              i              i              1              1 
593  Kev  state  of  Na22 




ti/2     -0.2  7  ±  .01  /x  sec. 





^73  Kev             ^~"\^ 


XI            In  coincidence  with 



/     \                 593  Kev  y 










V*               .       *.*.... 





1     .. 

109  Kev            197  Kev 

1                  T                    1          ,         A                  1 

20         30  40 

Channel  number 




Fig.  11.  (A)  Coincidence  rate  vs.  artificial 
delay  introduced  in  the  73-kev  channel,  showing 
a  0.266-microsecond  half-life  for  the  593-kev 
state  of  Na22.  (B)  Pulse-height  spectrum  in  coin- 
cidence with  593-kev  radiation  in  the  single- 
channel  gate,  showing  pronounced  73-kev  peak. 
It  is  this  peak  height  as  a  function  of  introduced 
delay  that  is  plotted  in  (A). 

self-conjugate  nuclei  between  major  shells 
shows  excellent  agreement  for  Na22.  The 
straight-line  relationship  indicates  the  de- 
pendence of  the  nuclear  radius  upon  the 
cube  root  of  the  atomic  weight. 

We  also  excited  Na22  in  another  manner, 
by  bombarding  Ne20  with  He3  ions,  induc- 
ing the  Ne20(He3,  ^)Na22  reaction,  and 
detecting  the  emerging  protons  in  a  thin 
scintillator  in  coincidence  with  y  radiation. 
A  great  many  proton  groups  are  revealed, 
and  they  fit  into  the  general  level  scheme 



as  summarized  for  all  reactions  in  figure 
13.  It  should  be  emphasized  that  neither 
reaction  is  of  the  self-conjugate  type,  such 
as  Mg24(d,  a)Na22,  where  all  nuclei  and 
particles  involved  are  self-mirrored.  This 
means  that,  because  of  the  expected  con- 
servation laws  applying  to  the  isobaric 
spin  quantum  number,  only  T  =  0  states 
should  be  excited  in  the  latter  reaction, 
since  in  Mg24  the  deuteron  and  a  particle 

.V  5.0  - 

4  0 

1                         1 
T=  l 


Tz  =0-*»Tz»| 

0+  -*■  0  +           / 


/  p30 






— 1 1 



4.0  5.0 

1/2  A2/3 

Fig.  12.  Systematic  plot  of  Coulomb-energy 
differences  due  to  addition  of  one  proton  for  the 
group  of  nuclei  having  equal  (odd)  numbers  of 
neutrons  and  protons  vs.  A2/3.  Corresponding 
members  of  the  T  =  1  isobaric  triplets  are  used. 
Na22  is  seen  to  fall  exactly  as  expected.  The 
near-straight-line  variation  between  nuclear  shells 
at  O16  and  Ca40  indicates  the  dependence  of  the 
nuclear  radius  on  the  cube  root  of  the  atomic 
weight  A  and  the  correctness  of  ascribing  the 
energy  difference  to  the  electrostatic  energy  in- 
volved in  adding  one  proton  to  these  nuclei. 

are  all  in  their  T  =  0  ground  states,  while 
our  two  reactions  allow  both  T  =  0  and 
T=l  states  to  be  formed  in  Na22  (since 
F19,  He3,  and  the  neutron  have  T=l/2). 
As  can  be  seen  from  figure  13  the  (d,  a) 
reaction  excites  all  low-lying  states  of  Na22 
except  the  new  one  at  666  kev.  This  con- 
stitutes additional  strong  evidence  for  its 
T—l  character. 

Fig.  13.  Summary  of  present  knowledge  on 
level  structure  of  Na22,  showing  information 
from  both  reactions  used  by  us,  and  the  Mg24- 
(d,  a)Na22  "self -con jugate"  reaction  capable  of 
exciting  only  T  =  0  states.  Tentative  spin  assign- 
ments are  shown  in  parentheses. 

Atomically  Polarized  Ion  Source 

Considerable  interest  in  both  nuclear  and 
elementary  particle  physics  has  centered 
over  the  past  few  years  in  the  production 


and  use  of  polarized  particles  for  the  study 
of  nuclear  interactions.  The  preparation  of 
polarized  beams  has  thus  far  always  in- 
volved a  first  polarizing  event,  either  a 
nuclear  reaction  such  as  the  (d,  d)  or  the 
Li7(/7,  n)  reaction  at  low  energies,  or  an 
elastic  or  inelastic  scattering  event  at  high 
energies  (of  the  order  of  100  Mev  or 
above) .  The  detection  of  preferential  spin 
orientation  in  outgoing  beams  after  these 
encounters  then  involves  another  interac- 
tion, such  as  a  scattering  of  protons  or 
neutrons  from  helium,  where  the  analyz- 
ing power  is  very  well  understood.  Un- 
fortunately such  "double  scattering"  ex- 
periments suffer  greatly  from  lack  of  in- 
tensity, since  the  "beams"  of  polarized 
particles  are  expressed  in  105  to  108  particles 
per  cm2  per  sec  rather  than  in  microam- 
peres (1012  per  sec)  as  is  appropriate  for 
nuclear  studies.  It  is  clear  that  a  beam  of 
polarized  particles  is  intrinsically  capable 
of  detecting  higher  "moments"  of  a  nu- 
clear encounter  than  an  unpolarized  beam, 
which  of  necessity  averages  over  all  spin 
orientations,  thus  sacrificing  a  certain 
amount  of  information  content. 

A  number  of  suggestions  have  been 
made  recently  in  an  attempt  to  improve 
the  situation  with  regard  to  polarized- 
particle  beams.  Most  of  them  have  cen- 
tered around  the  idea  of  polarizing,  say, 
protons  in  the  atomic  state,  thus  having 
the  benefit  of  working  with  an  atomic 
rather  than  a  nuclear  Bohr  magneton.  In 
principle,  then,  the  same  trick  is  to  be 
tried  as  is  employed  for  the  polarization 
of  radioactive  nuclei  through  the  inter- 
mediary of  the  strong  hyperfine  interac- 
tion between  the  electron  cloud  and  the 
nucleus  imbedded  within  it,  as  was  de- 
scribed in  Year  Book  54.  Therefore  a 
more  or  less  conventional  atomic  beam 
experiment  of  the  Rabi  type  with  atomic 
hydrogen  or  deuterium  is  contemplated, 
involving  an  arrangement  of  magnetic 
field  gradients  and  uniform  magnetic 
fields,  together  with  judiciously  placed 
mechanical  stops  and  radiofrequency-in- 
duced  transitions  in  such  a  way  as  to  per- 

mit only  one  of  the  hyperfine  components 
to  pass  through  the  array.  This  component 
then  consists  of,  say,  hydrogen  atoms  polar- 
ized transversely  to  the  beam  direction,  at 
thermal  energies.  This  implies  polarized 
hydrogen  nuclei  (protons),  since  they  have 
a  definite  orientation  with  respect  to  the 
electron  spin  (parallel  or  antiparallel) .  If 
this  beam  is  now  ionized  in  some  manner, 
as  by  electron  bombardment,  a  beam  of 
essentially  complete  proton  polarization  is 
obtained.  In  order  to  verify  the  existence 
of  polarization,  it  is  necessary  to  perform  a 
nuclear  experiment  of  some  sort,  such  as 
the  above-mentioned  scattering  from  he- 
lium at  about  2  million  electron  volts.  In 
other  words,  the  atomic  beam  source  will 
have  to  be  placed  in  an  accelerator  before 
we  shall  know  whether  any  depolarizing 
influences  have  been  overlooked  in  the 

Vernon  W.  Hughes  and  Charles  W. 
Drake,  of  Yale  University,  have  suggested 
a  collaborative  effort  to  attempt  to  produce 
this  type  of  polarized  beam,  making  use 
of  our  uniquely  suited  facilities.  Because 
of  the  vast  dimensions  of  our  pressurized 
Van  de  Graafif  generator  it  is  possible, 
without  giving  further  thought  to  minia- 
turization or  reduction  of  power  require- 
ments, to  install  a  conventional  atomic 
beam  apparatus  in  our  high-voltage  ter- 
minal. Neither  space  (about  3  cubic 
meters)  nor  power  requirements  (about  7 
kilowatts  for  the  source  proper)  are  prob- 
lems in  our  accelerator.  Furthermore,  our 
ability  to  accelerate  protons  up  to  2  Mev, 
the  region  of  immediate  interest,  without 
having  to  pressurize  the  generator  will 
greatly  facilitate  the  initial  alignment  pro- 
cedure. No  other  accelerator  of  this  type 
exists  in  the  world  at  present. 

Last,  but  not  least,  the  uncluttered  na- 
ture of  our  machine  schedule,  the  lack  of 
obligation  to  graduate  students  working  on 
theses,  or  to  other  investigators  committed 
to  different  research  projects,  makes  our 
situation  particularly  suited  to  this  kind 
of  pioneering  effort. 

The  source  proper  is  being  constructed 



and  assembled  at  Yale  University,  and 
will  be  subjected  to  preliminary  tests  there 
before  being  installed  in  our  high-voltage 
terminal.  We  expect  to  have  preliminary 
indications  of  the  feasibility  (or  lack 
thereof)  during  the  next  report  year. 


E.  T.  Bolton,  R.  J.  Britten,  D.  B.  Cowie,  and 
R.  B.  Roberts 


It  is  our  privilege  to  study  life's  most  fas- 
cinating problem — the  nature  of  life  itself. 
Such  a  tremendous  subject  must  of  course 
be  broken  down  into  lesser  parts  which 
can  be  comprehended,  or  the  investigator 
would  be  overwhelmed.  In  our  dissection 
of  the  problem  we  have  asked  a  subsidiary 
question,  but  still  a  profound  one.  Can  the 
physical  attributes  of  life  be  explained  in 
terms  of  the  properties  of  atoms  and  mole- 
cules ? 

Training  in  physical  science  does  not 
qualify  us  in  any  special  way  to  question 
whether  there  is  a  spiritual  as  well  as  a 
physical  aspect  of  living  creatures,  whether 
a  dualism  exists.  It  is,  however,  within  our 
capabilities  to  attempt  to  discern  limita- 
tions to  our  understanding  of  the  physical 
and  chemical  processes  of  life.  Limitations 
here  might  be  comparable  to  the  concep- 
tual limitations  encountered  in  attempting 
to  interpret  atomic  physics  without  the 
help  of  quantum  theory. 

When  the  problem  of  the  nature  of  life 
is  restricted  in  this  way,  the  biological  ma- 
terial chosen  as  most  amenable  for  study 
needs  only  to  have  the  essential  chemical 
and  physical  properties  of  life,  and  need 
not  exhibit  any  mental  or  spiritual  capaci- 
ties. But  what  are  the  essential  chemical 
and  physical  properties  of  life?  If  life  had 
been  synthesized  de  novo  from  nonliving 
ingredients  in  a  beaker  on  our  laboratory 
bench,  how  would  we  recognize  it? 

One  obvious  quality  of  living  organ- 
isms is  their  capacity  for  growth.  Another 
is  reproduction.  Yet  those  qualities  alone 
do  not  define  life,  and  in  framing  a  defini- 

tion it  is  necessary  to  choose  words  and 
concepts  that  include  such  creatures  as 
mules,  which  cannot  reproduce,  and  that 
exclude  salt  crystals,  which  can  grow. 
Furthermore,  such  objects  as  seeds,  and 
perhaps  even  viruses,  should  be  included. 
Both  have  the  capability  of  catalyzing  the 
synthesis  of  more  of  their  kind  even 
though  they  may  not  exhibit  growth  or 
other  metabolic  activity  for  long  periods 
of  time.  Thus,  there  seems  to  be  no  simple 
answer  to  how  newly  created  life  might  be 
recognized.  On  the  other  hand,  there  are 
several  useful  central  concepts  that  seem 
appropriate  to  considerations  of  life  at  the 
macromolecular  level.  Among  the  most 
vital  are  autocatalysis,  and  its  implications 
for  growth  and  reproduction,  and  the 
capacity  for  evolution,  the  ability  to  give 
rise  to  persistent  lines  even  though  they 
may  differ  from  parental  stock.  Also,  a 
minimum  level  of  "organization"  seems  to 
be  involved — sucrose  molecules  do  not 
qualify  as  living,  except  in  poetic  imagery; 
and  viruses,  which  are  nucleoprotein  mole- 
cules, perhaps  may. 

The  components  of  living  cells  that  may 
fall  within  the  scope  of  these  speculations 
about  the  nature  of  living  matter  are 
deoxyribose  nucleic  acid  (DNA)  and  ri- 
bose  nucleic  acid  (RNA)  or  perhaps  these 
materials  in  association  with  protein  (DNP 
and  RNP).  Certainly  these  together  qual- 
ify as  a  living  system  when  they  are 
within  the  protoplasm  of  a  growing  cell. 
Accordingly,  a  large  fraction  of  our  in- 
terest has  continually  been  focused  on  the 
biosynthesis  of  these  large  and  compli- 
cated molecules.  The  processes  supplying 
the  energy  and  the  material  required  for 
synthesis  are  complicated,  but  they  can  be 
understood  in  principle.  The  processes  by 
which  their  building  blocks  are  assembled 
in  proper  order,  however,  remain  obscure. 
Even  so,  there  is  no  indication  that  the 
usual  forces  of  chemistry  will  fail  to  pro- 
vide an  adequate  interpretation. 

In  this  year,  much  more  than  in  the 
years  gone  by,  we  seem  to  be  approaching 


a  climax  in  a  long  story.  Almost  twenty 
years  ago  Brachet  and  Caspersson  showed 
that  RNA  was  prominent  where  or  when 
protein  synthesis  was  in  progress.  Quite 
recently  it  has  been  demonstrated  that  in 
virus  infections  it  is  the  nucleic  acid  and 
not  the  protein  part  of  the  virus  that  speci- 
fies the  protein  of  the  virus  progeny.  An- 
other observation  of  recent  years  is  that 
most  of  the  RNA  of  cells  is  held  in  large 
part  in  granules  of  nucleoprotein  ranging 
from  1  to  4  million  in  molecular  weight. 
These  particles,  recently  named  ribosomes, 
are  found  in  such  widely  varied  sources 
as  bacteria,  pea  seedlings,  and  rat  liver. 
Accordingly,  there  has  been  a  growing 
belief  that  the  ribosomes  form  the  site 
and  machinery  for  protein  synthesis.  This 
theory  has  been  strengthened  by  the  find- 
ing in  several  laboratories  that  freshly  in- 
corporated amino  acids  appear  first  in 
protein  associated  with  the  ribosomes  and 
only  later  in  other  protein.  Furthermore, 
cell-free  preparations  of  purified  ribosomes 
(with  the  addition  of  certain  enzymes  and 
cofactors)  have  shown  some  incorporation 
of  amino  acids  into  protein. 

Nevertheless,  a  considerable  measure  of 
doubt  about  the  mechanism  of  protein 
synthesis  still  remains.  The  observed  in- 
corporation into  cell-free  ribosomes  re- 
quires only  the  addition  of  one  amino 
acid  at  a  time  and  thus  suggests  exchange. 
Two  other  laboratories  have  reported  dur- 
ing the  year  that  fragments  of  cellular 
membranes  carry  out  a  more  extensive 
synthesis  of  protein  and  do  in  fact  require 
the  presence  of  all  the  amino  acids.  Pos- 
sibly these  findings  may  be  reconciled  if 
it  turns  out  that  the  ribosome  must  unfold 
on  the  surface  of  the  membrane  to  be 
active  as  a  template  for  synthesis.  In  any 
event,  further  work  is  needed  before  the 
issue  is  settled. 

This  year's  work  in  our  Biophysics  Sec- 
tion has  again  centered  around  protein 
synthesis.  Ion-exchange  columns,  radio- 
active tracer  methods,  and  the  Spinco 
Model  L  centrifuge  were  used  to  study  the 
properties  of  the  ribosomes  and  the  kinetics 

of  their  formation.  With  the  new  Spinco 
Model  E  analytical  centrifuge  we  are  at- 
tempting to  understand  the  biological 
significance  of  the  different  sizes  of  ribo- 
somes. Bacteria  rapidly  growing  in  a  broth 
medium  contain  more  of  the  smaller  par- 
ticles than  bacteria  growing  in  a  glucose- 
salts  medium,  and  cells  treated  with 
chloramphenicol  in  order  to  halt  protein 
synthesis  contain  mostly  one  class  of  ribo- 
some. Such  results,  although  only  a  preface 
to  chapters  yet  to  be  written,  show  that 
the  ribosome  size  distribution  can  serve 
as  an  indicator  of  the  protein-synthesizing 
ability  of  the  cell.  It  is  hoped  that  current 
investigations  will  reveal  whether  the  size 
distributions  are  causes  or  effects  of  altered 
protein-synthesizing  capacity.  It  is  already 
clear  that  the  ribosomes  contain  protein 
distinct  from  the  other  protein  of  the  cell 
and  that  the  ribosomes  cannot  disintegrate 
to  supply  the  other  protein.  On  the  con- 
trary, synthesis  of  the  ribosomes  appears 
to  proceed  at  quite  a  leisurely  rate.  Studies 
with  radioactive  tracers  indicate  that  the 
nucleic  acid  portion  of  the  ribosome  is 
derived  from  nucleic  acid  macromolecules 
which  are  initially  chemically  associated 
with  little,  if  any,  protein.  This  result  sug- 
gests that  ribosomes  may  grow,  and  move 
from  one  size  class  to  another,  by  combin- 
ing with  protein  and  nucleic  acid  macro- 
molecules  rather  than  by  adding  small 

The  use  of  amino  acid  analogs  has  pro- 
vided other  information  about  the  mecha- 
nism of  protein  synthesis.  The  cells  can- 
not perfectly  distinguish  these  analogs 
from  the  usual  amino  acid,  and  they  are 
incorporated  into  the  protein.  Further- 
more, they  are  incorporated  into  different 
proteins  in  the  same  proportion,  suggest- 
ing that  the  mechanism  for  amino  acid 
selection  does  not  differ  from  one  protein 
to  another.  A  given  analog  affects  the  ac- 
tivity of  different  enzymes  to  different 
degrees.  This  observation  suggests  that  the 
amino  acid  complement  at  the  active  sites 
of  an  enzyme  may  be  examined  by  study- 



ing  the  sensitivity  of  the  enzyme  to  a 
spectrum  of  analogs. 

Other  studies  have  been  carried  out  with 
Hydra  and  Planaria.  Collaboration  has 
continued  with  Drs.  Flexner,  of  the  Uni- 
versity of  Pennsylvania,  in  a  study  of  pro- 
tein synthesis  in  mouse  tissues.  With  these 
more  highly  organized  animals  the  prob- 
lems are  more  complicated,  as  interactions 
between  different  types  of  cells  are  in- 
volved. The  experiments  show  the  usual 
processes  of  synthesis  as  they  occur  in  dif- 
ferent animals  and  different  tissues  and,  in 
addition,  provide  an  introduction  to  the  ef- 
fects of  higher  levels  of  organization.  It  is 
encouraging  to  find  that  much  of  the  ex- 
perience gained  with  microorganisms  is 
useful  in  dealing  with  complex  animals. 

The  experimental  work  leading  to  these 
conclusions  is  described  in  detail  below. 


A  single  cell  of  the  bacterium  Escher- 
ichia coli  contains  roughly  10,000  ribo- 
somes.  As  was  mentioned  above,  there  are 
numerous  indications  that  the  ribosomes 
play  some  important  role  in  protein  syn- 
thesis, but  additional  experimental  evi- 
dence is  needed  to  prove  the  point  con- 
clusively and  to  supply  information  on  the 
mechanism  involved.  One  approach  is  to 
study  the  properties  of  the  ribosomes:  their 
stability,  their  composition,  and,  if  pos- 
sible, their  structure.  Another  is  to  cor- 
relate the  proportion  of  different  sizes  of 
ribosomes  with  the  metabolic  state  of  the 
bacteria.  Still  another  is  to  observe  the 
incorporation  of  various  tracers  and  their 
transfer  from  one  class  of  molecules  to 

As  a  first  step  it  is  necessary  to  break  the 
cells  and  to  separate  out  the  various  types 
of  molecules.  Suppose,  for  example,  that 
one  class  of  particles  is  the  precursor  of 
another,  or  of  the  soluble  protein.  Kinetic 
measurements  of  tracer  incorporation  will 
show  this  clearly  if  the  various  classes  can 
be  separated.  Thus,  chemical  fractionation 
with  various  extracting  agents  separates 
the  small  molecules  from  lipide,  protein, 

and  nucleic  acid,  and  kinetic  measurements 
show  that  the  small  molecules  serve  as 
precursors  for  the  large  ones. 

Chemical  fractionation  does  not,  how- 
ever, distinguish  free  nucleic  acid  and  pro- 
tein from  nucleoprotein.  Nor  does  it  dis- 
tinguish which  components  are  bound  in 
the  cell  wall  or  membrane.  To  distinguish 
these  classes  we  have  tried  various  cen- 
trifuging  techniques,  electrophoresis,  and 
cellulose  ion-exchange  columns.  Chemical 
analysis  of  the  fractions  obtained  by  these 
methods  is  then  useful  to  determine  the 
composition  and  purity  of  the  fractions. 
Also,  the  fractions  obtained  by  one  method 
can  be  analyzed  further  by  another. 

The  French  pressure  cell  has  proved 
very  satisfactory  in  breaking  the  cells. 
Large  quantities  of  cells  (10  g)  can  be 
broken  rapidly,  and  the  breakage  is  99  per 
cent  complete.  When  necessary,  the  resid- 
ual unbroken  cells  can  be  removed  by 
centrifuging  four  times  successively  for  2 
minutes  in  the  Servall  centrifuge  to  give 
a  cell-free  juice.  This  step  also  removes  an 
appreciable  fraction  of  the  large  fragments 
of  cell  debris. 

The  bacterial  juice  obtained  from  the 
pressure  cell  has  a  low  viscosity,  as  DNA 
is  fragmented  on  passing  through  the  cell. 
Low  viscosity  is  usually  an  advantage,  for 
the  succeeding  fractionation  steps  are  diffi- 
cult when  a  mass  of  jelly  like  DNA  is 
present.  If  unbroken  DNA  is  required,  a 
different  way  of  breaking  the  cells  must 
be  used.  Grinding  with  alumina  or  os- 
motic shock  of  lysozyme-treated  cells  will 
yield  intact  DNA,  but  another  method  has 
proved  far  superior. 

In  this  method  the  cells  are  washed  in 
a  magnesium-free  tris-succinate  buffer 
(TS)  0.01  M  adjusted  to  pH  8.  Lysozyme 
(0.1  mg/ml),  ethylenediaminetetraacetic 
acid  (EDTA)  (0.01  M),  and  sodium  de- 
oxycholate  (1  per  cent)  are  added.  The 
lysozyme  and  deoxycholate  act  in  con- 
junction to  lyse  the  cells  promptly  and 
give  a  viscous  fluid  from  which  the  DNA 
can  be  removed  with  a  stirring  rod  after 
the  addition  of  ethanol.   The  usual  meth- 


ods  are  then  followed  to  purify  the  DNA. 
In  most  of  our  work  the  undegraded 
DNA  is  not  required  and  the  pressure  cell 
juice  is  used.  Cell  walls  and  debris  are 
quite  effectively  removed  by  a  single  cen- 
trifugation  for  15  minutes  at  100,000^. 
There  are  some  indications  that  the  cell- 
wall  fraction  is  not  homogeneous.  If  the 
pressure  cell  juice  is  spun  twice  at  40,000^ 
for  15  minutes  the  second  pellet  which  con- 
tains  the   smaller   fragments   has   almost 

in  2  hours)  and  small  bits  of  cell  mem- 
brane (this  pellet  has  a  high  lipide  con- 
tent) .  The  contamination  is  best  shown  by 
the  column  analysis  described  below.  The 
supernatant  fluid  contains  soluble  proteins, 
free  RNA  and  DNA. 

A  further  purification  of  the  ribosomes 
is  achieved  by  resuspending  this  pellet  and 
spinning  at  40,000^  for  15  minutes.  The 
proteins  and  membrane  contaminants  do 
not  resuspend  but  remain  aggregated  and 

Fraction  number 

Fig.  14.  Fractionation  of  particle  preparations  using  the  swinging  bucket  centrifuge.  Five-tenths 
milliliter  of  suspension  is  placed  on  top  of  4.5  ml  sucrose  gradient  in  the  centrifuge  tube.  After  45 
minutes  at  100,000^,  0.3-ml  fractions  are  taken  off  from  the  top  with  a  pipet. 

twice  the  proportion  of  lipide  found  in  the 
first  pellet,  indicating  that  it  may  contain 
more  lipoprotein  "membrane"  and  less  of 
the  structural  "wall."  The  nucleic  acid 
content  of  these  fractions  is  low,  indicat- 
ing only  a  small  contamination  by  ribo- 

Most  of  the  wall  or  membrane  having 
been  removed,  the  remaining  fluid  is  cen- 
trifuged  at  100,000g-  for  V/2  to  2  hours;  the 
resulting  pellet  is  (by  definition)  the  mi- 
crosomal fraction.  It  contains  ribosomes 
together  with  large  proteins  (roughly  70 
per  cent  of  p-galactosidase  is  sedimented 

are  sedimented.  The  ribosomes  go  into 
suspension  more  readily  and  are  subse- 
quently sedimented  by  centrifuging  at 
100,0(%  for  2  hours. 

Attempts  were  made  to  separate  the 
various  sizes  of  ribosomes  by  choosing  an 
appropriate  centrifuging  schedule.  Thus, 
the  analytical  centrifuge  showed  that  ma- 
terial which  sediments  in  15  minutes  at 
100,000g-  is  richer  in  the  large  particles 
than  the  pellet  obtained  by  centrifuging 
(2  hours,  100,000g)  material  remaining  in 
the  supernatant  fluid  after  three  successive 
15-minute  100,000g  spins.   This  approach 


showed  no  real  promise  of  giving  adequate 

A  better  separation  of  the  different  ribo- 
somes  can  be  obtained  by  means  of  the 
swinging  bucket  head  for  the  Spinco 
Model  L  centrifuge.  Microsome  pellets  are 
resuspended  and  layered  on  top  of  a 
sucrose  gradient.  After  a  period  of  cen- 
trifugation,  layers  are  taken  off  with  a 
pipet.  This  technique  is  adequate  to  dem- 
onstrate marked  differences  in  the  distribu- 
tions, depending  on  the  initial  material. 
Figure  14  shows  one  curve  for  a  resus- 
pended pellet  composed  mostly  of  large 
(80S)  particles,  another  for  the  smaller 
particles  (20  to  40S),  and  a  third  for  the 
nonsedimenting  material.  The  analytical 
centrifuge  shows  that  the  bottom  layers 
are  rich  in  the  heavy  particles  and  lack  the 
light  particles,  whereas  the  top  layers  have 
the  opposite  distribution. 

An  entirely  different  type  of  fractiona- 
tion results  from  chromatography  on 
columns  of  diethylaminoethyl  cellulose 
(DEAE).  Extremely  high  resolution  can 
be  achieved  giving  a  separation  of  various 
proteins,  as  shown  in  figure  15.  Nucleo- 
protein  appears  as  a  prominent  peak  in 
the  elution  diagram  of  the  total  cell  juice 
but  not  in  the  diagram  obtained  with  the 
100,000g-  supernatant  fluid  (fig.  16).  The 
corresponding  ultraviolet  diagrams  show 
two  main  peaks:  the  first  consists  of 
nucleoprotein  of  high  molecular  weight 
which  can  be  spun  down  in  the  centrifuge; 
the  second  is  partly  nucleoprotein  and 
partly  due  to  free  DNA  and  RNA  still 
remaining  in  the  100,000g  supernatant 

The  elution  pattern  is  not  sensitive  to 
the  size  of  the  particles.  The  same  pattern 
is  obtained  whether  the  microsome  pellet 
is  composed  mostly  of  the  large  (80S) 
particles  or  of  the  smaller  (20  to  40S)  ones 
that  result  from  magnesium  deficiency  or 
treatment  with  NaCl.  Compare  figures 
17 A  and  17 B. 

Microsome  pellets,  when  resuspended 
and  analyzed  on  the  column,  show  the 
nucleoprotein  peak  together  with  a  quan- 

tity of  other  protein  which  depends  on 
the  method  of  preparation  (fig.  17).  The 
least-contaminated  preparations  of  ribo- 
somes  are  those  obtained  by  resuspending 

40  60         80         100 

Fraction    number 

Fig.  15.  Cells  washed  in  TSM  and  broken 
with  French  pressure  cell;  0.5  g  wet  weight  of 
cell  juice  adsorbed  on  DEAE  column  (1  cm2  X  20 
cm)  and  eluted  with  concentration  gradient  0 
to  0.7  M  of  NaCl  in  tris-succinate  buffer  plus 
magnesium.  Lower  curve,  total  protein  indicated 
by  Folin  reaction;  upper  curve,  assay  for  activity 
of  three  different  enzymes.  One-milliliter  sam- 
ples collected  in  fraction  collector. 

a  microsome  pellet  and  centrifuging  again 
in  the  swinging  bucket  head  (fig.  17C). 
Unfortunately,  the  column  cannot  be 
used  to  prepare  purified  ribosomes  because 
the  material  eluted  from  the  column  is 
very    different    from    that    originally    ad- 



sorbed.  When  the  fractions  containing  the 
nucleoprotein  peak  are  centrifuged  (100,- 
OOOg-,  2  hours),  a  colorless  glassy  pellet  is 
formed  which  contains  approximately  65 
per  cent  of  the  protein  and  nucleic  acid. 
This  pellet  resuspends  easily  and  com- 
pletely. The  analytical  centrifuge  shows 
that  it  contains  peaks  in  the  20  to  40S 

to  molar  NaCl  show  a  reduction  in  size 
but  no  change  in  composition  or  elution 
pattern.  This  nucleoprotein  derived  from 
ribosomes,  but  having  a  changed  amino 
acid  to  base  ratio  as  a  result  of  degrada- 
tion on  the  column,  is  designated  CNP 
to  distinguish  it  from  the  original  NP  of 
the  ribosomes.    The  letter  "C"  in  CNP 

,-Totol  cell  extract 

100,000  g  SN 

Fig.  16.  Elution  patterns  of  total  cell  juice  and  supernatant  fluid  of  100,00%  2-hour  spin.  Upper 
curve,  optical  density  at  254  m[i  indicating  nucleic  acid  concentration;  lower  curve,  S35  radioactivity, 
indicating  protein.  Note  nucleoprotein  peak  which  is  missing  in  100,000^  SN.  Figures  16  through 
22  are  taken  by  permission  from  Roberts,  Microsomal  Particles  and  Protein  Synthesis,  published 
for  the  Washington  Academy  of  Sciences  by  the  Pergamon  Press,  1958. 

region,  whereas  the  80S  peak  was  most 
prominent  in  the  original  material.  The 
ratio  of  nucleic  acid  to  protein  in  the  pellet 
(measured  by  optical  density  at  260  m[j 
and  S35)  is  twice  that  of  the  starting  ma- 
terial, and  the  elution  pattern  obtained 
when  the  pellet  is  rerun  on  a  DEAE 
column  is  very  different  (fig.  18). 

These  changes  appear  to  be  caused  by 
the  column  material  and  not  by  the  salt 
of  the  eluting  fluid.    Ribosomes  exposed 

was  chosen  as  a  reminder  of  the  method 
of  preparation  by  elution  from  a  chromato- 
graphic column. 


The  ribonucleoprotein  of  E.  coli  occurs 
in  several  species  of  particles  which  can 
be  differentiated  by  their  sedimentation 
rates  in  the  analytical  ultracentrifuge.  The 
proportions  of  particles  in  the  ribosome 
species  may  be  varied  in  vitro  by  altering 


.2  1- 

TS  Microsome  pellet 

TSM    Microsome  pellet 

3      5     7      9     II      13     15    17     19    21     23  25    27   29   31    33   35 
Column  fraction  no. 

Fig.  17.  Elution  patterns  of  microsome  pellets.  (A)  100,000^  2-hour  pellet  with  magnesium  pres- 
ent; (B)  same  with  magnesium  lacking;  (C)  microsome  pellet  resuspended  and  fractionated  with 
swinging  bucket  centrifuge. 


the  composition  of  the  suspending  fluid 
and  in  vivo  by  altering  the  conditions 
under  which  the  bacteria  are  cultured. 

Figure  19  shows  typical  sedimentation 
diagrams  of  French  pressure  cell  extracts 
of  exponentially  growing  E.  coli  which 
were  disrupted  in  the  presence  of  TSM 
(tris-succinate-magnesium),  TS,  or  TSM 
and   phosphate  buffer.    Approximate  ap- 

or  relative  proportions  of  components.  The 
sedimentation  diagrams  of  bacterial  ex- 
tracts prepared  with  the  dilute  tris-suc- 
cinate  buffers  at  pYL  7.6  remain  unchanged 
as  a  result  of  20  hours'  storage  at  0°  to 
4°  C  (fig.  20).  Addition  of  the  chelater 
EDTA  (fig.  21),  however,  or  of  the  en- 
zyme ribonuclease  (fig.  22)  removes  all 
components     with     sedimentation     rates 

16  20  24 

Column   fraction  no. 

Fig.  18.    Nucleoprotein  peak  of  elution  pattern  (shaded  area)  spun  down  and  rechromatographed 
(lower  figure).    Note  change  to  elution  pattern  like  that  of  nucleic  acid. 

parent  sedimentation  rates  are  indicated 
along  the  abscissa  of  the  upper  diagram. 
It  is  evident  from  comparison  of  these 
diagrams  that  more,  and  larger,  com- 
ponents are  observed  when  magnesium 
has  been  included  in  the  buffer.  The  addi- 
tion of  phosphate  abolishes  the  more  rap- 
idly sedimenting  components.  Cysteine 
and  sucrose,  which  are  frequently  included 
in  suspending  media  for  subcellular  ele- 
ments, are  without  effect  upon  the  number 

greater  than  20S.  Deoxyribonuclease  (fig. 
22),  on  the  other  hand,  has  no  apparent 
effect  on  the  sedimentation  behavior  of  the 

The  sedimentation  patterns  of  the  ex- 
tracts may  also  be  varied  by  altering  the 
conditions  under  which  the  bacteria  are 
cultured.  As  is  clear  in  figure  23,  plate  2,7 
marked   differences   appear  even   though 

7  Plates  2  to  6  are  grouped  between  pages 
160  and  161. 


the  cells  were  harvested  and  washed  in 
the  same  TSM  medium.  Another  example 
shown  in  figure  23  is  the  change  which 
results  when  chloramphenicol  is  added  to 

20   40  60   80   S 


mm    60,000   RPM 

Fig.  19.  Sedimentation  diagrams  of  E.  coll 
disrupted  in  various  buffer  solutions.  The  con- 
centration of  the  bacterial  juices  differed  among 
the  runs. 

a  growing  culture  of  bacteria.  These 
changes  seem  an  important,  though  as 
yet  uninterpreted,  clue  to  the  role  of  the 
ribosomes  and  the  mechanism  of  their 

Composition  of  Ribosomes 

Ribosomes  purified  by  differential  cen- 
trifugation  contain  40  per  cent  protein  and 
60  per  cent  nucleic  acid  by  weight.  This 
distribution  is  based  on  protein  estimations 
made  by  chemical  and  radioactive  tracer 
methods,  and  on  ribonucleic  acid  measure- 
ments made  by  chemical  and  spectropho- 
tometric  means.  Since  the  average  molecu- 
lar weight  of  an  amino  acid  residue  in 
the  protein  is  about  108,  and  the  average 
molecular  weight  of  a  nucleotide  building 
block  in  the  nucleic  acid  is  about  325,  there 
are  on  the  average  two  amino  acids  for 
each  nucleotide  in  the  ribosome. 

The  amino  acid  composition  of  the  ribo- 
somes is  also  distinctive.  These  particles 
contain  somewhat  less  methionine  than 
the  average  bacterial  protein  and  less  than 
one-fortieth  the  cyst(e)ine.  They  probably 
do  not  lack  cystine  entirely,  since  the  en- 
zyme ribonuclease,  which  contains  it,  has 
been  found  to  occur  in  the  ribosomes. 
They  are  relatively  rich  in  lysine  and  poor 
in  aspartic  acid.  The  remaining  amino 
acids  appear  in  roughly  the  usual  propor- 
tions. The  nucleic  acid  is  probably  en- 
tirely of  the  ribonucleic  acid  type.  Chemi- 
cal tests  for  deoxyribonucleic  acid  have 
been  negative. 

When  the  purified  ribosomes,  which 
are  mostly  40  to  80S  particles,  are  chroma- 
tographed  by  ion  exchange  on  DEAE- 
cellulose,  a  single  nucleoprotein  peak  is 
observed  (cf.  fig.  17C).  This  peak  (CNP) 
is  a  mixture  of  high-molecular-weight 
components  (sedimentation  rates  about  20 
to  40S).  When  the  mixture  is  analyzed  for 
protein  and  nucleic  acid  an  average  of 
only  one  amino  acid  is  found  for  each 
nucleotide.  It  appears,  therefore,  that  the 
chromatographic  procedure  strips  off  about 
one-half  of  the  ribosomal  protein  and  de- 
creases the  size  of  the  particles.  The  pro- 
tein lost  from  the  particles  is  firmly  bound 
to  the  exchanger,  but  may  be  eluted  with 
sodium  hydroxide.  When  the  CNP  is 
again  chromatographed  it  breaks  up  into 
two  regions  (fig.  18).  Thus,  the  chromato- 


graphic  procedure  itself  degrades  the  ri- 
bosomes  and  yields  a  series  of  products 
that  indicate  heterogeneity  among  the 
building  blocks  of  the  ribosomes. 

The  heterogeneity  of  the  ribosomal  pro- 
tein was  studied  by  examining  the  ion- 
exchange  behavior  of  ribosomes  that  had 
been  degraded  by  4  M  urea,  an  agent 
which   decomposes   hydrogen   bonds   and 

S35  radioactivity  is  illustrated  in  the  up- 
per part  of  the  diagram.  Comparison  of 
the  S35  data  shows  that  urea  decomposes 
the  ribosomes  and  gives  rise  to  several 
new  S35-labeled  components.  These  are 
eluted  much  sooner  than  the  CNP  of  the 
ribosome.  Much  of  the  protein  radioac- 
tivity actually  migrates  with  the  front. 
This  behavior  indicates  that  the  proteins 

0   hours 

20  40      80      S 

M     I 

20  hours 

12   min     60,000  RPM 

Fig.  20.  Influence  of  storage  at  4°  C.  The  suspending  buffers  (TSM,  left;  TS,  right)  also  con- 
tained 0.25  M  sucrose,  although  subsequent  runs  have  shown  that  sucrose  has  no  effect  on  the  pat- 
tern of  components. 

denatures  many  types  of  protein.  Figure 
24  shows  tracings  from  an  automatic  re- 
cording device  which  monitors  the  radio- 
activity and  ultraviolet  absorption  of  the 
effluent  from  an  ion-exchange  column. 
Four  runs  are  illustrated:  ribosomes  that 
had  been  exposed  to  urea  for  2  minutes 
or  for  6  hours,  and  a  run  of  the  CNP  ob- 
tained from  the  ribosome  chromatogram. 
The  CNP  solution  was  centrifuged  at 
100,000g-  for  2  hours  and  the  pellet  was 
suspended  in  4  M  urea  for  2  hours  before 
chromatography  was  started. 

are  uncharged  or  are  positively  charged  at 
pH  7.6.  When  the  CNP  is  urea-degraded 
and  again  chromatographed  nearly  all  the 
labeled  protein  migrates  with  the  front. 
Thus,  the  CNP  appears  to  be  composed 
of  protein  subunits  which  are  almost  en- 
tirely basic  or  uncharged  in  4  M  urea  at 
pH  7.6.  The  particular  ribosome  prepara- 
tion used  for  these  comparisons  was  con- 
taminated with  nonribosomal  proteins, 
which,  even  in  the  absence  of  urea,  are 
eluted  earlier  than  the  CNP  (uppermost 
chromatogram) .  However,  ribosome  prep- 


arations  purified  by  two  cycles  of  differ- 
ential centrifugation  contain  very  little 
contaminating  material.  When  such  ribo- 
somes  are  degraded  with  urea  they  yield 

20  40    80         S 








6  min    60,000  RPM 

Fig.  21.  Effect  of  EDTA  on  the  sedimentation 
diagrams  of  E.  coli  juice.  The  two  lower  dia- 
grams are  from  preparations  containing  one-half 
as  much  material  as  those  for  the  upper  pattern. 

chromatograms    essentially    like   those   in 
figure  24. 

When  ribosomes  are  treated  with  urea 
a  clear  solution  results,  but  when  the  CNP 
is  treated  with  urea  part  of  the  protein 

precipitates.  Thus,  still  another  class  of 
protein  may  be  distinguished,  and  it  may 
be  inferred  that  urea  degradation  gives 
rise  to  different  products,  depending  on 
the  complexity  of  organization  of  the  start- 
ing material. 

The  lower  part  of  figure  24  shows  the 
ultraviolet-absorption  distribution  along 
the  four  chromatograms.  It  is  evident  that 


20   40       80      S 


RNA  -  ose 

6  min     60.000  RPM 

Fig.  22.  Effect  of  nucleases  on  sedimentation 
diagrams.  The  lower  pattern  is  from  a  prepara- 
tion one-half  as  concentrated  as  that  of  the  upper 

urea  brings  about  dramatic  alterations  in 
these  patterns.  When  the  treatment  has 
been  brief  (2  minutes  at  room  tempera- 
ture) the  location  of  the  ultraviolet-absorb- 
ing material  shows  that  it  is  still  highly 
polymerized.  This  region  is  now  devoid 
of  protein  and  contains  only  acid-precipi- 
table  ribonucleic  acid.  If  urea  treatment  of 
ribosomes  or  CNP  is  allowed  to  proceed 
for  several  hours  the  ultraviolet-absorbing 
material  is  no  longer  acid-precipitable  and 



the  ion-exchange  elution  pattern  resembles 
that  for  a  mixture  of  nucleotides  of  low 
molecular  weight.  The  degradation  of 
ribonucleic  acid  in  this  fashion  implies 
that  the  ribosomes  and  also  the  CNP  con- 
tain ribonuclease  activity.  To  test  this  im- 
plication yeast  ribonucleic  acid  was  added 
to  urea-degraded  ribosomes  or  CNP  after 
the  bacterial  nucleic  acid  had  become  com- 
pletely acid-soluble.  The  yeast  nucleic  acid 


migrates  with  the  front,  after  urea 
radation  of  CNP.  These  results  are  il- 
lustrated in  the  center  of  figure  24.  After 
urea  treatment,  ribonuclease  and  also  all 
the  rest  of  the  proteins  that  move  with  the 
chromatogram  front  are  precipitable  by 
trichloroacetic  acid,  and  may  be  redis- 
solved  at  pH  5  with  0.1  M  acetate  buffer. 
The  ribonuclease  retains  full  enzymatic 
activity    after    this    treatment.    Thus,    in 

Fig.  24.     Heterogeneity  among  the  proteins  of  ribosomes  and  the  destruction  of  ribosomal  nucleic 
acid  by  ribonuclease.   Description  of  these  data  in  the  text. 

was  hydrolyzed  by  both  preparations. 
Yeast  nucleic  acid  was  not  degraded  by 
either  preparation  in  the  absence  of  urea. 
Hence,  it  was  concluded  that  E.  coli  con- 
tains ribonuclease,  buried  in  the  ribonu- 
cleoproteins,  and  inactive  until  the  particles 
are  disrupted  by  the  action  of  urea.  Meas- 
urements to  date  indicate  that  all  the 
E.  coli  ribonuclease  is  contained  in  ribo- 
somes. Thus,  the  enzyme  ribonuclease 
numbers  among  the  various  protein  sub- 
units  of  the  ribosome. 

Further  work  has  shown  that  the  bac- 
terial ribonuclease  chromatographs  exactly 
like  the  CNP  before  urea  treatment,  and 

chromatographic  behavior  on  DEAE-cellu- 
lose  and  solubility  after  trichloroacetic  acid 
precipitation  the  ribonuclease  of  E.  coli 
resembles  the  ribonuclease  of  beef  pan- 
creas. The  demonstration  of  the  latent 
ribonuclease  activity  of  E.  coli  ribosomes 
confirms  and  extends  the  finding  of  Dr. 
David  Elson,  of  the  Weizmann  Institute 
of  Science  (Israel),  who  also  used  urea  to 
degrade  the  ribosomes  of  E.  coli. 


In  order  to  examine  the  precursor-prod- 
uct relationships  among  the  macromole- 


cules  of  the  cell  and  to  assess  the  possible 
role  of  the  ribosomes  in  the  synthetic  proc- 
esses of  the  cell,  kinetic  studies  have  been 
made  of  radioactive-tracer  incorporation 
into  cell  fractions  separated  by  means  of 
ion  exchange  and  the  ultracentrifuge. 

The  kinetic  studies  of  protein  synthesis 
using  S35  as  a  tracer  are  still  in  a  pre- 
liminary stage.  Exponentially  growing 
cells  were  exposed  to  the  tracer  for  varying 
periods  of  time  and  then  broken  and  their 
constituents  separated  by  centrifugation 
and  chromatography.  The  specific  radio- 
activity of  the  protein  fractions  was  meas- 
ured by  TCA-precipitable  S35  and  Folin 
reaction  color.  When  the  cells  are  exposed 
to  the  tracer  for  a  prolonged  period 
(steady  state)  the  specific  radioactivity 
varies  throughout  the  chromatographic 
elution  pattern  by  a  factor  of  roughly  3, 
being  lowest  in  the  nucleoprotein  fraction. 
These  variations  are  simply  due  to  differ- 
ences in  the  sulfur  content  of  the  different 
proteins.  Alternatively,  cells  were  grown 
for  three  generations  in  a  nonradioactive 
medium  after  exposure  to  the  tracer.  In 
this  treatment  any  intermediates  that  have 
a  rapid  turnover  should  lose  their  radio- 
activity. The  resulting  "persistent  pattern" 
was  entirely  similar  to  the  "steady-state 
pattern,"  and  no  protein  components  could 
be  identified  as  intermediates. 

Growing  cells  were  also  exposed  to  the 
tracer  for  short  periods.  After  a  4-minute 
exposure  the  resulting  "pulse  pattern"  was 
similar  to  the  "steady-state  pattern"  except 
that  the  radioactivity  of  the  nucleoprotein 
peak  was  only  half  that  expected  from  the 
"steady-state  pattern."  A  similar  result 
was  obtained  with  cells  exposed  for  4  min- 
utes to  a  mixture  of  C14-labeled  amino 

It  appears  from  these  observations  that 
the  nucleoproteins  (i.e.  CNP)  of  the  ribo- 
somes are  not  precursors  of  other  proteins 
of  the  cell.  Possibly,  however,  newly 
formed  proteins  might  still  be  found  in 
association  with  the  ribosomes  after  break- 
ing of  the  cells.  To  test  this  possibility,  a 

lOOjOOOg-,  2-hour  pellet  containing  ribo- 
somes (and  contaminating  protein)  from 
pulse  S35-labeled  cells  was  analyzed  on  the 
column.  It  was  found  that  the  traces  of 
protein  eluting  at  the  same  salt  concentra- 
tions as  the  bulk  of  the  cellular  protein 
had  in  fact  the  same  specific  radioactivity 
as  the  total  protein  of  the  cell — not  the 
high  specific  activity  indicative  of  pre- 
cursors. This  failure  to  observe  precursor 
proteins  in  association  with  the  ribosomes 
is  not  conclusive  evidence  that  such  an 
association  does  not  exist.  On  the  one 
hand  the  turnover  rate  could  be  so  high 
that  very  much  briefer  pulses  would  be 
required  for  the  observation,  or  on  the 
other  hand  the  association  may  be  so  labile 
as  to  be  destroyed  by  breakage  of  the  cells. 

The  search  for  precursors  of  nucleic 
acid  using  P32  as  a  tracer  has  been  more 
rewarding.  Figure  25  shows  the  nucleic 
acid  region  of  the  elution  patterns  obtained 
with  cells  exposed  to  the  tracer  for  in- 
creasing periods  of  time.  The  radioactivity 
appears  first  in  a  distinct  fraction  of  this 
region,  and  later  in  the  other  fractions.  In 
the  steady-state  and  persistent  patterns  the 
phosphorus  radioactivity  was  proportional 
to  the  amount  of  nucleic  acid  measured 
by  the  optical  density  at  260  mu.  Thus 
the  DEAE  column  is  capable  of  resolving 
the  nucleic  acid  and  nucleoprotein  into 
fractions  that  have  the  kinetic  behavior  of 
precursors  and  products.  Similar  kinetic 
differences  were  observed  earlier  in  this 
laboratory  by  E.  H.  Creaser,  using 
ECTEOLA  columns  to  analyze  alcohol- 
extracted  nucleic  acid. 

It  has  not  been  possible,  as  yet,  to  sepa- 
rate the  precursors  from  the  nonnucleo- 
protein  RNA  and  the  DNA,  owing  to  the 
overlap  of  the  elution  patterns.  A  number 
of  chemical  and  physical  characteristics  of 
the  precursors,  however,  have  been  deter- 
mined. The  precursors  are  principally 
macromolecular  RNA  and  together 
amount  to  about  10  per  cent  of  total  RNA 
of  the  cell,  as  shown  by  the  following 


The  precursor  peaks  are  TCA-precip- 
itable  after  elution  from  the  column.  They 
become  TCA-soluble  at  about  the  same 
rate  as  yeast  RNA  when  exposed  to  beef 
pancreatic  RNAase. 

itable  P32-labeled  RNA  has  an  average 
sedimentation  rate  about  half  that  of  the 
RNA  in  the  nucleoprotein  particles. 

The   quantity  of  RNA  present  in  the 
form  of  the  precursor  is  probably  small, 


Fraction  number 

Fig.  25.  The  labeling  of  macromolecular  nucleic  acid  at  early  times  after  the  addition  of  P32 
to  growing  cells.  Each  curve  shows  the  incorporated  P32  eluted  over  the  range  from  0.35  to  0.75  M 
NaCl,  using  a  DEAE-cellulose  column.  The  broken  curve  at  the  left  shows  the  pattern  of  the 
ultraviolet-absorbing  material.  Since  this  pattern  is  identical  for  all  the  runs  the  locations  of  the  two 
ultraviolet-absorbing  peaks  have  been  simply  marked  NP  and  B. 






1         1         1 

•    P*     c/s 

•  / 

s  * 






o    Optical  density    X  =  260   ~~ 






\           • 

°  \                            — 

I                  1                  1 

Fraction  number 

Fig.  26.  Sedimentation  of  P32  incorporated  into  macromolecules  during  4  minutes'  exposure  of 
growing  cells  to  P3204.  The  counting  rate  of  the  cold  TCA  precipitate  and  the  optical  density  at 
260  mu  are  shown  after  centrifuging  45  minutes  at  125,000^  in  the  swinging  bucket  rotor.  Initially 
a  small  sample  was  loaded  over  a  sucrose  gradient  in  the  centrifuge  tube. 

Figure  26  shows  the  results  of  analysis 
by  means  of  the  swinging  bucket  head  in 
the  ultracentrifuge  (as  described  earlier 
in  this  report)  of  growing  cells  exposed 
to  P32  for  4  minutes.    The  TCA-precip- 

since  the  steady-state  and  "persistent" 
labeling  show  uniform  specific  activity 
throughout  the  elution  pattern.  Alterna- 
tively, the  precursor  might  be  large  and 
circulation  might  occur  as  indicated  in  the 



following  diagram,  where   a   and   (3   are 
rates  of  flow: 






"J/         Product 
tT    RNAandCNP 

The  quantity  of  the  precursor  has  been 
estimated  from  the  kinetics  of  formation 
on  the  basis  of  several  simplifying  assump- 
tions. By  assuming  that  all  the  ultraviolet- 
absorbing  material  has  the  specific  ac- 
tivity of  the  nucleoprotein  peak,  the 
amount  of  P32  in  the  product  RNA  which 
is  not  resolved  from  the  precursor  can  be 
calculated.  The  correction  is  never  large 
and  probably  does  not  seriously  influence 
the  conclusion.  Figure  27  shows  the  P32 
in  the  precursor  and  product  RNA  calcu- 
lated on  this  basis  as  a  function  of  time 
after  addition  of  the  tracer  to  growing  cells. 

The  curve  for  the  total  RNA  has  the 
proper  shape  and  magnitude  expected 
from  the  delay  due  to  the  known  TCA- 
soluble  pool.  By  means  of  a  semiempirical 
curve,  shown  as  the  total  in  figure  27,  the 
two  lower  curves  were  calculated  by  as- 
suming that  the  precursor  contained  10 
per  cent  of  the  total  RNA  and  that  the 
circulation  was  negligible  ((3  =  0). 

Since  the  experimental  points  fit  the 
curves  well,  it  is  clear  that,  if  the  circula- 
tion is  indeed  negligible,  the  quantity  of 
precursors  is  not  far  different  from  10  per 
cent  of  the  total  cellular  RNA.  If  circu- 
lation were  present  the  precursor  would 
be  larger.  The  data  could  not  be  fitted, 
however,  on  the  assumption  that  one  base 
moved  from  precursor  to  product  for  each 
amino  acid  incorporated  into  protein,  since 
(3  would  be  greater  than  10a. 

These  findings  suggest  that  the  inter- 
mediate is  RNA  of  high  molecular  weight, 
free  or  associated  with  less  protein  than 
the  bulk  of  the  nucleoprotein.  It  should 
be  emphasized  that  neither  lipides  nor 
fragments  of  cell  wall  or  of  cell  membrane 
are  eluted  from  the  column,  and  it  is  ob- 
served that  a  large  part  of  the  P32  incor- 
porated in  short  exposures  is  irreversibly 
bound  to  the  column.  An  important  step 

in   the   flow   of   phosphorus   into   nucleic 
acids  may  thereby  be  missed. 

As  an  interesting  sidelight,  cells  were 
exposed  to  chloramphenicol  and  P3204  for 
2  hours.  The  ion-exchange  analysis  shows 
that   no    P32    was    incorporated    into    the 







TOTAL-— ^/ 






/       Yv 


7                /      o/ 


/            /          /    ^ 

'          /          /       PRODUCT 




Jr-'-T5!          | 

1         1         1         1         1         1         1 


Fig.  27.  Time  course  of  P32  incorporation 
into  bacterial  nucleic  acid.  The  ordinate  is  the 
newly  incorporated  phosphorus  expressed  as  the 
fraction  of  the  total  nucleic  acid  phosphorus 
eluted  from  the  column  (zero  time  run).  The 
two  curves  shown  as  precursor  and  product  are 
calculated  from  the  total  curve  assuming  that 
the  amount  of  precursor  is  10  per  cent  of  the 
product  nucleic  acid.  Cells  growing  at  37° 
with  a  generation  time  of  69  minutes.  Experi- 
mental points  are  shown  for  the  P32  incorporated 
into  the  total  (closed  circles),  the  precursor  (tri- 
angles), and  the  product  RNA  and  RNP  (open 
circles) . 

nucleoprotein  peak.  The  bulk  of  the  P32 
was  eluted  at  the  same  salt  concentration 
as  the  earlier  of  the  precursors  described 
above  (fig.  25) .  In  addition,  the  ultraviolet- 
absorbing  material  under  the  nucleoprotein 
peak  was  strongly  accentuated,  indicating 


that  RNA  having  properties  similar  to 
those  of  one  of  the  precursors  was  piled 
up  in  the  cell. 


For  the  past  two  summers  the  biophysics 
group  has  had  one  of  its  members  at  the 
Rocky  Mountain  Laboratory  of  the  U.  S. 
Public  Health  Service,  Hamilton,  Mon- 
tana. The  visits  have  proved  mutually 
beneficial.  Studies  of  interest  to  the  Bio- 
physics Section  were  carried  out  with  the 
electron  microscope  and  analytical  centri- 
fuge, which  were  not  then  available  at  the 
Department.  At  the  same  time,  some  of 
the  ion-exchange  techniques  developed 
here  for  study  of  bacterial  components 
were  applied  to  problems  of  virology. 

In  this  cooperative  venture  with  Dr. 
Bill  H.  Hoyer,  of  the  Rocky  Mountain 
Laboratory,  it  was  found  that  mammalian 
viruses  and  rickettsiae  could  be  purified 
by  means  of  cellulose  ion  exchange.  Dur- 
ing the  summer  of  1957,  poliomyelitis, 
type  2,  virus  was  grown  in  human  tissue 
culture  cells  and  purified  of  host  material 
by  a  single  passage  over  a  bed  of  the  cellu- 
lose anion-exchanger  "ECTEOLA."  A 
report  of  this  work,  which  also  includes 
subsequent  investigations  by  Dr.  Hoyer 
and  his  associates,  has  already  appeared  in 
print  (Hoyer,  Bill  H.,  Ellis  T.  Bolton, 
Richard  A.  Ormsbee,  George  LeBouvier, 
Daniel  B.  Ritter,  and  Carl  L.  Larson, 
Mammalian  viruses  and  rickettsiae,  Sci- 
ence,  727,859-863,1958). 

Two  other  examples  of  virus  chroma- 
tography are  shown  in  figure  28.  These 
chromatograms  were  obtained  as  follows: 
P32-labeled  tissue  cultures  (monkey  kidney 
cells)  were  allowed  to  synthesize  virus  in 
the  presence  of  P32-labeled  orthophosphate. 
When  the  host  cells  had  lysed,  the  radio- 
active viruses  were  harvested  by  ultracen- 
trifugation.  The  sediment  at  the  bottom 
of  the  centrifuge  tube  was  taken  up  in  a 
small  volume  of  dilute  neutral  phosphate 
buffer    and    loaded    onto    a    column    of 

ECTEOLA.  Buffered  salt  solutions  were 
then  passed  over  the  exchanger,  the  effluent 
being  collected  in  separate  test  tubes  at 
each  addition  of  buffer.  The  contents  of 
the  test  tubes  were  assayed  for  P32  radio- 
activity and  for  infectivity.  The  resulting 
data  are  summarized  in  figure  28.  Only 
a  small  part  of  the  initial  P32  was  eluted 
in  regions  which  contained  essentially  all 
the  viruses  that  had  been  loaded  onto  the 
exchanger.  Protein  was  barely  detectable 
even  by  the  most  sensitive  chemical  meth- 
ods available.  Thus,  a  high  degree  of  virus 
purification  could  be  obtained  with  great 
rapidity  and  relative  ease.  Such  a  procedure 
should  prove  useful  for  the  production 
of  the  purified  infectious  agents  required 
for  careful  physicochemical  investigations, 
and  also  for  the  production  of  vaccines 
free  of  unwanted  host  materials,  which 
sometimes  induce  deleterious  "side  effects" 
such  as  the  allergic  encephalitis  caused  by 
certain  vaccines,  rabies  in  particular. 

The  separation  of  viruses  from  host  ma- 
terials, and  from  one  another  as  figure  28 
shows,  suggests  that  the  ion-exchange  proc- 
ess may  also  find  application  in  problems 
requiring  virus  classification  and  identifi- 

The  viruses  studied  to  date  exhibit  a 
chromatographic  behavior  differing  from 
that  of  host  nucleic  acid,  virus  nucleic 
acid,  or  host  nucleoprotein.  Since  viruses 
themselves  are  nucleoproteins,  the  chro- 
matographic behavior  is  probably  deter- 
mined by  the  nature  of  the  protein  moiety. 

Formaldehyde  treatment,  as  commonly 
used  for  the  production  of  vaccines,  causes 
the  viral  protein  to  take  on  a  more  anionic 
character  than  it  previously  had,  and  the 
virus  becomes  attenuated  or  even  "killed." 
The  feasibility  of  separating  the  formalde- 
hyde-modified virus  ("dead"  ones)  from 
unmodified  ("live"  ones)  by  means  of  ion 
exchangers  has  also  been  demonstrated  for 
polio  virus,  type  2.  In  this  case  formalde- 
hyde-treated viruses  are  held  to  the  anion 
exchanger  more  stronglv  than  untreated 


viruses.  Thus,  the  two  types  can  be  sepa- 
rated (Hoyer,  B.  H.,  R.  A.  Ormsbee, 
Ellis  T.  Bolton,  and  D.  B.  Ritter,  Demon- 
stration of  formaldehyde  effect  on  viruses, 
Federation  Proc,  17,  517,  1958).  Although 


Considerable  quantities  of  certain  amino 
acid  analogs  may  be  incorporated  into  the 
proteins  of  Escherichia  coli.   The  analogs 


O.IM   NaCI 

Fig.  28.    Chromatography  of  animal  viruses  on  ECTEOLA-cellulose.    The  viruses  were  labeled 
with  P32.   Each  tube  of  effluent  was  assayed  for  radioactivity  and  infectivity. 

it  is  too  early  to  judge  whether  the  ion- 
exchange  method  will  contribute  in  a 
practical  way  toward  the  production  of 
safe  vaccines,  its  simplicity,  economy,  and 
efficiency  recommend  it  as  an  adjunct  to 
the  biologists'  tools  for  the  production  of 
highly  purified  infectious  agents. 

substitute  for  corresponding  naturally  oc- 
curring amino  acids  and  cause  various 
biological  effects.  In  general,  cellular 
growth  becomes  linear,  and  specific  en- 
zymatic functions  may  be  lost,  be  sup- 
pressed, or  remain  unaffected.  Such  effects 
depend  upon  the  degree  and  kind  of  sub- 


stitution  produced.  Since  the  degree  and 
kind  of  substitution  can  be  controlled, 
analogs  provide  a  means  for  the  quantita- 
tive examination  of  the  relationship  be- 
tween altered  molecular  structure  and 
enzymatic  activity.  Evidence  can  also  be 
adduced  concerning  susceptibility  of  bac- 
terial protein  types  to  analog  substitution. 
The  demonstration  that  amino  acid  ana- 
logs could  be  incorporated  into  bacterial 
proteins   immediately  raised  many   ques- 

The  analog,  norleucine,  substitutes  for 
methionine  in  the  proteins  of  E.  coli.  A 
reduction  of  38  per  cent  of  the  protein 
methionine  is  obtained  when  the  methio- 
nine-requiring  mutant  (ML  304d)  is 
grown  in  C  medium  containing  DL-nor- 
leucine  (2  X  10"2  M)  and  S35  L-methionine 
(10~4  M).  This  mutant  was  chosen  in 
order  to  eliminate  competitive  reactions 
involving  sulfur  compounds  other  than 
methionine  or  the  methionine  analog.  The 

50         60         70  80 

Fraction  number 

Fig.  29.  Elution  pattern  of  bacterial  extract  of  E.  coli  obtained  with  a  DEAE  ion-exchange  col- 
umn. Mutant  cells  (ML  304d)  grown  in  C  medium  containing  S35  L-methionine  (10-3  M),  thio- 
methyl  p-^-galactoside  (5  X  10~4  M),  and  maltose. 

tions  about  the  nature  of  the  proteins  pro- 
duced. In  collaboration  with  Dr.  Georges 
N.  Cohen,  of  the  Pasteur  Institute,  Paris, 
France,  and  H.  de  Robichon-Szulmajster, 
National  Institutes  of  Health,  U.S.P.H.S., 
Bethesda,  Maryland,  investigations  were 
carried  out  to  determine  whether  the  ana- 
logs are  contained  in  radically  different 
molecular  species  or  in  proteins  similar  to 
those  normally  synthesized.  These  investi- 
gations required:  (a)  an  analog  that 
would  substitute  for  only  one  naturally 
occurring  amino  acid,  and  (b)  a  quantita- 
tive method  for  analyzing  bacterial  pro- 

separation  of  bacterial  proteins  into  chro- 
matographically  resolvable  "protein  classes" 
was  achieved  through  the  use  of  the 
DEAE-cellulose  ion-exchange  column. 
Figure  29  shows  the  elution  pattern  of  an 
extract  of  E.  coli  grown  in  C  medium  con- 
taining S35  methionine.  Thiomethyl  fi-d- 
galactoside  was  added  to  induce  the  syn- 
thesis of  (3-galactosidase. 

The  bacterial  extract  was  prepared  from 
washed  cells,  ruptured  by  extrusion 
through  a  small  orifice  under  pressure,  the 
extruded  material  being  centrifuged  to  re- 
move whole  cells  and  large  cellular  frag- 
ments.   The  opalescent  supernatant   was 


then  used  for  the  column  analysis.  Evident 
in  this  elution  diagram  are  a  number  of 
well  resolved  regions  showing  a  close  corre- 
lation between  the  protein  pattern  (meas- 
ured by  the  Folin  reagent)  and  the  pattern 
of  distribution  of  the  radiomethionine. 
Two-dimensional  paper  chromatograms  of 
hydrolysates  of  an  aliquot  of  the  bacterial 
extract  showed  that  the  incorporated  radio- 
activity was  contained  solely  as  methionine. 
In  figure  30  are  shown  the  specific  radio- 

are,  of  course,  contained  in  this  region. 
Nevertheless,  the  partitioning  of  bacterial 
proteins  into  protein  classes  is  apparent. 

Column  analysis  of  bacterial  extracts  of 
cells  grown  in  C  medium  containing  dl- 
norleucine  (2  X  10-2  M)  and  S35  L-methio- 
nine  (10~4  M)  gave  elution  patterns  similar 
to  figure  29.  A  significant  difference  was 
a  uniform  reduction  in  the  specific  radio- 
activities of  these  bacterial  proteins  com- 
pared with  those  of  the  control  experiment 



1.60  - 

h-    1.40  - 

U    1.20 



9  ioo 


V     .80 





ft  /)          A 



















1            1            1            1            1            1            1            1            1            1            1            1 


20        30 

40       50        60        70        80 

90       100       110        120 

Fig.  30.  Specific  radioactivity  of  eluted  column  fractions.  Data  represent  the  ratio  of  radioactivity 
per  fraction  to  the  quantity  of  protein  newly  synthesized  after  the  addition  of  the  labeled  methio- 
nine to  the  culture. 

activities  of  the  individual  fractions.  The 
specific  radioactivity  is  the  ratio  of  the 
quantity  of  radioactivity  to  the  amount  of 
newly  synthesized  proteins.  This  figure 
demonstrates  that  the  methionine  content 
varies  among  the  protein  classes  resolved. 
Figure  31  shows  the  degree  of  resolution 
among  the  eluted  proteins.  Superimposed 
on  the  elution  diagram  are  the  locations 
of  three  enzyme  activities:  3-galactosidase 
(LAC),  phosphoglucomutase  (MUT), 
and  glucose-6-phosphate  dehydrogenase 
(ZW).  Each  enzyme  activity  is  correlated 
with  a  well  resolved  protein  peak.  Other 
proteins  having  similar  charge  properties 

(fig.  17).  The  existence  of  certain  mark- 
ers (peaks,  valleys,  enzymes,  etc.)  along 
the  elution  diagram  allows  a  quantitative 
comparison,  marker  for  marker,  among 
several  column  runs.  Figure  32  shows  the 
specific  radioactivities  of  seven  well 
marked  and  separated  regions  obtained 
with  the  norleucine-grown  cells.  These 
are  compared  with  the  same  regions  in  the 
control  experiment,  where  the  specific 
radioactivity  of  each  region  was  arbitrarily 
chosen  to  equal  100. 

The  regions  compared  in  figure  19  were  : 
two  well  resolved  and  isolated  protein 
peaks    A    and    B,    the    ribonucleoprotein 


peak,  the  three  peaks  of  enzyme  activity 
(LAC,  MUT,  and  ZW)  easily  measurable 
in  both  experiments,  and  region  C  shown 
in  figure   16.    Each  point  represents  the 

arithmetical  mean  of  the  specific  radioac- 
tivity of  the  maximum  peak  sample  and 
the  two  samples  immediately  preceding 
and  following  this  peak. 


30       40        50       60       70        80 

90      100     110 


Fig.  31.    Location  of  enzyme  activities  along  elution  diagram.  Glucose-6-phosphate  dehydrogen- 
ase (ZW) ;  phosphoglucomutase  (MUT) ;  (3-galactosidase   (LAC) . 

Figure  33  shows  an  elution  pattern  ob- 
tained from  cells  grown  in  DL-norleucine 
1-C14  (2X10-2  M)  and  nonradioactive 
L-methionine  (10~4  M).  In  this  experi- 
ment there  was  a  43  per  cent  substitution 
of  norleucine  for  methionine  in  the  bac- 
terial proteins.  Radioautographs  of  two- 
dimensional  paper  chromatograms  of  hy- 
drolysates  of  the  bacterial  extract  showed 
one  radioactive,  ninhydrin-positive  spot 
having  the  same  Rf  as  found  with  the 
labeled  norleucine  used  in  this  experiment. 

There  is  a  great  deal  of  similarity  in  the 
elution  diagrams  obtained  from  the  S35 
methionine  and  the  C14  norleucine  labeled 
cells.  One  significant  difference,  however, 
occurs  in  the  first  major  peak  of  the  elution 
diagrams.  In  these  early  fractions  of  eluted 
material  are  contained  the  nonprotein 
amino  acids  (or  analogs)  concentrated  by 
the  cell  from  the  environment.  The  quan- 
tity of  "free  amino  acids"  depends  upon 
their  external  concentrations,  and  in  these 


A         ZW    MUT  8 














/         *            V           .       * 









1                  1                  1 




40  50  60  70  80  90 

Fraction      number 

Fig.  32.  Comparison  of  specific  radioactivities 
of  definite  regions  along  elution  diagrams  ob- 
tained from  cells  grown  in  C  medium  contain- 
ing S35  L-methionine  (10~3  M)  (control)  and 
from  cells  grown  in  C  medium  containing  S35 
L-methionine  (10~4  M)  plus  DL-norleucine 
(2  X  10~2  M).  Linear  growth  was  obtained  in 
the  latter  culture,  and  the  cells  were  harvested 
for  analysis  after  more  than  a  doubling  of  bac- 
terial mass. 



experiments  the  ratio  of  C14  L-norleucine 
to  S35  L-methionine  in  the  media  was  100 
to  1.  Chemical  fractionation  of  the  eluted 
fractions  showed  that  TCA-soluble  ma- 
terial ("free  amino  acids")  was  mainly 
contained  in  the  first  20  samples  and 
dropped  rapidly  to  a  few  per  cent  by  the 
thirty-fifth  sample,  remaining  low  for  the 
rest  of  the  elution  process.  It  has  also  been 

nine  and  norleucine  (figs.  29  and  33)  also 
eliminates  the  hypothesis  that  only  certain 
proteins  are  susceptible  to  analog  substitu- 
tion. Indeed,  figure  32  demonstrates  that 
the  analog  is  incorporated  into  all  the  pro- 
teins and  in  the  same  proportion.  Each 
methionine  incorporation  site  thus  seems 
to  have  an  equal  probability  of  analog 
substitution.    The   formation   of   a   large 


Fig.  33.  Elution  pattern  of  bacterial  extract  of  1-C14  DL-norleucine-grown  cells.  Mutant  cells 
(ML  304d)  grown  for  more  than  a  doubling  in  C  medium  containing  labeled  norleucine 
(2  X  104  M)  and  S32  L-methionine  (10~4  M). 

noted  that  the  quantity  of  material  con- 
tained in  the  ribonucleoprotein  region 
varies  from  one  column  run  to  another, 
and,  if  the  cells  are  ruptured  in  media  con- 
taining phosphate  buffer,  or  in  buffer 
containing  no  magnesium,  this  ribonucleo- 
protein is  not  seen  at  all. 

The  above  results  demonstrate  that  most 
of  the  proteins  formed  in  the  presence  of 
the  analog  are  not  radically  different  mo- 
lecular species  but  are  physicochemically 
similar  to  the  proteins  normally  synthe- 
sized. The  similarity  of  the  elution  dia- 
grams obtained  with  the  labeled  methio- 

quantity  of  uncompleted  proteins,  caused 
by  the  joining  of  the  analog  by  a  peptide 
bond  to  one  of  its  neighboring  amino 
acids  but  not  to  the  other,  does  not  seem 
to  be  a  probable  event.  Should  such  un- 
finished molecules  be  present,  they  would 
markedly  alter  the  elution  patterns  ob- 
tained after  the  analog  is  incorporated. 

These  conclusions  are  strengthened  by 
data  obtained  in  experiments  using  other 
amino  acid  analogs.  Figure  34  shows  the 
elution  diagram  obtained  from  wild-type 
E.  coli  (ML  30)  grown  in  C  medium  con- 
taining C14  3-DL-phenylalanine   (10-4  M) 



and  DL-p-fluorophenylalanine  (5  X  10~3  M) . 
At  these  concentrations  there  is  approxi- 
mately a  50  per  cent  substitution  of  the 
analog  for  protein  phenylalanine,  and 
linear  growth  occurs.  The  elution  diagram 
obtained  (fig.  34)  appears  very  similar  to 
that  from  normal  cells  (fig.  29).  There  is 
no  evidence  of  different  types  of  protein 
classes  being  formed  as  a  result  of  analog 

Figure  35  demonstrates,  within  the  lim- 
its of  resolution  of  the  column  and  a  low 


a  20 

C    -Phenylalanine 
I  l  l  I  I  I  ) 

20       40        60       80       100       120      140      160 
Fraction     number 

Fig.  34.  Elution  pattern  of  an  extract  of  cells 
grown  in  the  presence  of  C14-3-DL-phenylalanine 
and  DL-p-fluorophenylalanine. 

specific  radioactivity  of  tracer,  that  /7-fluoro- 
phenylalanine  is  incorporated  into  all  the 
bacterial  proteins.  This  elution  diagram 
was  obtained  from  wild-type  E.  coli  grown 
in  C  medium  containing  C14  3-DL-p- 
fluorophenylalanine  (5xl0~3  M). 

The  use  of  another  amino  acid  analog 
gave  results  which  in  every  respect  confirm 
and  augment  the  conclusions  cited  above. 
Selenomethionine  completely  substitutes 
for  the  methionine  of  the  bacterial  protein. 
With  this  uniform  replacement  exponen- 
tial growth  was  observed  and  the  induc- 
tion and  synthesis  of  active  [3-galactosidase 
demonstrated.  The  constitutive  enzymes 
essential  for  exponential  growth  were  obvi- 
ously present  in  active  forms.  Under  these 

conditions  there  can  be  little  doubt  that 
active  altered  proteins  are  synthesized, 
having  biological  as  well  as  physicochemi- 
cal  properties  similar  to  those  of  the  nor- 
mal cell. 

The  use  of  amino  acid  analogs  other 
than  selenomethionine  has  always  resulted 
in  linear  growth  of  the  cells  whenever 
analog  substitution  in  the  protein  was  evi- 
dent. Thus,  it  might  be  argued  that  at 
least  one  growth-rate-limiting  enzyme  was 
unusually  susceptible  to  analog  substitu- 
tion and  that  the  enzyme  was  synthesized 
at  a  reduced  rate  if  at  all.    On  the  other 

20       40        60       80       100      120      140      160      180 

Fraction    number 

Fig.  35.  Elution  pattern  of  an  extract  of  cells 
grown  in  the  presence  of  C14-3-DL-p-fluorophen- 

hand,  analog  incorporation  might  result 
in  the  synthesis  of  protein  molecules — but 
these  would  be  proteins  without  the  ca- 
pacity for  enzymatic  function.  This  elimi- 
nation or  suppression  of  enzyme  activity 
would  depend  on  the  degree  and  kind  of 
substitution  involved  and  on  the  amino 
acid  composition  of  the  sites  of  enzyme 
action.  Some  evidence  supporting  the  lat- 
ter hypothesis  has  accumulated,  and  the 
investigation  of  this  question  is  currently 
under  way. 


Systems  free  of  complete,  undamaged 
cells  have  frequently  been  tested  for  their 
ability  to  synthesize  protein.  One  goal  in 



using  such  systems  is  to  find  the  minimum 
number  of  carefully  purified  and  charac- 
terized components  necessary  for  protein 
synthesis.  A  much  more  limited  endeavor 
is  the  use  of  fractured  cells  in  which  the 
components  may  be  selectively  destroyed 
by  enzymes.  A  number  of  these  systems 
have  been  tested  in  this  laboratory  by 
means  of  radioactive-tracer  techniques, 
which  are  extremely  sensitive.  Unfortu- 
nately, we  have  not  yet  found  any  cell-free 
system  that  gives  reproducible  and  con- 
vincing incorporation  of  amino  acids  into 

Various  mixtures  of  purified  ribosomes 
with  soluble  proteins  were  tested.  These 
preparations  could  be  made  assuredly  cell- 
free  by  repeated  high-speed  centrifugation. 
No  sign  of  incorporation  of  amino  acids 
of  high  specific  radioactivity  (12  per  cent 
C14)  was  observed  in  any  of  the  tests. 
When  cell  walls  were  added  to  the  mix- 
tures, or  tested  alone,  incorporation  was 
observed.  Furthermore,  it  was  the  large 
fragments  of  cell  walls  that  gave  the  great- 
est incorporation.  In  these  preparations 
whole  cells  were  not  readily  observable  in 
the  wet  material  under  the  phase  con- 
trast microscope.  Nevertheless,  slides 
stained  with  gentian  violet  to  show  whole 
cells  invariably  revealed  their  presence  in 
sufficient  quantity  to  account  for  the  ob- 
served incorporation. 

In  last  year's  report  the  formation  of 
"protomorphs" — large  globular  structures 
that  form  in  clear  solutions  of  cell  ex- 
tracts— was  described.  Their  formation 
and  their  incorporation  of  amino  acids 
were  erratic.  During  this  report  year  some 
of  the  causes  of  variation  have  been  deter- 
mined. The  formation  process  is  sensi- 
tive to  concentration.  A  twofold  dilution 
of  the  usual  pressure  cell  juice  is  suffi- 
cient to  prevent  formation,  but  cell  juices 
prepared  by  grinding  or  other  processes 
that  preserve  intact  DNA  will  form  proto- 
morphs from  much  more  dilute  solutions. 
Furthermore,  the  addition  of  DNAase  to 
the  solution  invariably  prevents  proto- 
morph  formation.  DNA  is  one  of  the  com- 

ponents of  protomorphs,  and  it  seems  to 
be  an  essential  one.  Phosphate  is  also  es- 
sential. When  the  cells  are  very  thoroughly 
washed,  protomorphs  will  not  form  unless 
P04  is  added  back  to  10"3  M.  The  quan- 
tity of  intact  DNA  could  easily  vary,  de- 
pending on  the  conditions  in  the  pressure- 
cell  orifice,  and  the  quantity  of  PCX  could 
vary  with  the  conditions  of  washing. 
These  factors  seem  to  be  sufficient  to  ac- 
count for  the  observed  variations  in  forma- 
tion of  protomorphs. 

The  incorporation  of  amino  acids  by 
protomorphs  was  at  times  convincing.  In 
particular,  the  incorporation  increased 
with  increased  concentration  of  amino  acid 
or  upon  supplementation  with  ATP  (con- 
ditions that  should  not  affect  intact  cells). 
Also,  the  pattern  of  amino  acid  incorpora- 
tion did  not  resemble  that  of  whole  cells. 

Unfortunately,  the  incorporation  by  the 
protomorphs  was  also  highly  erratic.  One 
series  of  tests  showed  that  the  capacity  of 
protomorphs  to  incorporate  amino  acids 
depended  on  the  presence  of  cell-wall  frag- 
ments. Again  it  seems  possible  to  ascribe 
the  observed  incorporation  to  whole  cells 
hidden  from  view  by  the  protomorphs. 

As  a  consequence  of  these  equivocal  re- 
sults, work  with  cell-free  systems  has  pro- 
ceeded only  occasionally.  The  cell  wall 
or  cell  membrane  appears  to  be  the  most 
likely  fragment  to  show  true  synthetic 
activity,  but  it  also  is  the  most  likely  to  be 
contaminated  by  intact  cells.  The  goal  of 
reconstituting  a  synthesizing  system  from 
purified  components  still  seems  very  re- 


For  several  years  the  ability  of  E.  coli 
to  concentrate  low-molecular-weight  com- 
pounds has  been  the  subject  of  intensive 
study.  In  the  past  year  our  experimental 
activities  shifted  to  other  aspects  of  the 
mechanisms  of  macromolecular  synthesis. 
Measurements  of  the  exchange  rate  of  the 
proline  pool  at  0°  C  provide  the  only  new 
experimental  information,  but  the  impli- 


cations  of  the  large  body  of  previously 
obtained  experimental  results  for  the 
models  of  the  pool-forming  process  are 
now  fairly  well  understood.  It  has  become 
clear  that  neither  the  simple  "permease" 
model  nor  the  "stoichiometric  site"  hy- 
pothesis is  sufficient.  A  model  in  which  a 
limited  quantity  of  carriers  is  utilized  to 
place  the  pool  compounds  on  sites  appears, 
however,  to  be  sufficient  to  explain  prac- 
tically all  the  observations. 

Exchange  Rate  of  the  Proline  Pool  at  0°  C 

Observations  of  exchange  between 
amino  acids  present  externally  and  those 
in  the  pool  can  be  conveniently  carried  out 
under  a  variety  of  conditions  by  means  of 
labeled  compounds.  If  they  are  carried  out 
at  the  normal  temperatures  used  for 
growth  and  pool  formation  (20°  to  40°  C) 
the  rates  of  pool  formation,  protein  incor- 
poration, and  exchange  are  all  comparable. 
Also,  in  quasi  steady-state  conditions,  the 
pool  size  and  external  concentration  can- 
not be  varied  independently.  At  0°  C, 
however,  the  pool  formation  and  protein 
incorporation  rates  are  very  much  more 
strongly  suppressed  than  the  exchange  rate 
(as  reported  in  Year  Book  56).  As  a  re- 
sult the  pool  size  will  not  show  significant 
change  in  several  hours,  even  when  it  is 
not  in  equilibrium  with  the  external  con- 
centration, and  this  is  sufficient  time  for 
exchange  equilibrium  to  occur  between 
the  external  and  pool  amino  acids.  Thus, 
exchange  studies  can  be  carried  out  over 
a  wide  variety  of  independently  varied 
concentrations  and  pool  sizes. 

Since  pool  formation  is  very  slow  at 
0°  C,  pools  of  a  desired  size  are  formed 
with  unlabeled  proline  at  25°.  The  sus- 
pension is  then  chilled  to  0°  C.  The  cells 
are  centrifuged  and  resuspended  in  buffer, 
and  after  an  hour  or  so  at  0°  C,  C14  proline 
is  added.  Since  the  external  quantity  is 
small  compared  with  the  amount  in  the 
pool,  a  very  efficient  labeling  of  the  pool 
by  exchange  is  achieved.  After  a  steady 
state  (equal  internal  and  external  specific 
activities)  is  reached,  the  external  concen- 

tration is  brought  up  to  a  chosen  value 
by  adding  C12  proline.  The  time  course 
of  exchange  is  then  followed  by  measuring 
the  loss  of  TCA-soluble  radioactivity  from 
the  cells.  Typical  measurements  under 
these  conditions  are  shown  in  the  curve  of 
figure  36. 

When  the  results  of  such  experiments 
over  a  wide  range  of  concentrations  and 
pool  sizes  are  examined  it  is  found  that 
the  time  course  of  exchange  can  be  re- 

Fig.  36.  Time  course  of  exchange  of  the  pro- 
line pool  at  0°  C.  Cells  with  a  C14  proline  pool 
at  1.5  X  10-6  mole  per  wet  gram  were  prepared 
as  described  in  the  text.  At  time  zero,  C12  pro- 
line was  added  to  a  concentration  of  7  X  10-4  M. 

solved  into  two  components  with  widely 
different  time  constants.  Figure  37  shows 
the  dependence  of  the  rate  of  exchange 
calculated  for  each  of  these  components 
as  a  function  of  the  pool  size.  The  figure 
beside  each  point  is  the  external  concen- 
tration during  exchange  in  micromoles 
per  liter.  It  appears  that  the  exchange  rate 
is  independent  of  the  external  concentra- 
tion except  possibly  at  low  concentrations. 
The  rapidly  exchanging  component  of 
the  pool  is  always  smaller  than  the  slowly 
exchanging  one.  It  appears  to  saturate  at 
less  than  1.0  micromole  per  gram  wet  cells 
and  is  not  easily  observable  when  the  total 
pool  is  greater  than   10  micromoles  per 


gram.  Such  exchange  experiments  are 
difficult  to  perform  since  the  larger  pools 
appear  to  be  unstable  at  0°  C. 

The  exchange  rate  of  the  large,  slow 
component  is  roughly  proportional  to  the 
total  pool  size.  Unfortunately,  the  accu- 
racy of  the  data  is  not  quite  sufficient  to 
determine  whether  the  exchange  rate  of 

owing  to  the  simplicity  of  exchange  proc- 
esses. If  a  single  homogeneous  component 
of  the  pool  exchanges  with  external  amino 
acid,  the  time  course  of  the  process  must 
follow  a  simple  exponential  represented  by 
a  single  time  constant.  This  statement 
holds,  regardless  of  the  nature  and  multi- 
plicity of  the  mechanisms  mediating  the 






^   ' 



^\  1000 
X690^/          *|00 

*  / 

X870                                                     , 


/  »950 


*6                 /                                                           / 

2  2         /                                                     1000/ 

35/                                      yVoo 

/  x690                                                   / 

/                                                  /        •  10.000 
/        X2                                                / 







690  / 


1                  1            1         1                              1                              1 

1            1 

1                              • 

10  2.0  4.0  6.0         8.0    10.0 


Fig.  37.  Rate  of  exchange  as  a  function  of  pool  size  (log  log  plot).  The  points  were  obtained 
from  experiments  such  as  that  shown  in  figure  14  by  fitting  the  time  course  of  exchange  to  curves 
derived  from  the  sum  of  two  exponential  decays.  The  numbers  beside  each  point  are  the  external 
concentrations  during  exchange  in  micromoles  per  liter.  The  straight  line  shown  would  result  if 
the  exchange  rate  were  proportional  to  pool  size. 

each  of  the  components  is  proportional  to 
its  own  size,  although  this  result  is  sug- 
gested by  the  evidence. 

In  addition,  it  is  not  certain  that  there 
are  only  two  components,  although  there 
are  clearly  more  than  one,  since  none  of 
the  exchange  curves  fits  a  simple  expo- 
nential. These  results  show  that  the  pro- 
line of  the  pool  is  held  in  the  cell  by  more 
than  one  mechanism,  or  at  least  with  vary- 
ing degrees  of  affinity. 

The  evidence  for  this  statement  is  good, 

exchange,  as  long  as  the  quantity  of  amino 
acid  associated  with  intermediate  steps  is 
negligibly  small  compared  with  the  size 
of  the  pool.  The  significance  of  these  re- 
sults will  be  discussed  below. 

Description  of  Models 

Figures  38  and  39  describe  three  models 
for  the  pool-forming  process.  The  first 
(fig.  38)  is  more  or  less  equivalent  to  the 
"permease"  model  suggested  by  Monod 
and  Cohen.  It  is  assumed  that  there  is  an 



association  of  the  external  amino  acid  with 
a  specific  transport  mechanism.  This 
mechanism,  being  coupled  with  a  source 
of  energy,  can  form  an  association  with 
the  amino  acid  in  the  dilute  external 
medium  and  then  release  it  into  a  region 

site-amino  acid  complex.  The  pool  satu- 
rates when  all  the  sites  for  the  particular 
amino  acid  are  occupied.  The  process  of 
dissociation  of  the  site  must  also  be  coupled 
to  the  energy  supply  since  the  pools  are 
maintained  in  the  absence  of  an  energy 


compound  =  A 


Pool  =  P 





->  P  +  Y 

2  Energy  coupled 

Fig.  38.    Permease  model. 

■>  A 

of  high  concentration  within  the  cell.  In 
order  that  the  pool  size  saturate  at  high 
external  concentration  it  is  assumed  that 
a  separate  mechanism  allows  the  pool  com- 
pound to  leak  out  of  the  cell.  Thus  at 
high  external  concentrations  the  concen- 
trating mechanism  becomes  saturated. 
When  its  maximal  rate  is  balanced  by  the 
leak  rate  the  pool  can  no  longer  be  in- 
creased. According  to  the  permease  model 
the  pool  is  maintained  by  means  of  a  circu- 
lating flow — a  sort  of  dynamic  steady  state. 
This  feature  is  borne  out  by  experiments 
showing  rapid  exchange  during  pool 

The  upper  model  indicated  in  figure  39 
is  the  simplest  site  model,  referred  to  by 
Monod  and  Cohen  as  the  "stoichiometric 
site  hypothesis."  It  is  assumed  that,  by 
means  of  an  unspecified  mechanism 
coupled  to  the  energy  supply  of  the  cell, 
free  amino  acid  diffusing  through  the  cell 
is  caused  to  form  a  labile  bond  with  a 
specific  site.  The  relationship  between  the 
pool  size  and  the  external  concentration 
is  maintained  by  a  balance  between  the 
rates  of  formation  and  dissociation  of  the 

supply  (although  they  are  not  formed  un- 
der these  conditions) .  It  must  be  assumed 
that  exchange  occurs  by  a  separate  process. 
The  lower  model  shown  in  figure  39  is 
entitled   the   "site   plus   carrier,"    since   a 

compound  =  A 

Pool  =  AR 

"Stoichiometric  site" 

+  R     <        \      > 


'Site  plus  carrier" 




AY  +  R 




Fig.  39.     Site  models. 


carrier  has  been  postulated  in  order  to 
account  for  a  wide  variety  of  observations 
of  the  pool-forming  process.  A  carrier- 
amino  acid  complex  is  formed  without  the 
participation  of  the  cell's  energy-supplying 
mechanisms.  The  carrier  complex  (AY) 
then  reacts  with  a  free  site  to  form  a  site 
complex  (AR)  and  free  the  carrier.  The 
free  carrier  may  in  turn  react  with  the  site 
complex  in  the  reverse  reaction  and  form 
AY.  Both  these  reactions  must  be  coupled 
to  the  energy  supply  of  the  cell  in  order 
that  the  pool  be  maintained  in  the  absence 
of  an  energy  supply.  A  given  carrier  is 
present  in  limited  quantity  and  is  specific 
for  a  particular  amino  acid.  The  carriers 
and  carrier-amino  acid  complexes  must  be 
free  to  diffuse  through  the  cell  in  order  to 
reach  and  react  with  the  sites  or  site  com- 
plexes. It  is  also  presumed  that  in  the 
presence  or  absence  of  an  energy  supply 
an  exchange  reaction  occurs: 

A*Y  +  AR  ^  AY  +  A#R 

The  carrier  indeed  carries  a  heavy  bur- 
den of  properties  that  may  be  conceptually 
difficult.  Nevertheless  the  resulting  model 
is  simple  enough  to  be  mathematically 
analyzed  in  detail  and  predicts  the  princi- 
pal properties  of  the  pool. 

Summary  of  Pool  Properties  and 
Implications  of  the  Models 

Table  8  lists  in  summary  form  the  princi- 
pal observations  of  the  behavior  of  the 
proline  pool.  In  the  columns  at  the  right 
the  score  for  each  of  the  models  is  listed, 
a  plus  sign  indicating  that  the  model  satis- 
factorily explains  the  observation  and  a 
minus  sign  that  there  is  a  contradiction  or 
that  a  modification  may  be  required  by 
the  experimental  evidence. 

Since  most  of  these  observations  have 
been  discussed  in  the  annual  reports  of 
previous  years,  no  attempt  will  be  made 
here  to  follow  in  detail  the  arguments 
leading  to  these  scores,  except  for  the  new 
evidence  concerning  the  exchange  rate 
at  0°  C. 

As  has  already  been  mentioned,  the  0° 
exchange  data  show  that  there  is  more 
than  one  component  in  the  proline  pool. 
Items  13  and  16  of  table  8  also  indicate 
that  there  are  different  mechanisms  for 
the  maintenance  of  compounds  in  the  pool. 
This  evidence  may  be  incorporated  in  the 
site  models  by  assuming  that  there  are 
sites  of  different  types  or  affinities  within 
the  cell. 

In  order  to  incorporate  this  evidence 
into  the  permease  model  it  might  be  as- 
sumed that  there  are  internal  regions 
within  the  cell  having  different  degrees 
of  accessibility  to  the  external  amino  acid 
(to  explain  the  exchange  data)  and  differ- 
ent sensitivities  to  osmotic  shock.  Alterna- 
tively, it  might  be  assumed  that  sites 
within  the  cell  retain  part  of  the  pool 
while  the  entry  mechanism  is  still  that 
of  the  permease  model. 

If  the  arguments  are  followed  in  detail 
it  is  found  that  the  addition  of  internal 
sites  to  the  permease  model  does  not  an- 
swer the  difficulties  arising  from  items 
2,  3,  4,  and  7  of  table  8,  although  it  does 
help  as  far  as  items  8  and  16  are  concerned. 
The  very  existence  of  a  high  rate  of  ex- 
change (when  pools  can  only  be  formed 
slowly)  raises  serious  difficulties  for  the 
permease  model  (items  3,  4,  and  7).  It 
would  be  necessary  to  specify  what  features 
would  be  added  to  this  model  to  allow 
for  exchange  and  maintenance  before  the 
rates  of  exchange  could  be  discussed. 

The  experimental  results  of  Cohen  and 
Monod  at  the  Institut  Pasteur  on  the  galac- 
toside-concentrating  system  of  E.  coli  are 
almost  entirely  consistent  with  the  perme- 
ase model.  In  fact,  it  was  these  results 
that  led  them  to  postulate  the  model.  They 
do,  however,  report  one  observation  which 
has  a  similar  consequence  to  that  of  items 
7  and  4  of  table  8.  The  concentration  of 
galactosides  is  blocked  by  the  lack  of  an 
energy  source  or  by  azide.  But  in  whole 
cells  the  rate  of  splitting  of  orthonitro- 
phenylgalactoside  by  the  internal  enzyme 
3-galactosidase  is  not  reduced  under  either 
of  these  conditions.  Thus,  although  under 


these  conditions  pools  may  not  be  formed, 
galactosides  may  enter  the  cell  at  a  rapid 
rate  when  the  energy  supply  for  the 
"Y  mechanism"  of  the  permease  model 
has  been  blocked. 

Both  the  site  models  predict  that  the  ex- 
change rate  should  be  proportional  to  the 

The  quantitative  observations  corre- 
sponding to  items  10,  11,  and  12,  and  the 
relationship  between  pool  size  and  external 
concentration,  supply  five  independent  re- 
lations between  the  constants  of  the  "site 
plus  carrier"  model.  The  total  number 
of  sites  is  known  from  the  saturation  value 

TABLE  8.    Relation  of  Observations  of  the  Proline  Pool  of  E.  coli  to  the  Models 





"Site  plus 








































+  * 

+  * 




1  Glucose  required  for  formation 

2  Glucose  not  required  for  maintenance 

3  Pools  maintained  at  0°  C 

4  Pools  formed  slowly  at  0°  C 

5  Pools  may  be  very  large 

6  Exchange  between  external  and  pool  AA  occurs  dur- 
ing formation 

7  Exchange  occurs  at  a  high  rate  in  absence  of  glucose 
or  at  0°  C 

8  Fast  and  slow  components  appear  in  exchange  curves 
at  0°  C 

9  The  0°    C  exchange  rate  saturates   at  low  external 

10  The  0°  C  exchange  rate  increases  with  pool  size 

11  Rate  of  loss  from  pool  in  absence  of  external  AA  is 
slower  than  initial  rate  of  formation 

12  Initial  rate  of  pool  formation  does  not  rise  as  fast  with 
concentration  as  pool  size  does 

13  Small  pools  not  generally  influenced  by  other  A  A 
but  large  pools  are  suppressed 

14  Pools    removed    by    sudden    reduction    in    osmotic 

15  Pools  may  be  immediately  re-formed  after  removal  by 

16  Different  pools  removed  to  different  extent  by  shock 

17  Maximum  pool  size  increases  with  osmotic  strength 
of  medium 


Assuming  that  the  sites  may  be  osmotically  sensitive. 

pool  size.  However,  the  limited  quantity 
of  carrier  postulated  in  the  "site  plus  car- 
rier" model  is  necessary  in  order  that  the 
exchange  rate  be  independent  of  the  ex- 
ternal concentration,  for  moderate  concen- 
trations. The  limited  quantity  of  carrier 
plays  a  role  in  forcing  the  predictions  of 
the  model  to  match  items  11  and  12  of 
table  8.  It  is  also  useful  in  understanding 
the  interactions  between  similar  amino 
acids  reported  in  Year  Book  55. 

of  the  pool.  From  these  relations  the  con- 
stants Ki,  K2,  K3,  K±,  and  K5  have  been 
numerically  evaluated,  except  for  a  con- 
stant factor  which  is  the  reciprocal  of  the 
total  number  of  carriers  present.  With  a 
reasonable  assumption  as  to  the  number  of 
carriers,  all  the  constants  are  consistent 
with  principles  of  chemical  kinetics. 

Thus,  we  have  been  able  to  explain  the 
wide  variety  of  observations  concerning 
the  pool  and  to  approach  an  evaluation  of 


the  reaction  rate  constants  of  the  postu- 
lated steps.  The  equations  which  in  the 
end  represent  the  pool-forming  process 
will  have  to  be  similar  to  those  derived 
from  this  model.  It  is  always  possible, 
however,  that  they  could  be  derived  from 
a  fundamentally  different  model.  Thus, 
the  existence  of  the  carrier  has  not  been 
directly  demonstrated  although  there  is  a 
requirement  for  some  equivalent  mecha- 
nism to  account  for  the  limited  rates  of 
formation  and  exchange. 

Such  a  carrier  may  also  be  important  in 
protein  synthesis  since  it  would  be  a  suit- 
able specific  mechanism  for  the  transport 
of  amino  acids  from  the  pool  sites  to  sites 
that  are  actively  engaged  in  that  process. 


During  the  year  our  collaborative  work 
with  Drs.  L.  B.  Flexner  and  J.  B.  Flexner 
on  protein  synthesis  in  mouse  tissues  has 
continued.  These  experiments  provide  the 
Biophysics  Section  with  an  opportunity  to 
maintain  a  working  knowledge  of  mam- 
malian cells.  It  is  of  the  greatest  interest 
to  observe  the  many  ways  in  which  these 
cells  resemble  microorganisms  and  others 
in  which  there  are  notable  differences. 

The  absolute  quantity  of  pool  amino 
acids  in  the  liver  and  cerebral  cortex  of 
both  newborn  and  adult  mice  was  deter- 
mined (table  9).  In  contrast  to  the  tissue 
proteins,  where  there  is  little  difference  in 
amino  acid  composition  between  adult 
and  newborn,  the  amino  acid  pools  change 
markedly  with  maturation.  These  values 
are  needed  to  calculate  the  flux  of  amino 
acids  from  pool  to  protein. 

Figure  40  shows  the  passage  of  radio- 
active glutamic  acid  from  blood  plasma  to 
tissue  pool  to  tissue  protein  in  the  adult 
mouse.  The  rapid  transfer  from  plasma 
to  cortex  is  of  particular  interest  because 
the  blood-brain  barrier  has  previously  been 
presumed  to  be  impermeable  to  glutamic 
acid  in  the  adult.  Similar  observations 
were  made  on  the  transfer  of  a  group  of 
essential  amino  acids  and  the  utilization  of 

glucose  for  synthesis  of  nonessential  amino 

The  rates  at  which  individual  amino 
acids  are  transferred  to  protein  were  then 
calculated.  If  amino  acids  were  incorpo- 
rated into  protein  solely  as  the  result  of 
synthesis  of  a  complete  molecule,  the 
quantity  of  amino  acid  incorporated  should 
be  proportional  to  the  quantity  in  the 
protein.  On  the  other  hand,  if  amino  acids 
were  incorporated  by  exchange  into  pre- 
existing protein,  the  quantity  should  de- 
pend only  on  the  exchange  rate.  The  ob- 
served rates  of  incorporation  of  different 
amino  acids  are  by  no  means  proportional 
to  their  concentrations  in  the  protein. 
Moreover,  maturation  has  different  effects 
on  the  rate  of  incorporation  of  different 
amino  acids.  For  some,  the  incorporation 
rate  is  higher  in  the  newborn  than  in  the 
adult;  for  others,  it  is  equal  or  less  in  spite 
of  the  close  correspondence  of  the  amino 
acid  composition  of  the  protein  at  the 
two  ages. 

Although  such  results  suggest  that  ex- 
change plays  an  important  part  in  the  in- 
corporation of  amino  acids  by  mammalian 
cells,  there  is  an  alternative  interpretation. 
Only  the  average  composition  of  all  the 
protein  of  the  tissue  was  measured,  and 
the  incorporation  rate  observed  was  an 
average  rate.  It  is,  then,  quite  possible 
that  certain  proteins  of  the  tissue,  which 
are  synthesized  and  degraded  rapidly,  ac- 
count for  most  of  the  incorporation, 
whereas  the  average  composition  is  deter- 
mined by  other  protein. 

This  uncertainty  can  be  resolved  by 
measuring  the  compositions  and  incorpora- 
tion rates  for  individual  isolated  proteins. 
In  addition,  it  is  of  considerable  interest 
to  observe  whether  there  is  a  range  of 
stability  among  the  proteins  of  a  tissue. 
Accordingly,  some  exploratory  experiments 
have  been  made  using  the  DEAE  columns 
described  above  to  separate  out  individual 
proteins  from  different  tissues  of  the 

For  these  experiments  a  much  higher 
level  of  radioactivity  was  needed  to  ob- 


serve  individual  proteins  instead  o£  the 
total.  S3504  (25  millicuries)  was  supplied 
to  a  culture  o£  E.  coli  which  converted 
roughly  60  per  cent  to  S35-labeled  cystine 
and   methionine  of  the  cells.    The  cells 

placed  on  the  column  and  eluted  with  a 
salt  gradient.  The  different  tissues  (liver, 
cortex,  muscle,  tumor)  gave  different  elu- 
tion  patterns,  each  one  showing  at  least 
four  resolved  protein  peaks.  The  radioac- 

TABLE  9.     Amino  Acid  Composition  of  Tissue  Pools  and  of  Tissue  Protein 

Results  on  the  amino  acid  pools  are  expressed  in  terms  of  the  mean  and  its  standard  error.  Num- 
ber of  samples  is  in  parentheses.  Since  glutamine  is  lost  on  the  column  the  observed  values  are 
falsely  low.  The  value  for  glutamic  acid  in  protein  includes  glutamine  converted  to  glutamic  acid 
by  acid  hydrolysis.    Ph.  al.  +  =  phenylalanine  +  leucine  +  isoleucine. 

/xg  in  pool/ 

100  mg  tissue 

mg/100  mg  protein 

Amino  Acid 











Newborn   Adult 






Ph.  al.  + 


22    +3.5(5) 
28    +3.6(5) 
13    +1.1(6) 
13    +1.2(6) 
>10    +1.7(5) 

26    +1.9(6) 


20    +1.7(6) 

>19    +4.0(4) 

20    +1.6(6) 
47    +4.2(6) 
10    +1.1(6) 
>13    +2.4(4) 

31    +3.0(7) 
110    +5.8(7) 
>22    +4.8(6) 

4  (1) 


21  (1) 

11  (1)       11  (1) 
16(1)       17(1) 
3(1)        3(1) 
4(1)        3(1) 
5(1)        5(1) 
14  (1)       17  (1) 

40  60  80     90    150  20  40 


80     90    :50 

Fig.  40.  Radioactivity  of  glutamic  acid  and  glutamine  in  blood  plasma  and  in  amino  acid  pools 
of  cortex  and  liver,  and  of  glutamic  acid  in  proteins  of  cortex  and  liver,  after  subcutaneous  injec- 
tion of  L-glutamic  acid  randomly  labeled  with  C14. 

were  harvested,  washed,  and  hydrolyzed, 
and  hydrolysate  was  injected  into  two 
mice.  One  mouse  was  killed  after  an  hour, 
the  other  after  2  hours. 

Various  tissues  were  dissected  out  and 
homogenized,  and  the  cells  were  broken 
with  the  pressure  cell.  Samples  were  then 

tivity  per  unit  protein  varied  considerably 
from  one  peak  to  another,  indicating  either 
a  difference  in  the  sulfur  content  or  a  dif- 
ference in  the  rate  of  incorporation  in  the 
different  proteins.  Even  this  single  ex- 
ploratory experiment  showed  unequivo- 
cally that  there  is  no  single  protein  con- 


stituent  of  the  brain  cortex  which  accounts 
for  most  of  the  incorporation. 

In  the  same  experiment  8-mu  slices  were 
made  of  the  cortex.  Radioautographs  of 
the  slides,  compared  with  the  same  section 
stained  with  thionine  seen  visually,  show 
that  it  is  largely  the  cell  bodies  of  the  brain 
and  not  the  nerve  processes  that  incor- 
porate S35  into  protein  (fig.  41,  pi.  3). 
From  the  intensity  obtained  it  seems  possi- 
ble that  useful  radioautographs  could  be 
made  after  exposures  to  the  tracer  of  5 
minutes  or  less.  In  such  short  times  it 
might  be  possible  to  correlate  the  areas  of 
the  brain  that  synthesize  proteins  with 
different  functional  states  of  the  animal. 


In  Year  Book  55,  some  advantages  of 
using  relatively  simple  multicellular  ani- 
mals for  investigations  of  growth,  differ- 
entiation, and  reproduction  were  discussed 
and  preliminary  studies  with  radioactive 
substrates  were  reported.  Further  investi- 
gations on  Hydra  and  some  preliminary 
studies  on  Planaria  have  been  carried  out 
during  the  past  year. 

Feeding  and  Digestion 

Because  of  the  relative  impermeability 
of  Hydra  to  free  substrates,  it  was  found 
more  efficient  to  feed  radioactive  tissues 
to  the  hydras.  The  animals  were  fed  small 
bits  of  S35-labeled  mouse  lung  from  the 
carcass  of  a  mouse  used  in  the  experiments 
carried  out  in  collaboration  with  Dr.  L. 
Flexner,  as  described  elsewhere  in  this 
report.  Each  piece  of  sulfur-labeled  tissue 
was  soaked  in  10"4  M  glutathione  (GSH) 
for  5  minutes  and,  with  the  aid  of  watch- 
maker's forceps,  placed  individually  in  the 
center  of  the  ringlet  of  tentacles  near  the 
hypostome.  As  demonstrated  by  W.  F. 
Loomis,  a  hydra  will  swallow  nearly  any 
solid  object  of  acceptable  dimensions  in 
the  presence  of  GSH.  With  very  little 
practice  the  investigator  was  able  to  feed 
about  50  hydras  in  a  half  hour. 

Usually  10  to  12  S35-labeled  hydras  were 

sufficient  for  chemical  fractionation.  The 
animals  were  washed,  placed  in  0.2  ml 
distilled  water,  and  disrupted  by  sonica- 
tion.  The  TCA-soluble  fraction  was  fil- 
tered; the  activity  of  the  remaining  parti- 
cles was  considered  in  the  protein  calcula- 
tions. In  addition,  the  TCA-precipitable 
alcohol-soluble  fraction  was  separated  from 
the  TCA-precipitable  alcohol-insoluble 
fraction  by  filtration;  the  latter  fraction 
was  counted  directly  on  the  dry  filter. 
Experiments  performed  during  the  first 
6  hours  after  ingestion  required  measure- 
ments of  the  radioactivity  of  the  food 
within  the  hydra's  tissues  and  of  the  un- 
egested  food  within  the  gastrovascular 
cavity.  The  hydras  were  bisected  longi- 
tudinally with  a  scalpel  and  washed  free 
of  all  visible  material  as  discerned  using 
the  dissecting  microscope.  The  hydra 
tissue,  free  of  cavity  contents,  was  then 
fractionated;  the  washings  were  saved  and 

The  hydra  has  a  large  gastrovascular 
cavity  in  which  its  solid  food  is  broken 
into  pieces  small  enough  to  be  engulfed  by 
endodermal  cells.  Upon  engulfment  di- 
gestion continues  intracellularly.  This 
was  demonstrated  by  cutting  the  animal 
open,  washing  it  free  of  all  visible  food, 
and  chemically  fractionating  it.  The  food 
not  engulfed  was  also  fractionated.  The 
relative  proportions  of  the  TCA-soluble 
and  -insoluble  fractions  of  the  hydra  were 
similar  to  those  of  the  ingested  food  (cf. 
fig.  42).  Examination  of  the  alcohol-solu- 
ble (AS)  and  alcohol-insoluble  (AI)  frac- 
tions of  the  hydra,  however,  revealed  that 
a  change  in  the  proportions  of  these  radio- 
active components  occurred  between  1  and 
2  hours  after  the  food  was  swallowed. 
Since  the  food  protein  within  the  cells  at 
1  hour  is  of  similar  composition  to  that 
of  the  food  before  being  given  to  the  ani- 
mal, and  changes  only  after  this  time,  it 
appears  that  the  hydra  digests  its  food 
protein  intracellularly. 

This  was  further  demonstrated  by  ana- 
lyzing the  food  in  the  gut  that  had  not 
been  engulfed  3  hours  after  ingestion;  this 


food  had  the  same  proportions  of  AS  and 
AI  proteins  as  the  original  food,  although, 
as  shown  in  figure  42,  the  food  within  the 
cells  had  been  almost  completely  inter- 
converted  at  this  time.  Thus,  the  food  is 
ingested  into  the  gastrovascular  cavity, 
where  it  is  partially  degraded  into  small 
particles.  The  particles  are  engulfed  by 
the  gastrodermal  cells  and  are  subsequently 
digested  and  re-formed  into  hydra  protein. 
Initially,  the  engulfed  food  is  slowly  di- 

TABLE  10 

Radioactivity  (cps) 









per  cent 























>    80 


ir  eo 






°-    20 





a        Q 

Q                                               " 

a            o 


Q       TCA   PPT. 


















°         o 

0     __1— — ^— — — 


°     TCA    SOL. 
1                       l 


1               1 




Fig.  42.    Per  cent  distribution  of  S35-amino  acids  in  hydras'  food  (mouse  lung)  and  also  hydras' 
tissues  which  were  analyzed  at  different  periods  of  starvation  after  feeding  the  S35-mouse  lung. 

gested,  but  after  a  lag  it  is  rapidly  digested. 
This  finding  suggests  that  the  intracellular 
mechanisms  for  digestion  respond  adap- 
tively  to  engulfed  food.  By  measuring  the 
total  radioactivity  within  the  hydra  cells, 
and  the  total  radioactivity  of  either  the 
egested  food  or  the  contents  of  the  gastro- 
vascular cavity,  the  rate  of  protein  engulf- 
ment  by  the  hydra  was  determined.  The 
results  (table  10)  indicate  that  only  about 
10  per  cent  of  the  ingested  food  is  engulfed 
at  1  hour.  By  5  hours,  however,  the  en- 
gulfment  of  the  particles  was  essentially 
completed  and  the  percentage  of  radioac- 
tivity   retained   within   the   cells   did   not 

differ  significantly  from  that  of  the  ani- 
mals that  had  egested  their  wastes  and 
were  counted  9  hours  after  ingestion.  The 
data  presented  in  table  10  also  indicate 
that  protein  digestion  in  Hydra  is  rela- 
tively efficient  because  the  animals  retain 
80  to  90  per  cent  of  the  engulfed  radioac- 
tive protein. 

Whole  animal  radioautography  was  also 
used  to  study  the  digestion  and  redistri- 
bution of  the  sulfur-labeled  food  protein. 
Figure  43  (see  plate  4  for  figs.  43-46)  is  a 
radioautograph  of  an  animal  1  hour  after 
feeding.  The  partly  degraded  food  tissue 
is  seen  in  clumps  within  the  gastrovascular 


cavity;  no  radioactivity  is  seen  in  the  tenta- 
cles. After  3  hours  (fig.  44)  the  radioac- 
tivity is  fairly  equally  distributed  through- 
out the  animal,  except  for  the  tentacles 
and  the  lower  region  of  the  tube.  The 
absence  of  any  radioactivity  in  the  tentacles 
until  6  hours  after  ingestion,  when  traces 
were  first  seen,  suggests  that  the  endoderm 
of  the  tentacles  does  not  function  in  en- 
gulfing solid  particles  of  food.  At  6  hours 
after  ingestion  the  hydra  had  regurgitated 
the  undigested  food.  A  radioautograph  at 
this  time  (fig.  45)  revealed  that  most  of 
the  S35  was  well  distributed  except  for  the 
lower  part  of  the  body  tube,  which,  like  the 
tentacles,  does  not  seem  involved  in  the 
digestion  process.  There  does  seem  to  be 
some  incorporation  in  the  basal  disc,  a 
site  thought  to  be  relatively  active  in  syn- 
thesizing adhesive  substances.  There  are 
also  two  nonradioactive  "pockets"  in  the 
upper  side  of  the  body  tube,  correspond- 
ing to  two  testes  which  did  not  become 
labeled  during  6  hours.  This  observation 
suggests  that  after  the  testes  are  formed 
these  rapidly  differentiating  structures  are 
partly  independent  of  the  rest  of  the  ani- 
mal for  their  nutrition. 

S35-labeled  protein  was  shown,  by  means 
of  the  following  experiment,  to  reach  the 
extremities  of  the  tentacles.  The  upper 
third  of  a  radioactive  hydra  was  removed, 
and  in  its  place  was  grafted  a  nonradioac- 
tive hypostome  of  tentacles.  A  radioauto- 
graph of  this  "hybrid"  (Rg.  46),  taken 
after  2  days,  revealed  an  equal  distribution 
of  radioactivity  throughout  the  grafted 
tentacles.  It  is  probable  that  the  mecha- 
nism of  depositing  much  of  the  S35-labeled 
material  in  the  tentacles  involves  cell  mi- 
gration; cell  migration  is  now  definitely 
known  to  occur  in  the  specialized  cnido- 
blast  cells  whose  migration  is  described 

The  paths  of  protein  breakdown  and 
synthesis  in  hydras  starved  for  1  day  and 
then  fed  S35-labeled  mouse  tissue  are 
shown  in  figure  42.  No  significant  differ- 
ence is  observed  by  comparing  the  TCA- 
soluble  and  TCA-precipitable  fractions  of 

the  ingested  food  to  the  same  fractions  of 
the  hydra  starved  for  as  long  as  5  days 
after  feeding.  But  a  rapid  rise  in  the 
amount  of  an  alcohol-soluble  (AS)  protein 
occurs  within  6  hours;  this  protein  remains 
at  a  high  level  for  ll/2  days.  During  the 
next  12  hours  the  amount  of  the  AS  pro- 
tein falls  to  about  30  per  cent  of  the  total 
radioactivity,  and  it  remains  constant  at 
this  level  for  at  least  5  days.  At  the  time 
these  changes  in  the  amount  of  AS  protein 
were  occurring,  the  alcohol-insoluble  (AI) 
protein  followed  the  reciprocal  kinetics  of 
degradation  and  synthesis.  Analysis  of  the 
AS  fraction  demonstrated  that  it  did  not 
migrate  on  a  chromatogram  suspended  in 
a  butanol-formic  acid  solvent;  thus,  this 
fraction  appears  to  be  nonlipide  and  is 
probably  entirely  protein.  A  chromato- 
gram of  a  HC1  hydrolysate  of  the  AS  pro- 
tein revealed  the  presence  of  all  the  usual 
Hydra  amino  acids  including  hydroxypro- 
line  (see  below).  The  S35  was  distributed 
between  cysteine  and  methionine,  the 
former  being  more  radioactive;  this  distri- 
bution of  radioactivity  is  similar  to  that 
of  the  mouse  AS  and  AI  proteins.  On 
extraction  with  ether,  some  of  the  AS 
protein  precipitated. 

The  results  of  this  kinetic  experiment 
indicate  that  once  intracellular  digestion 
starts  the  AI  food  protein  is  rapidly  de- 
graded and  re-formed  into  Hydra  AS  pro- 
tein. It  is  not  yet  certain,  however,  whether 
the  degradation  of  the  ingested  protein 
proceeded  to  the  amino  acid  stage  before 
the  AS  protein  was  built  up,  or  whether 
the  Hydra  AS  protein  is  merely  "chopped- 
up"  food  in  the  form  of  low-molecular- 
weight  polypeptides  or  proteins.  During 
the  starvation  period  after  feeding,  the  AS 
protein  supplies  amino  acids  or  polypep- 
tides to  form  Hydra  AI  protein.  This 
"protein-to-protein"  transfer  may  well  re- 
flect one  of  the  mechanisms  that  help 
Hydra  keep  its  body  proportions  relatively 
constant  even  over  long  periods  of  starva- 

The  presence  of  hydroxyproline  in  the 
Hydra  AS  fraction  is  interesting  because 


it  has  recently  been  shown  that  the  cap- 
sule of  the  Hydra  nematocyst  is  an  un- 
usual member  of  the  collagen  family  of 
proteins  and  contains  more  than  20  per 
cent  hydroxyproline.  It  appears  that  the 
AS  fraction  may  contain  a  precursor  to 
the  capsule. 

The  rate  of  excretion  of  S35  compounds 
during  starvation  was  measured  by  sam- 
pling the  radioactivity  at  12-hour  intervals 
in  the  water  around  hydras  that  had  pre- 


S35     IN    HYDRA                                       ^-  -^.0 



S35     IN    MEDIUM 

4  00 


*l                  1                  1                   1                   1 


Fig.  47.  Loss  of  S35-label  by  a  hydra  on  star- 
vation, and  increase  of  S35-label  in  the  environ- 
ment over  the  same  period. 

viously  regurgitated  and  had  been  washed. 
As  demonstrated  in  figure  47,  the  hydras 
lose  about  25  per  cent  of  their  total  radio- 
active material  at  a  constant  rate  during 
5  days'  starvation.  If  the  S35  radioactivity 
can  be  interpreted  as  representative  of  all 
Hydra  protein,  these  results  suggest  that 
for  at  least  5  days  the  animals  are  either 
burning  up  their  body  protein  or  leaking 
free  amino  acids  to  the  outside. 

The  efficiency  with  which  hydras  utilized 
labeled  tissue  prompted  a  trial  with  Plana- 
via,  a  flatworm.  Planaria,  like  Hydra,  is  car- 
nivorous, but  is  much  more  highly  organ- 
ized; it  is  triploblastic,  has  a  more  compli- 

cated life  cycle,  and  has  well  developed 
organ  systems.  Upon  feeding  planarias  S35- 
labeled  mouse  lung  it  was  found  that  a 
relatively  large  amount  of  S35  was  incor- 
porated. Figures  48  and  49,  plate  5,  are 
radioautographs  of  sections  of  planarias 
killed  18  and  36  hours,  respectively,  after 
feeding.  This  simple  experiment  demon- 
strates the  possibility  of  using  tracer  tech- 
niques for  studies  on  lower  forms  and  in 
triploblastic  organisms  analogous  to  those 
already  made  with  Hydra.  Especially  in- 
triguing problems  are  the  conservation  of 
protein  throughout  a  complicated  life 
cycle  and  through  periods  of  stress,  and 
the  role  of  pre-existing  protein  during  re- 
generation of  an  amputated  part.  In  an- 
other aspect  to  the  work  on  Planaria,  egg 
cases  were  hydrolyzed  and  chromato- 
graphed.  The  chromatograms  revealed  the 
presence  of  a  ninhydrin-positive  spot  at 
the  position  corresponding  to  a-amino- 
adipic  acid.  This  amino  acid  has  not  been 
reported  previously  from  animal  protein 

Asexual  Reproduction 

In  the  earlier  studies  from  this  labora- 
tory it  was  found  that  25  to  30  per  cent  of 
parental  P32  was  contributed  to  the  off- 
spring of  each  asexual  generation.  Table 
11  presents  the  results  of  a  similar  study 
using  S35.  In  the  present  experiments,  the 
first  bud  contained  about  45  per  cent  of 
the  parental  S35  and  subsequent  genera- 
tions received  15  to  25  per  cent.  Figure  50, 
plate  5,  shows  that  the  first  bud  immedi- 
ately receives  and  incorporates  some  of  the 
ingested  S35-labeled  food  protein.  The 
fourth  and  fifth  buds,  although  of  similar 
size  to  their  earlier  sibs,  contain  a  smaller 
proportion  of  the  parental  S35  since  the 
parent  has  attached  buds  and,  therefore, 
more  tissue.  Figure  50  demonstrates  that 
most  of  the  bud  from  a  S35-labeled  parent 
is  evenly  labeled,  although  the  lower  por- 
tion of  the  parent  is  scantily  labeled.  The 
absence  of  appreciable  radioactivity  in  the 
gastrovascular  cavity   (about  0.1  per  cent 

Plate  2 

Department  of  Terrestrial  Magnetism 

No  glucose 






Fig.  23.  Sedimentation  patterns  of  ribonucleo- 
proteins  of  E.  coli.  Each  pair  obtained  in  a  sin- 
gle run  by  using  a  pair  of  analytical  cells  in  the 
ultracentrifuge.  Upper:  The  effect  of  lack  of 
glucose;  note  the  decrease  in  the  amount  of  the 
smaller  components.  Center:  The  effect  of 
broth;  note  the  absence  of  the  largest  component 
(^-80  S)  and  the  increase  in  amount  of  the 
smaller  ones.  Lower:  The  effect  of  chloram- 
phenicol; note  the  virtual  absence  of  the  largest 

Department  of  Terrestrial  Magneti 



-3    bO 
I     ^    °- 



Department  of  Terrestrial  Magnetism 


I  ■  ■ 
43     !        . 




Fig.  43.  Radioautograph  of  a  hydra  having  S35-labeled  mouse  lung  in  its  gastrovascular  cavity  for 
1  hour. 

Fig.  44.  Radioautograph  of  a  hydra  having  S35-labeled  mouse  lung  in  its  gastrovascular  cavity  for 
3  hours. 

Fig.  45.  Radioautograph  of  a  hydra  after  it  had  regurgitated  its  undigested  wastes  6  hours  after 
ingestion  of  S35-labeled  mouse  lung. 

Fig.  46.  Radioautograph  of  "hybrid"  hydra  consisting  of  S35-labeled  body  tube  grafted  to  an  un- 
labeled hypostome  and  tentacles.    The  radioautograph  was  made  2  days  after  the  grafting  operation. 

Plate  5 

Department  of  Terrestrial  Magnetism 





Fig.  48.  Raclioautograph  of  a  longitudinal  section  of  a  "coiled"  planaria  that  had  been  fed  S35- 
labeled  mouse  lung  18  hours  before  it  was  killed. 

Fig.  49.  Radioautograph  of  a  cross  section  of  a  planaria  that  had  been  fed  S35-labeled  mouse  lung 
36  hours  before  it  was  killed. 

Fig.  50.  Radioautograph  of  a  bud  (left)  attached  to  its  parent  (right).  The  parent  was  fed 
S35-labeled  mouse  lung,  and  subsequently  fed  unlabeled  brine  shrimp  until  the  bud  completed  its 

Fig.  51.     Radioautograph  of  a  hydra  exposed  to  C1402  for  16  hours. 

Plate  6 

Department  of  Terrestrial  Magnetism 

<u  ~      — 

*->  '£  m 


o   <u 
-Q  -73 

O    aj 

as    £    O     as    o 

O  T3 

co   JJ2   JJ   CJ 

O    as    a      . 

■^^       S3 

—i  j3  a  o 
o   2  .2  vo 

°  3  &■ 


^  o-o 

v>    O 

Oh    >-       . 

<X     X   72   --< 

<u    >    aj 

Mr*  03     H 

<U  .5 

J3    C 


•     as 
£     Pi 


£  ~  p* 

2  >,S 


a,  cr 


-5  ~o 


in    5 



2  § 

55    <u 

Q-l     Q 

X  "as 

o  <u 

3  2 

OS  rj 

O  aj 

bC  in 

O  ^ 

.3  -a 

^  T3. 

O  G 

^  O 

o  cr 

U  o 





of  the  total)  precludes  the  possibility  of 
passage  of  any  significant  protein  via  this 
cavity.  It  appears,  therefore,  that  buds  are 
derived  from  the  more  heavily  labeled  re- 
gion in  the  upper  part  of  the  parent  by 
processes  that  involve  cellular  migration 
as  a  chief  element  in  the  transport  of 
protein  from  parent  to  offspring. 

TABLE   11.     Transfer  of  S35  to  Buds 

Per  Cent  * 

of  Parental  S3" 

BnH  No 































*  This  proportion  was  based  on  the  amount  of 
S35  in  the  parent  just  before  the  bud  detached. 

Migration  of  C14-Labeled  Cnidoblasts 

The  migration  of  the  cnidoblast,  which 
contains  the  developing  nematocyst,  is  in- 
volved in  the  formation  of  the  tentacles. 
The  results  below  demonstrate  that  starved 
hydras  can  specifically  incorporate  C14 
from  labeled  carbon  dioxide  into  the  cnido- 
blast. It  has  been  possible  by  radioautog- 
raphy  to  demonstrate  the  migration  of  the 
cnidoblast  from  its  site  of  origin  in  the 
body  tube  to  its  site  of  function.  For  these 
studies  hydras  in  a  closed  vessel  were  ex- 
posed to  C1402  for  various  periods  of  time. 
When  the  required  exposure  had  been 
completed,  the  organisms  were  washed, 
laid  out  on  a  membrane  filter,  killed  by 
heat-drying,  and  radioautographed  with 
Eastman  NTB3  nuclear  track  plates.  Use- 
ful radioautographs  generally  required  2 
to  4  weeks'  exposure. 

Radioautographs  of  animals  exposed  to 
C1402  for  16  hours  (fig.  51,  pi.  5)  revealed 
that  most  of  the  radioactivity  was  confined 
to  small  but  discrete  "loci"  throughout  the 

upper  two-thirds  of  the  body  tube.  Few,  if 
any,  of  these  loci  were  present  in  the 
tentacles.  If,  on  the  other  hand,  hydras 
were  allowed  to  starve  in  a  nonradioactive 
environment  for  48  hours  after  being  ex- 
posed to  labeled  carbon  dioxide,  many  of 
the  loci  would  appear  in  the  tentacles  (fig. 
52,  pi.  6).  This  migration  of  the  labeled 
loci  from  the  body  tube  to  the  tentacles 
suggests  that  the  cnidoblasts,  which  are 
making  nematocysts  while  in  the  body 
tube,  migrate  to  the  tentacles  where  the 
nematocysts  are  to  be  used.  A  radioauto- 
graph  of  the  tentacles  and  hypostome  of  a 
similarly  treated  animal  demonstrates  the 
dense  concentration  of  radioactive  nema- 
tocysts in  the  tentacles,  and  the  presence 
of  probable  migratory  cnidoblasts  in  the 
hypostome  region  while  in  the  midst  of 
their  journey  (fig.  53,  pi.  6). 

Experiments  have  been  carried  out  in 
which  the  hydras  were  exposed  to  C1402 
for  16  hours  during  a  period  when  a  bud 
was  just  beginning  on  each  of  the  animals. 
The  bud  was  allowed  to  complete  its  de- 
velopment in  a  nonradioactive  environ- 
ment, and  a  radioautograph  was  made  of 
that  bud  after  it  detached.  As  shown  in 
figure  54,  plate  6,  the  batteries  of  the  bud's 
tentacles  are  strongly  labeled,  as  are  also 
the  cnidoblasts  in  the  body.  Since  the  bud 
developed  from  the  prelabeled  parent,  it 
apparently  did  not  make  all  its  own  nema- 
tocysts but  was  given  a  large  share  of 
nematocysts  preformed  by  its  parent. 

In  experiments  in  which  a  nonradioac- 
tive hypostome  with  a  ringlet  of  tentacles 
was  grafted  onto  a  C14-labeled  body  tube 
of  another  animal,  the  radioactive  cnido- 
blasts migrated  from  the  body  tube  and 
were  deposited  along  the  tentacles.  Thus, 
even  tissues  from  another  individual  can 
be  invaded  by  cells  from  the  body  tube. 

Chemical  analysis  of  C14-labeled  hydras 
has  shown  that  the  C14  is  distributed 
among  TCA-soluble,  alcohol-soluble,  and 
alcohol-insoluble  fractions.  Most  of  the 
radioactivity  is  found  in  glutamic  and  as- 
partic  acids.  This  finding  hints  that  the 
Krebs  cycle  operates  in  Hydra  as  it  does 


in  other  forms.  In  this  connection  it  is  of 
interest  that  the  C14  label  appears  in  those 
cells  that  are  most  active  in  synthesizing 
protein:  cnidoblasts,  which  serve  to  re- 
arm the  batteries  of  nematocysts;  and 
testes  (fig.  55,  pi.  6)  and  ovaries  (fig.  56, 
pi.  6),  whenever  these  organs  are  being 

differentiated.  Thus,  not  only  does  a  Krebs 
cycle  appear  to  be  functioning  but  it  may 
serve,  as  in  E.  coli,  to  provide  materials 
for  protein  synthesis.  These  results  may 
have  some  bearing  on  the  C02  induction 
of  sexual  differentiation  as  demonstrated 
by  Loomis. 



Cooperation  with  various  institutions 
and  organizations  in  this  country  and 
abroad  has  been  continued  during  the  year. 
These  include  the  American  Geophysical 
Union,  Applied  Physics  Laboratory,  Asso- 
ciated Universities,  Inc.,  Bernard  Price  In- 
stitute of  Geophysics  (S.  Africa),  U.  S. 
Coast  and  Geodetic  Survey,  Common- 
wealth Scientific  and  Industrial  Research 
Organization  (Australia),  Department  of 
Mines  and  Technical  Surveys  (Canada), 
Evans  Signal  Laboratory,  Geological  Sur- 
vey of  Canada,  U.  S.  Geological  Survey, 
Geophysical  Institute  of  Huancayo  (Peru), 
Institut  Pasteur  (France),  International 
Scientific  Radio  Union,  International 
Union  of  Geodesy  and  Geophysics,  Johns 
Hopkins  University,  Lamont  Geological 
Observatory,  National  Bureau  of  Stand- 
ards, National  Institutes  of  Health,  Na- 
tional Radio  Astronomy  Observatory,  Na- 
tional Research  Council,  National  Science 
Foundation,  Oak  Ridge  National  Labora- 
tory, Rocky  Mountain  Laboratory  of  the 
U.  S.  Public  Health  Service,  Harvard, 
Stanford,  and  Yale  Universities,  and  the 
Universities  of  Minnesota,  Wisconsin,  and 
Western  Australia. 

We  have  continued  to  work  closely  with 
the  Geophysical  Laboratory  on  the  de- 
termination of  mineral  ages  by  means  of 
isotope  measurements. 

Our  cosmic-ray  investigations  have  been 
aided  by  the  cooperation  of  observatories 
at  Christchurch,  New  Zealand;  Climax, 
Colorado;  Fredericksburg,  Virginia;  God- 
havn,  Greenland;  Huancayo,  Peru;  and 
Mexico,  D.  F.;  and  by  the  continuation 

of  our  contract  with  the  Office  of  Naval 
Research  covering  the  loan  of  government 

Licenses  from  the  Atomic  Energy  Com- 
mission remain  in  effect  for  the  procure- 
ment of  special  nuclear  material,  and  for 
by-product  materials  used  in  biophysical 

A  second  contract  with  the  Office  of 
Naval  Research,  for  investigations  of  the 
earth's  crust,  gave  invaluable  assistance 
to  our  Carnegie  Andes  Expedition  through 
the  loan  of  equipment. 

The  National  Science  Foundation  has 
made  two  additional  grants  to  the  Institu- 
tion, a  total  of  seven  such  grants  having 
been  administered  during  the  report  year. 
One  new  grant  provides  funds  for  investi- 
gations and  construction  of  photoelectric 
image  tubes  for  research  in  astronomy; 
the  other  supports  reduction  of  cosmic-ray 
ionization-chamber  data  in  the  Interna- 
tional Geophysical  Year  Cosmic  Ray  Pro- 

Eight  members  of  our  staff  participated 
in  the  seismic  expedition  to  Peru,  Bolivia, 
and  Chile  in  the  fall  of  1957.  Three  of 
the  members  extended  their  travels  to  con- 
fer with  present  and  potential  colleagues 
in  Argentina,  Brazil,  Ecuador,  and  Vene- 
zuela. Another  stafT  member  spent  nearly 
two  months  in  South  America  in  connec- 
tion with  the  International  Geophysical 
Year  project  concerned  with  the  height  of 
the  equatorial  electrojet  in  the  ionosphere. 
Several  staff  members  have  continued  to 
serve  on  committees  and  panels  for  the 
Biophysical  Society,  Federal  Civil  Defense 
Administration,  International  Geophysical 



Year,  International  Scientific  Radio  Union, 
and  the  National  Academy  of  Sciences. 


Publication  of  the  Journal  of  Geophysi- 
cal Research  has  continued  during  the 
report  year,  with  partial  subsidy  from  the 

Institution.  Preliminary  plans  have  been 
made  for  the  American  Geophysical  Union 
to  take  over  publication  of  the  Journal  in 

Rental  of  more  than  100  acres  for  radio 
astronomy  activities  has  been  continued  at 
the  River  Road  site. 


Abelson,  P.  H.   See  Major  Publications  below. 

Anderson,  C.  E.    See  Heydenburg,  N.  P. 

Balsley,  J.  R.     See  Graham,  J.  W. 

Barloutaud,  R.,  T.  Grjebine,  P.  Lehmann,  A. 
Leveque,  J.  Quidort,  and  G.  M.  Temmer. 
Recherche  de  l'excitation  coulombienne  du 
deuxieme  niveau  de  194Pt.  Compt.  rend, 
acad.  sci.  Paris,  244,  187-190  (1957). 

Baum,  W.  A.,  W.  K.  Ford,  Jr.,  and  J.  S.  Hall. 
Recent  astronomical  tests  of  thin-film  image 
converters.  (Abstract.)  Astron.  J.,  63, 
47-48  (1958). 

Bolton,  E.  T.,  and  J.  Leahy.  Association  of 
metabolic  pool  material  with  structural 
components  of  Escherichia  coli.  (Abstract.) 
Natl.  Biophys.  Conf.,  Columbus,  Ohio,  pp. 
11-12  (1957). 

Bolton,  E.  T.,  and  H.  G.  Mandel.  The  effects 
of  6-mercaptopurine  on  biosynthesis  in 
Escherichia  coli.  J.  Biol.  Chem.,  227,  833- 
844  (1957). 

Bolton,  E.  T.  See  also  Cowie,  D.  B.;  Hoyer, 
B.  H.;  and  Major  Publications  below. 

Britten,  R.  J.,  and  F.  T.  McClure.  Osmotic  ef- 
fects in  Escherichia  coli.  (Abstract.)  Natl. 
Biophys.  Conf.,  Columbus,  Ohio,  p.  14 

Britten,  R.  J.  See  also  Cowie,  D.  B.,  and  Major 
Publications   below. 

Buddington,  A.  F.    See  Graham,  J.  W. 

Burke,  B.  F.  Confusion  effects  in  surveys  of 
radio  sources.  (Abstract.)  URSI  Program 
of  Joint  Meeting,  National  Bureau  of  Stand- 
ards, Apr.  23-26,  1958,  p.  16. 

Burke,  B.  F.,  and  J.  W.  Firor.  Limiting  ac- 
curacy of  open-wire  transmission  lines  in 
positional  radio  astronomy.  (Abstract.) 
URSI  Program  of  Joint  Meeting,  National 
Bureau  of  Standards,  Apr.  23-26,  1958,  p.  16. 

Burke,  B.  F.,  and  K.  L.  Franklin.  Jupiter  as  a 
radio  source.  I.A.U.  Symposium  4,  Radio 
Astronomy  (Manchester  meeting,  1955), 
edited  by  H.  C.  van  de  Hulst,  Cambridge 
University  Press,  pp.  394-396  (1957). 

Burke,  B.  F.,  and  K.  L.  Franklin.  Observations 
of  discrete  sources  with  the  22  Mc/s  Mills 
Cross.  I.A.U.  Symposium  4,  Radio  Astron- 
omy (Manchester  meeting,  1955),  edited  by 

H.  C.  van  de  Hulst,  Cambridge  University 
Press,  pp.  151-153  (1957). 

Cohen,  G.  N.    See  Cowie,  D.  B. 

Cowie,  D.  B.,  and  E.  T.  Bolton.  The  use  of 
metabolic  pools  of  purine  compounds  for 
nucleic  acid  synthesis  in  yeast.  Biochim.  et 
Biophys.  Acta,  25,  292-298  (1957). 

Cowie,  D.  B.,  and  R.  J.  Britten.  Kinetic  studies 
of  metabolic  pools  in  microorganisms. 
(Abstract.)  Natl.  Biophys.  Conf.,  Colum- 
bus, Ohio,  p.  22  (1957). 

Cowie,  D.  B.,  and  G.  N.  Cohen.  Biosynthesis 
by  Escherichia  coli  of  active  altered  proteins 
containing  selenium  instead  of  sulfur.  Bio- 
chim. et  Biophys.  Acta,  26,  252-261  (1957). 

Cowie,  D.  B.  See  also  Lederberg,  J.,  and  Major 
Publications  below. 

Davis,  G.  L.    See  Tilton,  G.  R. 

Erickson,  W.  C.  A  mechanism  of  non-thermal 
radio-noise  origin.  Astrophys.  J.,  126,  480- 
492  (1957). 

Firor,  J.  W.  A  radio  telescope.  QST,  41,  32- 
36  (1957). 

Firor,  J.  W.  Brightness  distribution  of  the  sun 
at  1.45  metres.  I.A.U.  Symposium  4,  Radio 
Astronomy  (Manchester  meeting,  1955), 
edited  by  H.  C.  van  de  Hulst,  Cambridge 
University  Press,  pp.  294-297   (1957). 

Firor,  J.  W.  The  slowly  varying  component  of 
solar  radiation  at  340  Mc.  (Abstract.) 
URSI  Program  of  Joint  Meeting,  National 
Bureau  of  Standards,  Apr.  23-26,  1958,  p.  15. 

Firor,  J.  W.    See  also  Burke,  B.  F. 

Forbush,  S.  E.  Large  increase  of  cosmic-ray 
intensity  following  solar  flare  on  February 
23,  1956.  (Letters  to  Editor.)  /.  Geophys. 
Research,  61,  169-170  (1957). 

Forbush,  S.  E.  See  also  Major  Publications  be- 

Ford,  W.  K.,  Jr.    See  Baum,  W.  A. 

Frankel,  S.,  P.  G.  Hansen,  O.  Nathan,  and 
G.  M.  Temmer.  Detection  of  Ga66  positron 
polarization  by  the  annihilation-in-flight  rate 
in  polarized  matter.  Phys.  Rev.,  108,  1099- 
1101  (1957). 

Franklin,  K.  L.    See  Burke,  B.  F. 

Graham,  J.  W.  The  role  of  magnetostriction  in 
rock  magnetism.  Phil.  Mag.  Supplement,  6, 
362  (1957). 


Graham,  }.  W.,  A.  F.  Buddington,  and  J.  R. 
Balsley.  Stress-induced  magnetizations  of 
some  rocks  with  analyzed  magnetic  min- 
erals. /.  Geophys.  Research,  62,  465-474 

Grjebine,  T.    See  Barloutaud,  R. 

Hall,  J.  S.     See  Baum,  W.  A. 

Hansen,  P.  G.     See  Frankel,  S. 

Heydenburg,  N.  P.,  and  G.  F.  Pieper.  Coulomb 
excitation  of  rotational  levels  in  dysprosium. 
Phys.  Rev.,  107,  1297-1299  (1957). 

Heydenburg,  N.  P.,  G.  F.  Pieper,  and  C.  E.  An- 
derson. Coulomb  excitation  of  krypton. 
Phys.  Rev.,  108,  106-107  (1957). 

Heydenburg,  N.  P.  See  also  Pieper,  G.  F.;  and 
Temmer,  G.  M. 

Hoyer,  B.  H.,  E.  T.  Bolton,  R.  A.  Ormsbee, 
G.  LeBouvier,  D.  B.  Ritter,  and  C.  L.  Lar- 
son. Mammalian  viruses  and  rickettsiae. 
Science,  127,  859-863  (1958). 

Lange,  I.    See  Major  Publications  below. 

Larson,  C.  L.     See  Hoyer,  B.  H. 

Leahy,  J.     See  Bolton,  E.  T. 

LeBouvier,  G.    See  Hoyer,  B.  H. 

Lederberg,  J.,  and  D.  B.  Cowie.  Moondust. 
Science,  127,  W3-W5  (1958). 

Lehmann,  P.     See  Barloutaud,  R. 

Leveque,  A.    See  Barloutaud,  R. 

Mandel,  H.  G.     See  Bolton,  E.  T. 

McClure,  F.  T.  See  Britten,  R.  J.;  and  Roberts, 
R.  B. 

Nathan,  O.     See  Frankel,  S. 

Ormsbee,  R.  A.    See  Hoyer,  B.  H. 

Pieper,  G.  F.,  and  N.  P.  Heydenburg.  Coulomb 
excitation  of  iron-57.  Phys.  Rev.,  107,  1300- 
1302  (1957). 

Pieper,  G.  F.    See  also  Heydenburg,  N.  P. 

Quidort,  J.     See  Barloutaud,  R. 

Ritter,  D.  B.    See  Hoyer,  B.  H. 

Roberts,  R.  B.,  and  F.  T.  McClure.  Models  of 
Escherichia  coli  relating  structure  and  func- 
tion. (Abstract.)  Natl.  Biophys.  Conf., 
Columbus,  Ohio,  p.  59  (1957). 

Roberts,  R.  B.  See  also  Major  Publications  be- 

Scott,  W.  E.  List  of  recent  publications.  /.  Geo- 
phys. Research,  62,  500-507,  650-658  (1957); 
63,257-264,429-436  (1958). 

Tatel,  H.  E.  21-cm  meridian  plane  surveys. 
I.A.U.  Symposium  4,  Radio  Astronomy 
(Manchester  meeting,  1955),  edited  by  H.  C. 
van  de  Hulst,  Cambridge  University  Press, 
pp.  67-70  (1957). 

Tatel,  H.  E.,  and  M.  A.  Tuve.  Carnegie  seis- 
mic expedition  to  the  Andes,  1957.  (Ab- 
stract.) Trans.  Am.  Geophys.  Union,  39, 
580  (1958). 

Tatel,  H.  E.,  and  M.  A.  Tuve.  Seismic  obser- 
vations at  one  kilometer  depth.     Contribu- 

tions in  Geophysics:  Honor  of  Beno  Guten- 
berg, Pergamon  Press,  London,  pp.  152- 
157  (1958). 

Temmer,  G.  M.  Science  in  the  USSR.  "Nu- 
clear Physics,"  Science,  126,  1097  (1957). 

Temmer,  G.  M.,  and  N.  P.  Heydenburg.  Low- 
lying  excited  states  of  Na22.  (Abstract.) 
Bull.  Am.  Phys.  Soc,  ser.  II,  3,  200  (1958). 

Temmer,  G.  M.  See  also  Barloutaud,  R.;  and 
Frankel,  S. 

Tilton,  G.  R.,  G.  L.  Davis,  and  G.  W.  Wetherill. 
The  age  of  the  Baltimore  gneiss.  (Abstract.) 
Program  Am.  Geophys.  Union,  May  5-8, 
1958,  p.  42. 

Tuve,  M.  A.    See  Tatel,  H.  E. 

Vestine,  E.  H.  Geomagnetic  field.  The  Earth 
and  Its  Atmosphere,  edited  by  D.  R.  Bates, 
Basic  Books,  Inc.,  New  York,  pp.  89-96 

Wells,  H.  W.  Flux  measurements  of  Cassiopeia 
A  and  Cygnus  A  between  18.5  Mc  and  107 
Mc.    Proc.  IRE,  46,  205-208  (1957). 

Wells,  H.  W.  International  cooperation  in  radio 
research.  Dept.  State  Bull.,  37,  897-899 

Wells,  H.  W.  Large-scale  movements  of  the 
layers.  /.  Atmospheric  Terrest.  Phys.  (Spe- 
cial Supplement),  Proc.  Polar  Atmos.  Sym- 
posium Part  II,  Ionosphere  Section,  Oslo, 
pp.  33-40  (1957). 

Wells,  H.  W.  Preliminary  observations  of 
point  source  at  12.5  and  15.5  Mc/s.  I.A.U. 
Symposium  4,  Radio  Astronomy  (Manches- 
ter meeting,  1955),  edited  by  H.  C.  van  de 
Hulst,  Cambridge  University  Press,  pp.  148— 
150  (1957). 

Wells,  H.  W.  Unusual  propagation  at  40  Mc 
from  the  USSR  satellite.  Proc.  IRE,  46,  610 

Wetherill,  G.  W.  Radioactivity  of  potassium  and 
geologic  time.    Science,  126,  545-549  (1957). 

Wetherill,  G.  W.    See  also  Tilton,  G.  R. 


Cosmic-Ray  Results.  Huancayo,  Peru,  January 
1946-December  1955;  Cheltenham,  Mary- 
land, March  1936-December  1955;  Christ- 
church,  New  Zealand,  January  1947-De- 
cember  1955;  Godhavn,  Greenland,  January 
1947-December  1950.  By  I.  Lange  and  S.  E. 
Forbush.  Carnegie  Inst.  Wash.  Publ.  175, 
XX.  Quarto,  v  +  224  pp.,  431  tables.  Au- 
gust 1957. 

Studies  of  Biosynthesis  in  Escherichia  coli.  By 
R.  B.  Roberts,  P.  H.  Abelson,  D.  B.  Cowie, 
E.  T.  Bolton,  and  R.  J.  Britten.  Carnegie 
Inst.  Wash.  Publ.  607..  Octavo,  xiv  +  521 
pp.,  114  figs.,  second  printing  with  brief 
addenda.     September  1957. 




M.  A.  Tuve 

Staff  Members 

L.  T.  Aldrich 

E.  T.  Bolton 

R.  J.  Britten 

B.  F.  Burke 

D.  B.  Cowie 

J.  W.  Firor 

S.  E.  Forbush 

W.  K.  Ford,  Jr.  (from  September  3,  1957) 

J.  W.  Graham  (resigned  February  8,  1958) 

N.  P.  Heydenburg 

R.  B.  Roberts 

H.  E.  Tatel  (deceased  November  15,  1957) 

G.  M.  Temmer 

H.  W.  Wells 

G.  W.  Wetherill 

Section  Chairmen 

Atmosphere:     H.  W.  Wells 

Biophysics.    R.  B.  Roberts 

Earth's  Crust:     H.  E.  Tatel  to  November  15, 

1957;  L.  T.  Aldrich  from  December    12, 


Nuclear  Physics:     N.  P.  Heydenburg 
Theoretical  Geophysics:     S.  E.  Forbush 

Fellows  and  Associates 

G.  N.  Cohen,  Institut  Pasteur,  Paris,  France 
(part  time) 

D.  W.  Dewhirst,  University  of  Cambridge, 
Cambridge,  England  (entire  report  year) 

W.  C.  Erickson  (to  August  1957) 

H.  L.  Heifer  (to  August  1957) 

J.  J.  Leahy  (to  August  1957) 

H.  M.  Lenhoff  (from  February  1,  1958) 

D.  H.  Lindsley,  Johns  Hopkins  University 
(part  time) 

J.  L.  Pawsey,  Australian  Commonwealth  Sci- 
entific and  Industrial  Research  Organiza- 

Vi si  ting  In 

Carl  E.  Anderson,  Yale  University  (Septem- 
ber 1957) 

J.  A.  Clegg,  United  Kingdom  (September 

R.  D.  Davies,  Jodrell  Bank  Experimental  Sta- 
tion, Cheshire,  England  (July  1957) 

L.  B.  and  Mrs.  J.  B.  Flexner,  University  of 
Pennsylvania  (part  time) 

S.  Gorodetsky,  Institute  of  Nuclear  Research, 
Strasbourg,  France  (March  1958) 

B.  H.  Hoyer,  Rocky  Mountain  Laboratory 
(April  1958) 

C.  Lomnitz,  University  of  Chile,  Santiago, 
Chile  (June  1958) 

H.  G.  Mandel,  George  Washington  Univer- 
sity (part  time) 

tion,  Sydney,  Australia  (from  August 
1957  to  May  1958) 

G.  F.  Pieper,  Yale  University  (to  August 

I.  Z.  Roberts,  Trinity  College  (part  time) 

H.  de  Robichon-Szulmajster,  National  Insti- 
tutes of  Health  (part  time) 

J.  O.  Thomas,  University  of  Cambridge, 
Cambridge,  England  (April  1958) 

H.  Weaver,  University  of  California  (to  Au- 
gust 1957) 


D.  R.  Marshall,  University  of  Chicago  (Oc- 
tober 1957) 

F.  T.  McClure,  Applied  Physics  Laboratory 
(part  time) 

Catharine    Mouquin,    Trinity   College    (part 

time  from  February  1958) 
L.  O.  Nicolaysen,  Bernard  Price  Institute  of 

Geophysics,  Johannesburg,  S.  Africa  (April, 

May  1958) 

E.  C.  Pollard,   Yale  University   (September 

P.  German  Saa,  S.J.,  Universidad  del  Norte, 
Antofagasta,  Chile  (January  1958) 

G.  J.    Wasserburg,    California    Institute    of 
Technology  (April  1958) 


Research  Assistants 

J.  B.  Doak 
E.  T.  Ecklund 

C.  A.  Little,  Jr. 
W.  E.  Scott 

Laboratory  Assistants 

Miss  R.  E.  Bresnahan  Miss  B.  D.  North 

S.  J.  Buynitzky  R.  W.  Rasmussen  (from  April  1,  1958) 

Miss  E.  F.  French  (to  August  23,  1957)  R.  W.  Reuschlein  (to  February  28,  1958) 
P.  A.  Johnson 


Chief,  Administrative  and  Operating  Section: 
M.  B.  Smith  (retired  December  31,  1957) 

Chief,  Fiscal  Section:  Miss  H.  E.  Russell 
(from  January  1,  1958);  Accountant  (to 
December  31,  1957) 

Director's    Secretary    and    Office    Manager: 

W.  N.  Dove 
Librarian:     Mrs.  L.  J.  Prothro 
Stenographers:     Mrs.   C.   Ator;   Mrs.   A.   L. 

Hill  (from  May  14,  1958);  Mrs.  C,  Wind- 

muller  (to  May  23,  1958) 


Chief  of  Subsection — Main  Shop  and  Main- 
tenance:    W.  F.  Steiner 

Senior  Instrument  Makers:  B.  J.  Haase; 
L.  A.  Horton  (retired  June  30,  1958); 
J.  G.  Lorz 

Machinist:     F.  J.  Caherty 
Machinist-Instrument   Maker:     M.   Seemann 
(from  June  16,  1958) 

Foreman:     C.  Balsam 
Caretaker:     E.  Quade 

Mrs.  L.  Beach 
R.  W.  Bluehdorn 
M.  S.  Halversen 
R.  C.  Kile 
F.  R.  Norcross 
N.  E.  Peppell 

Buildings  and  Grounds 

Assistant    Caretakers:     C.    R.    Forshier;    S. 
Gawrys;  S.  Swantkowski 

Part-Time  and  Temporary  Employees 

R.  L.  Peppell 
K.  W.  Schwarz 
Mrs.  M.  T.  Sheahan 
C.  F.  Stroebel 
A.  R.  Turpin 
W.  W.  Wyatt 


Washington,  District  of  Columbia  PHILIP  H.  ABELSON,  Director 



Introduction    169 

Experimentation  at  High  Pressure 170 

Development  of  high-pressure  appa- 
ratus     170 

Single-stage  apparatus 170 

Two-stage  apparatus   172 

Melting  of  diopside  under  high  pres- 
sure    173 

Melting  points  of  alkali  halides  at  high 

pressure    174 

The  Age  of  Roc\s  and  Minerals 176 

Appalachian  ages 177 

Other  ages    179 

North  American  ages 181 

Acknowledgments    181 

Mineral  Assemblages  in  the  Green  River 

Formation    181 

Geochemistry  of  Artificial  Isotopes 184 

The  beneficiation  of  soils  contaminated 
with  strontium  90;  beneficial  ef- 
fects of  potassium 184 

Rain  fallout 186 

Tritium  hydrology 189 

Experimental  Petrology   189 

Effect  of  water  on  the  melting  of  sili- 
cates      189 

Iron-rich  chlorites 191 

Biotites 192 

Phase  relations  of  hydrous  silicates 

with  intermediate  Mg/Fe  ratios.  .  193 

The  phlogopite-annite  join 194 

Ferrous-ferric  biotites    195 

Cordierite-H20  system 195 

Polymorphism     196 

Composition    196 

Lower  stability  limits 196 

Anhydrous  alkali-free  cordierites 197 

Alkali  amphiboles   199 

Magnesioriebeckite    199 

Glaucophane  202 

Riebeckite    204 

Composition  of  a  pseudoleucite  from 
the  Bearpaw  Mountains,  Mon- 
tana      204 

Feldspar  investigations   206 

Spinels    209 

Magnetite-hercynite  relations 209 

Chester  emery  deposits 210 

The  quaternary  system   Na20-MgO- 

Al203-Si02  210 

Systems    with    rock-forming    olivines, 

pyroxenes,  and  feldspars 212 

Paleobiochemistry    214 

Ore  Minerals 214 

The  Cu-S  system 215 

The  Fe-S  system 218 

The  solidus  in  the  system  Cu-Fe-S  be- 
tween 400°  and  800°  C 222 

The  Fe-Zn-S  system 227 

The  Fe-As-S  system 229 

The  CoAs2-NiAs2-FeAs2-As  system .  .  232 

Sulfide-water  systems   234 

Ore  solutions 234 

Diffraction  Effects  of  Short-Range  Order- 
ing in  Layered  Sequences 240 

The  distribution  of  run  lengths 240 

Generation  of  diffraction  masks  of  lay- 
ered sequences    241 

The  optical  diffractometer 242 

The  calculation  of  intensity  profiles .  .  .    242 

Crystallography    243 

Neutron  diffraction  studies 243 

Magnetic  structure  of  chalcopyrite .  .   243 
Symmetry  of  magnetic  structures .  .  .   244 

Sulfides    246 

High-temperature  chalcopyrite 246 

Arsenopyrite    , 246 

Bornite     248 

Silicates    249 

Pyroxenes    249 

Synthetic  mica  of  type  3M 252 

Miscellaneous  Administration   253 

Penologists'  Club 253 

Seminars    253 

Johns  Hopkins  University  and  Geo- 
physical Laboratory  graduate  sem- 
inars on  "Researches  in  geochem- 
istry"    254 

Symposium  on  C14  dating,  Pleistocene 
stratigraphy,      and      archaeologic 

chronology 255 

Lectures  255 

Summary  of  Published  Wor\ 256 

Bibliography   259 

Personnel 259 

Carnegie  Institution   of  Washington  Year  Boo\  57,  1957-1958 


Geochemical  and  geophysical  research 
are  currently  making  exciting  progress. 
Unfortunately,  publications  on  these  topics 
are  widely  scattered  and  are  often  effec- 
tively lost  to  most  graduate  students  and 
professional  earth  scientists.  The  Geophysi- 
cal Laboratory  has  one  successful  partial 
solution  to  this  problem.  Requests  for  our 
annual  reports,  which  contain  most  of  our 
results,  come  from  individuals  and  libraries 
all  over  the  world.  More  than  1500  copies 
are  sent  out  each  year,  over  a  third  of  them 
to  foreign  countries. 

During  the  past  year  this  Laboratory 
made  an  additional  effort  in  the  collection 
and  dissemination  of  scientific  information 
related  to  our  area  of  activity.  A  series  of 
weekly  seminars  and  discussions  on  the 
subject  "Researches  in  Geochemistry"  was 
held  both  at  the  Johns  Hopkins  University 
in  Baltimore  and  at  this  Laboratory  during 
the  academic  year  1957-1958.  Some  of  the 
leading  geochemists  of  the  country  partici- 
pated. The  manuscripts  they  submitted, 
covering  their  remarks,  have  been  edited 
at  the  Laboratory  and  will  be  published 
early  in  1959  by  John  Wiley  &  Sons,  Inc., 
New  York.  The  volume,  containing  23 
chapters,  should  be  helpful  in  many  ways 
to  students  and  to  more  mature  investi- 
gators. It  provides  reviews  of  progress  in 
many  interesting  areas  by  men  currently 
active  in  research.  It  describes  much  new 
and  previously  unpublished  work,  and  in- 
cludes many  excellent  bibliographies.  A  list 
of  speakers  and  titles  is  given  in  a  later 
section  of  the  present  report. 

Our  diversified  experimental  program 
made  substantial  progress  in  many  areas. 
Boyd,  S.  Clark,  and  England  have  been  de- 
signing, constructing,  and  testing  new 
equipment  for  geochemical  and  geophysi- 
cal studies  at  high  pressures.  One  new  de- 
vice is  capable  of  maintaining  50,000  at- 
mospheres at  1700°  C  simultaneously. 

Tilton   and   Davis,    together   with   col- 

leagues at  the  Department  of  Terrestrial 
Magnetism,  have  measured  the  ages  of 
many  rocks  and  minerals,  including  a 
group  from  the  Appalachian  chain.  The 
pre-existing  gneisses  have  been  dated  at 
1100  million  years  (zircons)  and  the  orog- 
eny at  about  300  million  (micas). 

Eugster  and  Milton  (U.  S.  Geological 
Survey)  have  applied  phase-equilibria 
chemical  considerations  to  explain  some 
of  the  bizarre  mineral  assemblages  of  the 
Green  River  shale. 

Libby  has  carried  on  a  series  of  re- 
searches on  geochemistry  of  fission  prod- 
ucts, particularly  Sr90. 

Yoder  has  expanded  his  studies  of  the 
effect  of  water  on  the  melting  relations  of 
rock-forming  silicates,  and  summarizes  all 
the  available  examples. 

Eugster,  Wones,  and  Turnock  have 
studied  pressure-temperature  stability  char- 
acteristics of  a  series  of  hydrous,  iron-bear- 
ing micas  and  chlorites. 

Additional  investigations  in  experimen- 
tal petrology  include  a  study  of  cordierites 
by  Yoder,  Schairer,  and  Schreyer;  of  am- 
phiboles  by  Ernst;  of  pyroxenes  by 
Schairer  and  Morimoto;  and  of  spinels  by 
Wones  and  Turnock. 

The  ore  minerals  have  been  the  subject 
of  intensive  research  by  a  group  consisting 
of  Kullerud,  Arnold,  Barnes,  L.  Clark,  and 
Roseboom.  More  than  a  score  of  systems 
are  under  study,  and  many  new  data  which 
will  ultimately  be  applicable  to  geother- 
mometry  of  ore  deposits  have  been  ob- 

Chayes  has  continued  his  experimental 
investigations  of  order-disorder  diffraction 
effects  both  with  optical  analogs  and  by 
high-speed  calculations. 

Crystallographic  studies  have  been  pur- 
sued intensively  by  Donnay  and  Morimoto, 
Donnay  being  particularly  active  in  studies 
of  sulfides.  A  collaborative  study  of  chal- 
copyrite  by  neutron  diffraction  procedures 




has  provided  many  new  data  about  the 
magnetic  structure  of  this  mineral,  and  in- 
dicates  that  the   substance  is  best  repre- 

sented by  the  formula  Cu+Fe+++S2.  This  and 
other  researches  are  described  in  more  de- 
tail in  the  report  that  follows. 



F.  R.  Boyd  and  J.  L.  England 

Research  at  high  pressures  this  year  has 
been  primarily  devoted  to  the  development 
of  equipment,  although  some  preliminary 
phase-equilibria  data  have  been  obtained 
and  are  presented  below.  Equipment  de- 
velopment is  necessary  to  reach  the  pres- 
sures and  temperatures  required  to  find 
answers  to  many  problems  relating  to  the 
nature  of  the  upper  portion  of  the  earth's 
mantle.  With  the  "squeezer"  apparatus, 
previously  described,  we  have  been  able  to 
reach  pressures  approaching  100,000  atm. 
Such  pressures,  however,  may  be  attained 
with  this  device  only  at  relatively  low  tem- 
peratures (300°  to  400°  C).  Moreover,  our 
work  with  the  transition  quartz  ^  coesite 
in  the  squeezer  revealed  pressure  gradients 
up  to  30  per  cent  of  the  total  pressure.  Such 
gradients  are  probably  not  present  in  all 
systems  investigated  with  this  apparatus, 
but  they  lead  to  an  uncertainty  that  is 
large  in  relation  to  the  applications  to 
which  the  data  may  be  put.  It  has  been 
our  aim  to  develop  equipment  which 
would  extend  the  temperature  and  pressure 
range  beyond  that  attainable  with  the 
squeezer,  and  which  would  be  a  pressure 
system  hydrostatic  within  small,  known 

We  have  been  working  with  two  types 
of  apparatus.  Our  single-stage  apparatus  is 
now  on  a  fully  operational  basis  at  pres- 
sures up  to  50,000  bars  and  temperatures 
up  to  1750°  C.  The  limitation  on  tempera- 
ture is  a  measurement  limit;  a  platinum/ 
platinum-10  per  cent  rhodium  thermo- 
couple melts  at  a  temperature  slightly 
above  1750°  C.  We  are  experimenting  with 
a  tungsten-iridium  thermocouple  in  an 
effort  to  extend  our  measurable  range 
above  1750°  C. 

At  a  pressure  of  about  50,000  bars  the 
limit  of  strength  of  the  piston  in  our  single- 
stage  apparatus  is  reached.  We  have  devel- 
oped a  method  of  supporting  this  piston 
to  reach  pressures  above  50,000  bars.  Ap- 
paratus with  a  supported  piston,  described 
below  as  our  two-stage  apparatus,  has  been 
tested  to  a  pressure  of  65,000  bars  at  a 
temperature  of  1100°  C. 

We  have  begun  to  accumulate  data  on 
a  variety  of  systems.  A  particular  interest 
is  in  obtaining  data  on  the  change  in  melt- 
ing point  of  various  silicates  with  pressure. 
Such  data  are  important  because  they  place 
a  limit  on  the  geothermal  gradient  within 
the  mantle.  Preliminary  data  on  the  melt- 
ing of  diopside  are  presented  below.  In 
addition  to  the  work  on  melting  curves,  we 
are  looking  into  new  solid-state  reactions 
and  checking  diagrams  previously  worked 
out  with  the  squeezer. 

Single-Stage  Apparatus 

Our  single-stage  apparatus  is  shown  in 
figure  1.  In  developing  this  design  we  have 
benefited  from  the  experience  of  Tracy 
Hall  and  Loring  Coes,  who  have  used 
somewhat  similar  apparatus.  The  sample 
and  furnace  assembly  are  contained  in  a 
carbide  pressure  vessel  supported  by  a  steel 
ring.  Pressure  is  applied  by  driving  a 
carbide  piston  into  the  pressure  vessel. 
Power  for  the  furnace  is  supplied  through 
a  stainless-steel  plug  insulated  by  a  ceramic 
ring.  Up  to  four  thermocouple  leads  in  a 
ceramic  tube  may  be  introduced  into  the 
pressure  chamber  through  a  hole  in  the 
power  lead.  The  sample,  which  consists  of 
about  25  mg  of  powder  in  a  platinum  cap- 
sule, is  located  in  the  center  of  a  graphite 
furnace  ll/8  inches  in  length. 

Tests  have  shown  that  the  pressure  gra- 
dient within  the  furnace  and  run  assembly 
is  less  than  5  per  cent.  By  a  calibration  pro- 



cedure  described  below,  the  pressure  on  the 
run  may  be  measured  with  an  uncertainty 
of  less  than  ±  3  per  cent.  Despite  the  small 
size  of  the  furnace,  the  temperature  gra- 
dient is  small.  At  500°  C  we  have  meas- 
ured a  temperature  difference  of  5°  be- 
tween two  thermocouples,  one  placed  in 
the  position  of  the  run  and  one  placed 
Y8  inch  below. 

these  vessels  has  failed  explosively.  The 
steel  ring  is  machined  from  AISI  4340, 
hardened  to  Rockwell  C45,  and  stretched 
on  a  mandrel.  Stretching  a  ring  expands 
its  bore  1  per  cent  and  increases  its  yield 
point  by  a  factor  of  about  2.  The  carbide 
core  with  an  outside  diameter  of  2  inches 
is  fitted  to  the  ring  with  an  interference  of 
0.018  inch.    Tapering  the  ring  and  core 


Power  connection 
Fiber  insulation 

Power  connection 

100  Ton  ram 

I  Inch 

Fig.   1.    Apparatus  for  phase-equilibrium  investigations  in  the  range   10  to  50  kilobars  and  at 
temperatures  up  to  1750°  C.   Steel  parts  are  ruled,  carbide  parts  stippled. 

Our  single-stage  apparatus  has  provision 
both  for  end-loading  the  carbide  pressure 
vessel  and  for  water  cooling.  End-loading 
the  pressure  vessel  extends  its  life.  Water 
cooling  makes  it  possible  to  run  at  a  steady 
state  at  1750°  C.  Without  water  cooling, 
the  steel  supporting  ring  expands  away 
from  the  carbide  and  causes  its  rapid  de- 

Our  pressure  vessels  have  a  life  of  more 
than  20  runs  at  high  temperature.  None  of 

with  a  total  angle  of  1°  enables  them  to  be 
readily  pushed  together  in  a  hydraulic 
press.  The  carbide  core  gradually  cracks 
up  with  use.  When  no  longer  usable,  the 
core  can  be  pushed  out  and  the  ring  fitted 
with  new  carbide. 

Pressure  calibration  in  our  apparatus  is 
achieved  by  measuring  the  transitions  in 
bismuth.  These  lie  at  25  and  27  kilobars  at 
room  temperature,  and  they  have  been  ac- 
curately determined  in  special  apparatus 


by  Bridgman.  We  have  used  them  to  cali- 
brate our  apparatus,  and  to  study  the  effec- 
tiveness of  various  solid  pressure  media. 
The  bismuth  transitions  are  readily  de- 
tected by  measuring  the  change  in  elec- 
trical resistance  in  bismuth  with  pressure. 
The  bismuth  in  the  form  of  a  strip  of  foil 
J4  inch  long  and  about  0.002  inch  thick  is 

range  5  to  7  per  cent.  Jacketing  the  speci- 
mens of  solid  pressure  media  in  lead  foil 
greatly  reduces  the  friction.  Unjacketed 
teflon  gives  a  friction  of  16  per  cent;  with 
lead  foil  the  friction  is  reduced  to  5  per 
cent.  The  hysteresis  on  these  friction  runs 
is  normally  about  three  times  the  friction 
measured  on  raising  the  pressure. 

Power  connection 

Hardened  steel- 
supporting  ring 

f-Fiber  insulation 


I  Inch 

Fig.  2.  Apparatus  for  use  in  the  pressure  range  over  50  kilobars.  Steel  parts  are  ruled,  carbide  parts 
stippled.   The  furnace  and  run  assembly  placed  in  V2  is  similar  to  that  illustrated  in  figure  1. 

mounted  between  two  cylinders  of  a  solid 
pressure  medium.  Gold  leads  are  used  for 
electrical  connections.  The  bismuth  foil 
can  be  mounted  at  any  level  in  a  column 
of  solid  pressure  medium,  and  hence  pres- 
sure gradients  can  readily  be  studied.  The 
solid  pressure  media  we  have  studied  are 
boron  nitride,  talc,  lavastone  (pyrophyl- 
lite),  teflon,  and  Solenhofen  limestone. 
With  boron  nitride  jacketed  in  lead  foil 
we  have  measured  friction  as  low  as  2  per 
cent  at  25  kilobars.  The  other  pressure 
media,  jacketed  in  lead,  yield  values  in  the 

Two-Stage  Apparatus 

Figure  2  illustrates  the  present  stage  of 
evolution  of  our  two-stage  apparatus.  The 
main  pressure  vessel  is  constructed  and 
supported  in  the  same  manner  as  in  our 
single-stage  apparatus,  but  a  second  stage, 
which  supports  the  piston,  is  added.  There 
are  two  volumes,  Vi  and  V2,  under  pres- 
sure. V2,  the  high-pressure  volume,  houses 
the  sample  and  furnace  assembly;  Vi  con- 
tains a  material  (teflon  in  most  of  our 
runs)  which  supports  the  piston.  The  rela- 
tions between  the  volumes,  the  pressures 


Pi  and  P2  in  these  volumes,  the  compressi- 
bilities Bi  and  B2  of  the  materials  in  Vi 
and  V2,  and  the  areas  Ai  and  A2  over 
which  the  piston  bears  are  given  by  the 

P2     (Vi/Ai)Bi 

If  the  volumes  are  cylindrical,  as  in  the 
design  in  figure  2,  the  equation  may  be 
simplified  to 

P1/P2  —  L2B2/L1B1 

where  Li  and  L2  are  the  lengths  of  the 
volumes.  By  adjusting  the  lengths  of  the 
volumes  and  selecting  pressure  media  with 
different  compressibilities,  the  ratio  of  run 
pressure  to  supporting  pressure  can  be 
varied.  We  are  currently  working  with  a 
ratio  of  P2:Pi  of  about  3:1. 

In  principle,  the  pressure  distribution 
between  the  two  volumes  can  be  calculated 
from  a  knowledge  of  the  total  load  on  the 
piston,  the  friction,  the  lengths  of  Vi  and 
Vz,  and  the  relative  compressibilities  of 
the  materials  in  Vi  and  V2.  In  practice,  V2 
will  ordinarily  contain  a  complex  furnace, 
thermocouple,  and  run  assembly  with  ini- 
tial voids,  and  the  compressibility  of  the 
assembly  is  not  known  well  enough  to 
permit  adequate  computation  of  P2. 

We  have,  consequently,  developed  a 
method  of  calibration  that  permits  us  to 
measure  P2  with  an  accuracy  of  about  ±  5 
per  cent.  We  calibrate  by  putting  known 
loads  on  Vi  and  measuring  the  advance  of 
the  piston  as  a  function  of  load.  During 
calibration  the  run  assembly  in  V2  is  re- 
placed by  a  piston  driven  by  a  separate 
hydraulic  press  the  load  on  which  is 
known.  With  a  run  assembly  in  V2,  meas- 
urement of  piston  advance  yields  at  once 
the  load  on  Vi,  and  by  difference  with  the 
total  load  we  have  the  load  on  V2.  This 
method  has  the  particular  advantage  that 
the  friction  in  Vi  need  not  be  independ- 
ently estimated,  since  it  is  automatically 
taken  into  account  in  the  calibration.  The 
friction  in  V2  can  be  estimated  by  making 

a  run  on  a  known  transition.  We  have 
found  the  friction  in  V2  with  lavastone  as 
the  pressure  medium  to  be  6  per  cent  at  the 
bismuth  point,  in  agreement  with  our  data 
from  single-stage  apparatus. 

Our  range  with  the  two-stage  apparatus 
is  currently  about  65,000  at  temperatures 
above  1000°  C;  we  believe  that  minor  im- 
provements will  extend  the  range  consider- 
ably beyond  65,000  bars. 


F.  R.  Boyd  and  J.  L.  England 

Accurate  estimate  of  the  variation  of 
temperature  with  depth  in  the  earth's  man- 
tle is  of  major  importance  to  the  solution 
of  many  geophysical  problems.  Seismic 
data  tell  us  that  rocks  in  the  earth's  mantle 
behave  as  predominantly  crystalline  solids. 
Knowledge  of  the  effect  of  pressure  on 
the  melting  points  of  certain  refractory 
silicates  that  we  can  infer  to  be  constituents 
of  the  mantle  permit  us  to  place  an  upper 
limit  on  the  geothermal  gradient. 

Preliminary  data  on  the  effect  of  pres- 
sure on  the  melting  of  diopside  are  pre- 
sented in  figure  3.  These  data  were  ob- 
tained in  our  single-stage  apparatus.  Yoder 
(1952)  found  that  the  melting  point  of 
diopside  increases  with  pressure  at  the  rate 
of  13.0°  C  per  1000  bars  in  the  range  1  to 
5000  bars.  Present  data  extend  the  melting 
curve  of  diopside  to  32,000  bars  and  1740° 
C.  The  slope  of  the  melting  curve  de- 
creases with  pressure;  in  the  range  20,000 
to  30,000  bars  the  average  slope  is  10.3°  per 
1000  bars.  A  smooth  curve  can  be  drawn 
through  our  points  and  the  data  obtained 
by  Yoder. 

Diopside  liquid  cannot  be  quenched  to 
a  glass  above  about  11,000  bars,  but  we 
have  been  able  to  use  a  pronounced  tex- 
tural  change  in  the  run  to  locate  the  melt- 
ing curve  at  higher  pressures.  The  curve 
presented  has  been  corrected  for  friction  in 
our  apparatus  but  not  for  the  effect  of  pres- 
sure on  the  emf  of  the  thermocouple.  Pres- 
ent data  indicate  that  this  effect  is  small. 








I         I  CoMgSi206 


20  25 





Fig.  3.    Preliminary  curve  for  the  melting  of  diopside  under  pressure. 


S.  P.  Clark,  Jr. 

The  effect  of  pressure  on  melting  tem- 
peratures is  of  considerable  interest  in  the 
physical  theory  of  liquids.  It  is  also  of  im- 
mediate geophysical  importance.  Pressure 
increases  with  depth  in  the  earth,  and  pres- 
ent data  indicate  that  the  temperature  does 
likewise.  Since  large  volumes  of  lava  have 
been  poured  out  at  the  surface  of  the  earth 
in  every  period  of  geologic  time,  partial 
fusion  of  the  material  at  depth  in  the  earth 
must  commonly  take  place. 

Previous  work  on  melting  relations  at 
high  pressure  has  largely  been  confined  to 
metals,  organic  compounds,  and  van  der 

Waals  solids  such  as  He,  H2,  and  N2. 
Simon  found  that  many  experimentally  de- 
termined melting  curves  could  be  repre- 
sented by  the  equation 


where  To  is  the  melting  temperature  at 
zero  pressure  (or  at  the  solid-liquid-vapor 
triple  point),  and  Po  and  c  are  adjustable 

Subsequent  theoretical  investigation  has 
shown  that  this  equation,  originally  con- 
sidered to  be  empirical,  can  be  derived 
from  the  Lindemann  melting  criterion. 
The  derivation  also  leads  to  theoretical 
values  of  Po  and  c,  which  prove  to  be  re- 
lated to  parameters  appearing  in  the  Mie- 
Gruneisen  equation  of  state. 


Although  the  applicability  of  the  Simon 
equation  can  be  tested  on  material  of  any 
kind,  the  relation  between  P0  and  c  as  de- 
termined from  melting  curves,  and  their 
theoretical  values,  can  be  examined  only 
if  the  theoretical  values  can  be  calculated. 
The  calculation  assumes  that  the  substances 
are  fairly  accurately  represented  by  the 
simple  models  of  solids  developed  by 
Debye,  Gruneisen,  and  Born,  and  it  fur- 
ther requires  reasonably  complete  thermo- 


1100   - 

1000  - 

shown  in  figures  4  and  5.  The  breaks  in 
slope  in  the  curves  for  KCl,  RbCl,  and  CsCl 
result  from  the  intersection  of  the  melting 
curves  with  curves  representing  the  ap- 
pearance of  solid  polymorphs.  The  high- 
pressure  form  of  CsCl  on  the  melting  curve 
is  the  same  as  the  form  stable  at  room 
temperature  and  atmospheric  pressure. 
The  other  polymorphic  inversions  are  to 
true  high-pressure  phases  which  had  pre- 
viously been  discovered  at  low  tempera- 

I  900  - 

2    800 

700  - 



10,000  15,000 

Pressure  bars 



Fig.  4.  Melting  curves  of  the  alkali  chlorides. 

chemical  data.  For  these  reasons,  the  ini- 
tial experimental  work  has  been  confined 
to  the  alkali  halides. 

The  melting  curves  of  five  alkali  chlo- 
rides and  four  sodium  halides  were  fol- 
lowed to  a  pressure  of  nearly  25,000  bars. 
The  measurements  were  made  in  the  high- 
pressure  apparatus  developed  by  Francis 
Birch  at  Harvard  University.  Sealed  cap- 
sules containing  the  charges  were  packed 
around  a  thermocouple  inside  the  pres- 
sure vessel,  and  melting  and  freezing  were 
detected  by  arrests  in  heating  and  cooling 

Preliminary  results  for  the  eight  salts  are 

ture.  The  triple  point  of  KCl  may  be  par- 
ticularly useful  in  the  calibration  of  other 
types  of  high-pressure  equipment  at  high 

The  observed  slopes  of  the  melting 
curves  at  low  pressure  agree  with  the 
values  calculated  from  Clapeyron's  equa- 
tion only  for  NaF.  For  the  other  salts  the 
measured  slope  is  always  less  than  that 
calculated  from  thermochemical  data;  in 
some  instances  the  discrepancy  is  as  large 
as  50  per  cent. 

The  cause  of  this  discrepancy  has  not 
yet  been  determined,  and  the  low-pressure 
ends  of  the  melting  curves  are  being  re- 


examined  in  a  different  apparatus.  Results 
obtained  with  it  to  date  are  in  good  agree- 
ment with  those  found  earlier,  and  indicate 
that  the  observations  are  reproducible  and 
free  of  instrumental  bias. 

Other  possible  causes  of  the  discrepancy 
are  systematic  errors  in  the  thermochemical 
data  arising  from  premelting  phenomena, 

and  contamination  of  the  salts  used  in  the 
present  work.  The  main  contaminant  is 
likely  to  be  water,  which  is  notoriously 
difficult  to  remove  from  alkali  halides.  Ex- 
periments are  being  continued  with  mate- 
rial which  has  been  carefully  dewatered, 
and  they  should  aid  in  clarifying  the  situ- 



5000  10,000  15,000 

Pressure  bars 
Fig.  5.  Melting  curves  of  the  sodium  halides. 




(A    cooperative   program    of   the   Geophysical  Laboratory  and  the  Department 

of  Terrestrial  Magnetism  of  the  Carnegie  Institution  of  Washington) 

G.  L.  Davis,  G.  R.  Tilton,  L.  T.  Aldrich}  and  G.  W.  Wetherill  1 

Measurement  of  ages  of  coexisting  min- 
erals in  a  series  of  metamorphic  rocks  has 
revealed  new  opportunities  for  the  appli- 
cation of  dating  methods.  This  was  dis- 
covered first  for  zircon  and  biotite  from 
the  Baltimore  gneiss,  part  of  the  basement 
complex  of  the  Appalachian  Piedmont. 
The  zircon  is  1100  million  years  old,  on 
the  basis  of  nearly  concordant  uranium- 
lead  ages,  and  the  biotite  age  is  established 
as  about  300  million  years  by  rubidium- 

1  Department  of  Terrestrial  Magnetism. 

strontium  and  potassium-argon  methods. 
It  appears  that  the  gneiss  crystallized  1100 
million  years  ago,  and  that  the  biotite  age 
is  related  to  local  metamorphism  some  800 
million  years  later.  It  has  been  established 
that  rocks  with  ages  of  1000  to  1150  million 
years,  previously  known  in  the  southern 
Ontario  and  the  Adirondack  Mountains, 
are  found  in  the  Appalachian  Mountain 
belt  from  the  Catskills  in  New  York  to 
Shenandoah  National  Park,  Virginia,  in- 
creasing considerably  the  geographic  extent 
of  this  group  of  rocks. 


Appalachian  Ages 

The  Appalachian  orogenic  belt,  which 
extends  along  the  Atlantic  coast  of  the 
United  States  and  southeastern  Canada, 
experienced  several  periods  of  deformation 
between  250  and  400  million  years  ago, 
during  Paleozoic  time.  The  process  may 
be  visualized  as  follows.  Great  thicknesses 
of  sediments  were  deposited  on  a  basement 
of  gneisses  and  granites;  both  sediments 
and  basement  sank  into  the  crust,  where 
they  were  subjected  to  the  effects  of  ele- 
vated temperature  and  pressure.  In  this 
environment  some  of  the  crystalline  rocks 
and  sediments  became  altered  to  meta- 
morphic  rocks.  The  metamorphism  was 
accompanied  by  the  formation  or  injection 
of  granitic  rocks.  Later  uplift  and  erosion 
exposed  the  resulting  assemblage,  includ- 
ing gneiss,  schist,  marble,  quartzite,  and 
granite.  Such  a  process  might  be  expected 
to  produce  rather  complicated  age  patterns 
for  some  of  the  rocks.  Several  questions 
may  be  asked  about  this  over-all  process: 
(1)  How  old  are  the  crystalline  rocks  of 
the  basement  complex  ?  (2)  What  was  the 
effect  of  Paleozoic  metamorphism  on  the 
age  of  minerals  in  these  rocks  ?  (3)  What 
was  the  sequence  of  the  metamorphic 
events  from  one  part  of  the  belt  to  an- 
other? Did  these  events  take  place  simul- 
taneously along  the  whole  belt? 

The  Appalachian  orogenic  zone  is  al- 
most ideally  suited  for  studies  of  this  type. 
It  is  old  enough  so  that  erosion  has  pro- 
duced good  exposures  of  rocks  of  widely 
different  ages.  It  is  also  old  enough  for 
daughter  products  to  have  accumulated 
from  the  various  radioactive  parents  in 
sufficient  quantity  to  allow  accurate  age 
determination.  On  the  other  hand,  it  is 
young  enough  to  enable  small  time  inter- 
vals of  the  order  of  10  million  years  to  be 
distinguishable  by  careful  analytical  work. 
In  old  Precambrian  minerals  differences 
of  this  order  would  be  lost  in  analytical 
errors.  Finally,  many  parts  of  the  Appala- 
chians are  well  mapped,  and  some  strati- 
graphic  control  is  available  from  fossil  evi- 

The  Appalachian  age  data  are  given  in 
table  1.  The  first  three  entries — the  Can- 
ada Hill  gneiss,  Storm  King  granite,  and 
Mary's  Rock  gneiss — represent  rocks  hav- 
ing biotite  and  zircon  with  ages  about  1000 
million  years.  An  age  of  1000  to  1150  mil- 
lion years  deduced  from  the  zircon  prob- 
ably dates  the  time  of  crystallization  of 
these  rocks  best.  This  value  is  based  pri- 
marily on  the  U235-Pb207  and  Pb207-Pb206 
ages,  which  should  be  the  more  reliable 
ones.  It  is  apparent  that  metamorphic 
events  which  took  place  300  to  400  million 
years  ago  had  little  effect  on  the  age  rec- 
ord given  by  zircon  and  biotite.  (The  pat- 
tern of  ages  observed  in  these  zircons  could 
not  be  produced  by  removing  lead  from  a 
much  older  zircon,  for  example,  with  an 
age  of  1800  million  years,  in  Paleozoic 
time  since  the  two  U-Pb  ages  would  then 
be  discordant.)  Rocks  with  ages  of  1000 
to  1150  million  years  have  long  been 
known  in  the  Canadian  Shield  in  south- 
ern Ontario,  and  a  zircon  of  this  age  has 
been  found  in  the  Adirondack  Mountains. 
It  now  appears  that  rocks  of  this  age  ex- 
tend along  the  Appalachian  Mountain  belt 
at  least  as  far  south  as  Shenandoah  Na- 
tional Park.  Long  and  Kulp,  of  the  La- 
mont  Geological  Observatory,  have  meas- 
ured Rb-Sr  and  K-A  ages  of  about  900 
million  years  on  biotite  near  Hampton, 
Tennessee.  This  may  be  taken  as  a  good 
indication  that  rocks  of  this  age  group  are 
found  even  farther  to  the  south. 

In  other  parts  of  the  belt  the  mica  ages, 
and  some  zircon  ages  as  well,  were  affected 
to  a  greater  extent.  Examples  of  these 
constitute  the  remainder  of  table  1.  The 
results  from  the  Baltimore  gneiss  are  par- 
ticularly important.  Wasserburg,  Petti- 
john,  and  Lipson  have  recently  determined 
the  K-A  ages  on  micas  from  a  number  of 
metamorphic  rocks,  including  the  Balti- 
more gneiss  and  pegmatites  in  the  Balti- 
more area.  The  values  they  observed  were 
in  the  range  from  300  to  350  million  years, 
and  established  the  time  of  most  recent 
metamorphism  of  the  rocks  at  about  330 
±20  million  years.  Our  values  for  Rb-Sr 


and  K-A  ages  for  biotite  from  the  Balti- 
more gneiss  at  Baltimore  (table  1)  are  in 
this  same  range.  Zircon  from  the  Balti- 
more gneiss,  on  the  other  hand,  gives 
nearly  concordant  and  much  older  values, 
which  are  in  the  range  observed  at  Bear 

years  ago ;  or  the  rock  existed  as  an  uncon- 
solidated sediment  with  detrital  zircon  un- 
til 300  million  years  ago,  then  crystallized 
in  such  a  manner  as  not  to  alter  the  zir- 
con ages  appreciably.  Because  the  zircon 
and  microcline  ages  agree  quite  well  with 

TABLE  1.     Appalachian  Ages 


Age,  mill: 

ion  years 











Bear  Moutain,  N.  Y. 

Storm  King  granite 








Canada  Hill  gneiss 








Shenandoah  National  Park,  Va. 

Gneiss,  Mary's  Rock  Tunnel 








Baltimore,  Md. 

Baltimore  gneiss 

Phoenix  dome 






1200  ± 


Towson  dome 








Woodstock   dome 



Philadelphia,  Pa. 

Baltimore  gneiss 








Hibernia,  N.  J. 

Dark  gneiss 




Light  gneiss 




Spruce  Pine,  N.  C. 

Cranberry  gneiss 








Washington,  D.  C. 

Kensington  gneiss,  Sample  A 







Kensington  gneiss,  Sample  B 







Mountain  and  Mary's  Rock.  Microcline 
from  the  Baltimore  gneiss  at  the  Phoenix 
dome  gives  a  Rb-Sr  age  that  is  in  agree- 
ment with  the  zircon  age;  but  it  appears 
to  have  lost  all  its  argon  during  Paleozoic 
metamorphism.  Two  interpretations  may 
be  given  for  this  array  of  mineral  ages — 
the  rock  crystallized  1100  million  years  ago 
and  biotite  was  recrystallized  300  million 

those  found  for  zircon  and  biotite  at  Bear 
Mountain  and  Mary's  Rock,  and  examina- 
tion of  thin  sections  indicates  that  a  detri- 
tal origin  for  the  microcline  in  the  Balti- 
more gneiss  is  improbable,  it  is  believed 
that  the  gneiss  belongs  to  the  1000-  to  1150- 
million-year  group,  and  that  it  was  suffi- 
ciently metamorphosed  300  to  350  million 
years  ago  to  cause  the  microcline  to  lose 


argon  and  the  biotite  to  lose  both  stron- 
tium and  argon. 

The  Rb-Sr  age  of  390  million  years  for 
the  biotite  in  the  Baltimore  gneiss  at  Phila- 
delphia may  indicate  an  older  biotite 
which  did  not  lose  quite  all  its  strontium 
during  metamorphism;  this  is  so  far  the 
only  mica  age  for  this  area,  however. 
The  higher  K-A  age  is  suggestive.  Last 
year's  report  called  attention  to  similar  dis- 
cordances in  the  ages  of  micas  from  cer- 
tain metamorphic  rocks  from  the  Sudbury 
district,  Ontario.2  Thus,  K-A  ages  higher 
than  Rb-Sr  ages  appear  to  be  characteris- 
tic of  micas  in  some  metamorphic  rocks. 
The  K-A  age  is  somewhat  higher  than 
the  Rb-Sr  age  for  biotite  in  the  Baltimore 
gneiss  at  Towson  dome,  possibly  owing  to 
the  same  effect  and  indicating  that  the  Rb- 
Sr  age  is  a  better  measure  of  the  time  of 
most  recent  metamorphism. 

The  Mary's  Rock-Baltimore  gneiss 
comparison  is  in  agreement  with  field 
observation  that  the  intensity  of  meta- 
morphic processes  varies  across  the  oro- 
genic  belt,  the  rocks  to  the  west  being  less 
affected  than  those  of  the  Piedmont  prov- 
ince to  the  east. 

At  Hibernia,  New  Jersey,  only  mica  de- 
terminations have  been  completed.  The 
Rb-Sr  age  of  the  biotite  from  the  dark 
gneiss  is  similar  to  mica  ages  found  at  Bear 
Mountain  and  Mary's  Rock.  The  ages  for 
the  mica  from  the  light  gneiss  are  lower 
and  would  indicate  a  higher  degree  of 
metamorphism.  Both  micas  probably  be- 
long to  the  1000-  to  1150-million-year 
group,  but  the  light  gneiss  was  more 
highly  metamorphosed  300  to  350  million 
years  ago. 

Results  for  the  Cranberry  gneiss  resem- 
ble those  from  the  Baltimore  gneiss  ex- 
cept that  the  zircon  ages  are  discordant. 
The  Pb207-Pb206  age  of  1270  million  years 
may  indicate  the  presence  of  rocks  older 
than  the  1000-  to  1150-million-year  group 

2  This  observation  has  been  confirmed  inde- 
pendently by  measurements  made  at  the  Massa- 
chusetts Institute  of  Technology. 

found  farther  north,  and  may  be  related 
to  the  occurrence  of  rocks  with  ages  of 
1300  to  1450  million  years  farther  to  the 
west  (see  fig.  6).  The  Rb-Sr  age  of  370 
million  years  for  the  biotite  agrees  with 
ages  established  for  both  uraninite  and 
muscovite  from  the  Chestnut  Flat  pegma- 
tite near  Spruce  Pine  and  appears  to  date 
a  period  of  metamorphism. 

Discordant  ages  are  found  in  zircon  from 
the  Kensington  gneiss,  which  indicate  that 
the  mineral  was  affected  by  the  meta- 
morphism that  occurred  300  to  350  million 
years  ago.  The  age  data  are  consistent 
with  the  hypothesis  that  the  zircon  is  1100 
million  years  old  but  lost  about  90  per  cent 
of  its  lead  300  million  years  ago. 

Other  Ages 

Zircon  from  a  gneiss  at  Koli,  Finland, 
was  also  studied.  The  gneiss  forms  part 
of  the  basement  complex  in  the  Karelidic 
orogenic  belt,  occupying  a  position  analo- 
gous to  that  of  the  Baltimore  gneiss  in  the 
Appalachian  belt.  The  zircon  was  supplied 
by  Kouvo,  who  has  measured  the  age  of 
biotite  from  several  gneisses  in  the  area 
and  found  Rb-Sr  and  K-A  ages  of  1800  mil- 
lion years.  The  results  are  given  in  table 
2.  Unfortunately,  the  ages  are  discordant. 
The  Pb207-Pb206  age  suggests  that  the  zir- 
con is  2600  to  2700  million  years  old.  Pos- 
sibly the  gneiss  is  2600  to  2700  million 
years  old  and  was  metamorphosed  1800 
million  years  ago.  This  may  thus  be  an- 
other example  of  the  effect  observed  for 
the  Baltimore  gneiss. 

Although  most  of  the  central  United 
States  is  covered  with  sediments,  so  that 
the  crystalline  basement  rocks  can  be  ob- 
tained only  as  drill  cores,  exposures  of 
crystalline  rocks  do  occur  in  the  Arbuckle 
Mountains  in  southern  Oklahoma  and  in 
the  St.  Francis  Mountains  and  the  Deca- 
turville  uplift  in  southern  Missouri.  Ru- 
bidium-strontium ages  of  about  1400  mil- 
lion years  have  been  measured  in  these 
areas,  and  are  given  in  table  2.  Many 
micas  with  similar  ages  have  been  found 



in  the  western  United  States  (see  fig.  6).  sota,  and  Kenora,  Ontario.  When  these 
The  Wichita  Mountains,  100  miles  to  the  results  are  combined  with  previous  ones 
west  of  the  Arbuckles,  have  a  decidedly      obtained  here  and  in  other  laboratories — 

Fig.  6.  Reliably  dated  localities  in  North  America. 
TABLE  2.     Miscellaneous  Ages 

Age,  million  years 


Mineral        U238 



pb206  pb20^ 







Koli,  Finland   (55  km  north  of 

Joensuu),  gneiss 

Zircon           1890 

Troy,  Okla. 

Ten  Acre  granite 


Decaturville,  Mo. 



Fredericktown,  Mo. 

Einstein  Mine 


International  Falls,  Minn. 

Biotite  knot  in  Vermilion 



Kenora,  Ont. 









younger  age  of  500  to  550  million  years, 
as  shown  in  last  year's  report. 

Two  new  occurrences  of  micas  with 
ages  of  around  2600  million  years  have 
been  found  at  International  Falls,  Minne- 

particularly  those  by  P.  W.  Gast  and  L.  E. 
Long,  of  the  Lamont  Geological  Observa- 
tory, in  the  Bighorn  Basin  of  Wyoming 
and  Montana — a  belt  of  rocks  with  an  age 
of  2600  to  2700  million  years,  reaching  from 


central  Ontario  to  western  Wyoming  and 
Montana,  appears  to  have  been  defined. 
The  locations  of  these  rocks  are  given  in 
figure  6.  Ages  of  2500  to  2700  million  years 
have  been  determined  by  the  K-A  method 
at  the  University  of  Minnesota  for  mica 
in  rocks  from  northern  Minnesota  and  in 
the  Minnesota  River  valley  in  the  south- 
ern part  of  the  state.  Although  Rb-Sr  ages 
have  not  been  completed,  the  argon  work 
undoubtedly  establishes  the  presence  of 
rocks  with  ages  of  2600  to  2700  million 
years  in  Minnesota. 

North  American  Ages 

Figure  6  shows  the  locations  in  North 
America  where  reliable  mineral  ages  have 
been  obtained.  These  values  are  based  on 
samples  for  which  at  least  two  isotopic  ages 
are  in  agreement,  for  example,  the  two 
U-Pb  ages  of  a  zircon  or  uraninite  or  the 
Rb-Sr  and  K-A  ages  of  a  mica.  The  only 
exceptions  are  the  Arbuckle  Mountains, 
Missouri,  and  Kenora  samples,  for  which 
only  Rb-Sr  measurements  have  been  com- 
pleted. There  are  many  other  localities  for 
which  single  Rb-Sr  or  K-A  ages  or  dis- 
cordant lead  ages  are  available,  but  they 
have  been  omitted.  The  results  shown  in- 
clude the  work  of  several  other  labora- 

Figure  6  shows  rocks  with  ages  of  2600 
to  2700  million  years  extending  probably  as 
a  belt  from  central  Ontario  to  western 
Wyoming.  To  the  south  and  southeast  the 
rocks  appear  younger.  It  is  now  apparent 

that  rocks  with  ages  of  1000  to  1150  mil- 
lion years  are  common  in  the  Appalachian 
Mountains,  uplifted  during  the  Paleozoic 
era;  and  rocks  with  ages  of  2600  to  2700, 
1300  to  1450,  and  1000  to  1150  million  years 
are  not  rare  in  the  region  of  the  Laramide 
orogenic  belt  where  deformation  took 
place  60  million  years  ago. 

It  is  apparent  from  figure  6  that  most 
of  the  localities  that  have  been  measured 
lie  south  of  the  2600-  to  2700-million-year 
belt.  Information  north  of  the  belt  and 
around  the  ends  is  needed.  When  ages 
have  been  accumulated  for  the  whole  of 
the  continent  it  may  be  possible  to  make 
a  critical  evaluation  of  its  mode  of  forma- 
tion. The  question  is  whether  the  con- 
tinent has  grown  from  an  old  "nucleus"  by 
addition  of  materials  from  depth  during 
successive  orogenies  or  whether  it  has 
always  had  considerable  area. 


C.  A.  Hopson  and  A.  C.  Waters  (Johns 
Hopkins  University)  have  provided  valu- 
able assistance  in  the  collection  of  speci- 
mens of  the  Baltimore  gneiss  and  other 
local  rocks.  Thanks  are  also  due  to  W.  E. 
Ham  (Oklahoma  Geological  Survey)  and 
C.  A.  Merritt  (University  of  Oklahoma) 
for  guidance  in  the  collection  of  samples 
in  the  Wichita  and  Arbuckle  Mountains 
in  Oklahoma.  W.  C.  Hayes  (Missouri 
Geological  Survey)  and  Clayton  Johnson 
(University  of  Missouri)  gave  similar  aid 
in  the  collection  of  samples  from  Missouri. 


H.  P.  Eugster 

Since  the  fundamental  analysis  of  marine 
salt  deposits  by  van'tHotf  (1905,  1906), 
the  mineral  assemblages  of  such  deposits 
have  often  been  considered  the  best  ex- 
amples to  demonstrate  the  principles  gov- 
erning the  coexistence  of  minerals.  The 
sequences  of  minerals  obtained  on  evap- 

3  U.  S.  Geology  Survey,  publication  approved 
by  Director. 

id  C.  Milton  3 

oration  from  sea  water  and  upon  subse- 
quent metamorphism  as  well  as  the  asso- 
ciated reaction  temperatures  have  become 
known  through  the  efforts  of  a  number  of 
investigators.  Least  known  today  are  some 
of  the  saline  beds  deposited  from  fresh 
waters,  particularly  those  poor  in  chlorine. 
The  Green  River  formation  of  Wyo- 
ming, Utah,  and  Colorado  contains  such 



saline  beds,  consisting  chiefly  of  sodium 
carbonates,  as  well  as  a  number  of  other 
unusual  minerals.  The  beds  containing 
saline  minerals  are  intercalated  between 
calcareous  (dolomite  and  calcite)  oil  shale 
and  detrital  (often  tuffaceous)  beds  and 
must  have  precipitated  from  vast  lakes 
with  extensive  fluctuation  in  the  water 
levels.  The  compositions  of  the  lake  wa- 
ters were  somewhat  unusual,  since  chlo- 
rine, potassium,  and  sulfate  ions  played  a 
very  minor  role. 

The  coexistence  of  certain  minerals  in  the 
three  major  basins  (Green  River  in  Wyo- 
ming, Uinta  in  Utah,  and  Piceance  Creek 
in  Colorado)  gives  valuable  insight  into 
the  general  conditions  of  evaporation  and 
precipitation  as  well  as  into  local  differ- 
ences. Minerals  considered  are  nahcolite 
(NaHCOs),  trona  (NaHC(VNa2C03- 
2H20),  shortite  (Na2C(V2CaC03),  eite- 
lite  (Na2CCVMgC03),  calcite  (CaCOs), 
dolomite  (CaC03'MgC03),  northupite 
(Na.CCVMgCCVNaCl),  analcite  (Na- 
AlSi2Oe  •  H20) ,  searlesite  (NaBSi206  * 
H20),  reedmergnerite  (NaBSi308),  rie- 
beckite  (Na2  (Mg,Fe)  3Fe2Si8022  (OH)  2) , 
acmite  (NaFeSi206),  sepiolite  (Mg2Si306- 
(OH)4),  and  loughlinite  ((Na2,Mg)Si306- 

The  spatial  distribution  of  trona  and 
nahcolite  is  most  revealing.  Trona  occurs 
in  Wyoming  only  and  in  a  single  bed  10 
feet  thick  and  covering  several  hundreds 
of  square  miles.  Nahcolite  is  characteristic 
of  the  saline  beds  in  Utah  and  Colorado 
and  forms  concretions  and  pockets  inches 
to  feet  in  diameter  in  a  dolomitic  matrix. 
The  absence  of  nahcolite  in  Wyoming  and 
of  trona  in  Colorado  and  Utah  is  indica- 
tive of  local  differences  between  the  evap- 
orating basins. 

Figure  7  shows  a  Pco2-T  diagram  for 
the  system  NaHC03-Na2C03-H20  at  1 
atm  total  pressure,  recalculated  from  data 
determined  by  Freeth  (1923).  The  posi- 
tions of  the  phase  boundaries  are  not  ac- 
curately known,  since  the  calculations  in- 
volve dissociation   constants  and   activity 

coefficients  in  very  concentrated  solutions. 
Determination  of  the  phase  boundaries  by 
direct  experiments  is  under  way.  Nahco- 
lite occupies  the  area  of  high  CO2  content 
throughout,  whereas  soda  (  =  natrite, 
Na2CO3'10H2O)  is  more  characteristic  of 
low  temperatures.  Trona  is  not  stable 
below  19.7°  C,  the  temperature  of  the 
isobaric  invariant  point  nahcolite  +  soda + 

20  30  40 

— —  Temperature    in    °C 

Fig.  7.  Pco2-T  diagram  of  the  system 
NaHC03-Na2C03-H20.  The  C02  content  of 
present-day  air  has  been  indicated  on  the  right. 
Locations  of  phase  boundaries  were  calculated 
from  the  data  of  Freeth  (1923). 

trona  +  solution.  Thermonatrite  will  form 
only  above  34°  C  and  at  low  C02  pressures. 
The  soda-trona  boundary  is  very  steep; 
hence  the  occurrence  of  soda  rather  than 
trona  is  primarily  a  function  of  tempera- 
ture. In  some  lakes  trona  precipitates  dur- 
ing the  warm  seasons  and  soda  during  the 
colder  ones. 

For  the  saline  beds  of  the  Green  River 
formation  the  location  of  the  trona-nahco- 
lite  boundary  is  critical.  As  a  point  of  ref- 
erence the  C02  content  of  present-day 
average  air  (300  to  400  ppm)  has  been  in- 
dicated on  figure  7.  A  solution  of  sodium 
carbonates  equilibrated  with  such  air  will 
precipitate  on  evaporation  both  nahcolite 
and  trona  at  a  temperature  of  about  36°  C. 


The  occurrence  of  trona  in  Wyoming 
rather  than  nahcolite  must  be  due  to 
higher  temperatures  or  lower  C02  pres- 
sures, or  both,  than  those  characteristic  of 
the  Utah  and  Colorado  basins. 

The  concentrations  of  the  brines  neces- 
sary to  precipitate  sodium  carbonates  give 

of  trona  alone  requires  even  greater  con- 
centrations. From  a  comparison  of  figure 
8  with  figure  7  it  becomes  clear  that  the 
highest  concentrations  of  Na+  are  required 
to  saturate  solutions  with  the  lowest  C02 
pressures.  At  very  high  CO2  pressures, 
that  is  from  solutions  virtually  devoid  of 


Undersoturoted     solutions 

1  N 

2  N 

\                            \    cu              Nohcolite 
X                           \ 


\                 \ 


\               \ 


\               \ 


\            ^> 


\               \ 

?  4  n 


A                     \ 

sodium  (No 


\                          O  0-                                    ^^^_ 


\             *o 


Sodo                                 \ 


\                     Trono 



2    6N 


0  \ 

0   ) 



1                      1                      1 


Thermonatrite                         — »_^ 
1                      l                      l 


20  30 

Temperature    in    °C 




Fig.  8.  Concentration  of  sodium  ion  in  solution  as  a  function  of  temperature.  Solid  lines  are  for 
the  system  NaHC03-Na2C03-H20  as  determined  by  Freeth  (1923).  Broken  lines  are  for  the  systems 
NaHC03-H20  and  Na2C03-H20,  respectively  {International  Critical  Tables). 

some  further  clues.  Figure  8  shows  the 
normality  in  sodium  ions  of  saturated  solu- 
tions as  a  function  of  temperature  for  the 
systems  Na2C03-H20,  NaHC03-H20, 
and  NaHC03-Na2C03-H20.  The  solid 
field  boundaries  correspond  to  those  of 
figure  7.  Nahcolite  and  trona  coprecipitate 
at  normalities  between  4  and  5,  depending 
on  the  temperature,  whereas  the  formation 

C03=  ions,  nahcolite  precipitates  at  very 
much  lower  concentrations  of  Na+  (1  to  2 
normal),  as  indicated  by  the  curve  labeled 
"NaHCOs  only." 

The  brines  from  which  trona  precipi- 
tated in  Wyoming  must  have  been  very 
concentrated,  whereas  evaporation  in  Utah 
may  have  been  much  less  extensive.  In 
Wyoming  saturation  occurred  simultane- 



ously  over  a  very  wide  area,  probably  a 
shallow  and  warm  pool,  equilibrated  with 
air.  Average  temperatures  were  higher 
than  35°  C  if  the  C02  content  of  Eocene 
air  was  similar  to  that  of  present-day  air. 
These  temperatures  are  reasonable  since 
modern  briny  lakes  have  been  found  to  be 
as  hot  as  56°  C.  In  the  Utah  and  Colorado 
basins  the  brines  were  probably  covered 
with  fresh  water  before  saturation  was 
reached.  As  a  consequence  of  the  increased 
depth  of  the  lakes  the  temperature  of  the 
brines,  which  were  preserved  in  the  bottom 
layers,  was  lowered,  while  the  C02  con- 
tent increased  owing  to  both  greater  pres- 
sure and  slower  exchange  of  biogenic  C02 
with  the  atmosphere.  Calcium  carbonate 
muds,  which  began  to  precipitate,  isolated 
the  brines  into  pockets,  from  which  on 
further  desiccation  nahcolite  had  to  precip- 
itate.   Hence  the  fundamental  difference 

between  the  Wyoming  and  the  Utah  and 
Colorado  basins  may  have  been  simply  one 
of  depth. 

The  trona-nahcolite  problem  is  only  one 
of  a  great  number  of  similar  questions 
raised  by  the  unusual  mineral  assemblages 
of  the  Green  River  formation.  Little  or  no 
information  is  yet  available  about  condi- 
tions under  which  such  carbonates  as 
shortite,  eitelite,  dawsonite,  burbankite, 
pirssonite,  gaylussite,  bradleyite,  and  nor- 
thupite  and  such  silicates  as  analcite, 
searlesite,  reedmergnerite,  acmite,  riebeck- 
ite,  sepiolite,  loughlinite,  garrelsite,  leuco- 
sphenite,  elpidite,  and  labuntsovite  can  ex- 
ist. They  all  must  have  formed  at  or  near 
room  temperature  and  probably  under 
equilibrium  conditions.  The  interpretation 
of  their  associations  will  eventually  give  us 
direct  insight  into  the  history  of  a  group 
of  ancient  and  very  complex  lakes. 


W.  F.  Libby 

The  Beneficiation  of  Soils  Contaminated 

with  Strontium  90;  Beneficial  Effects 

of  Potassium 

There  appears  to  be  a  possibility  that 
part  of  the  strontium  in  soils,  like  calcium, 
exists  in  forms  unavailable  to  plants  and 
thus  to  the  biosphere.  Evidence  from  the 
fallout  data  of  the  Sunshine  Project  dis- 
closed disparities  between  total  fallout  as 
judged  by  actual  pot  collection  of  rain  and 
the  plant  contents  and  soil  analyses  which 
could  be  due  to  some  type  of  chemical  ag- 
ing or  to  development  of  chemical  inacces- 
sibility by  the  radiostrontium  carried  in  the 
rain.  Experiments  by  several  investigators 
have  shown  that  as  much  as  30  per  cent  of 
the  radiostrontium  in  soils  is  not  acces- 
sible to  plants. 

These  results  indicate  the  possibility  of 
a  beneficiation  of  heavily  contaminated 
soils  by  the  use  of  ordinary  fertilizers  in 
reasonable  amounts.  The  consequences  of 
reactor  accidents  or  local  fallout  during 
wartime  might  thus  be  reduced  consider- 

ably. It  seemed  reasonable  that  the  forma- 
tion of  certain  insoluble  inorganic  com- 
pounds like  strontium  sulfate  might  pro- 
duce such  effects.  Strontium  sulfate  occurs 
in  some  soils  as  the  mineral  celestite,  and 
it  might  be  expected  to  be  sufficiently  in- 
soluble to  accomplish  at  least  a  partial 
segregation  of  soluble  strontium  introduced 
into  the  soil.  It  is  so  insoluble  (solubility 
product  7.6  XlO'7  at  25°  C)  that  it  seemed 
likely  that  in  contrast  to  gypsum,  CaS04* 
2H20,  with  a  solubility  product  of  2.4  X 
10~5  at  25°  C — which  apparently  can  feed 
calcium  into  plants — strontium  sulfate 
strontium  might  be  truly  unavailable  to 
plant  life.  On  the  other  hand,  similar 
considerations  on  barium  have  been  tested 
by  Bradfield  (1932),  and  Robinson,  Whet- 
stone, and  Edgington  (1950),  and  their  re- 
sults show  that  barium  sulfate,  which  is 
even  less  soluble  than  strontium  sulfate, 
can  be  utilized  by  plants  in  certain  soils. 
Since,  however,  the  possibility  seemed  to 
exist  that  the  addition  of  sulfate  to  con- 



taminated  soils  might  be  helpful,  an  in- 
vestigation was  undertaken.  Because  cer- 
tain earlier  work  had  indicated  that  potas- 
sium might  have  a  considerable  beneficial 
effect  on  radiostrontium  absorption,  a 
search  for  a  specific  potassium  effect  was 
undertaken  also.  This  note  is  a  report  of 
the  experiments  to  test  these  two  theories. 
Soil  from  Washington,  D.  C.  (garden  at 
Geophysical  Laboratory),  was  mixed,  2 
parts  to  1  of  the  commercial  soil  thinner 
Vermiculite  and  1  part  of  horse  manure 
fertilizer.  The  soil  used  to  make  the  mix- 
ture had  32  milliequivalents  of  exchange- 
able calcium  per  100  g.  To  about  2  pounds 
of  this  mixture  was  added  in  very  dilute 

Pot  B  was  prepared  in  exactly  the  same 
way  except  that  no  sulfate  was  added. 
After  the  first  crop,  265  mg  of  potassium 
nitrate  was  added  to  test  the  potassium 
effect.  Pot  C  had  no  additions  whatsoever 
except  the  tracer  radiostrontium;  it  served 
as  a  control.  Pot  D  was  filled  with  the 
pure  soil,  unfertilized  and  untreated  with 
Vermiculite.  To  it  was  added  radioactive 
strontium  as  the  solid,  insoluble  strontium 
sulfate,  690  mg  of  the  radioactive  stron- 
tium sulfate  being  used  to  726  g  of  soil,  the 
two  being  intimately  mixed  before  plant- 
ing of  the  radish  seeds. 

The  pots  were  planted  with  radish  seeds 
and  cultivated  by  setting  in  the  ground 

TABLE  3.    Effect  of  Sulfate  and  Potassium  Treatment  of  Radiostrontium-Contaminated  Soils  on 

the  Availability  to  Radish  Crops 



Sr90  Content  of  Radish  Ash 
Carbonates,  arbitrary  units  f 

Pot  A  (370  g  mixture  of 

soil  and  Vermiculite) 
Pot  A  (370  g  mixture  of 

soil  and  Vermiculite) 
Pot  B  (356  g  mixture) 
Pot  B  (356  g  mixture) 
Pot  C  (380  g  mixture) 
Pot  D  (726  g  soil  only) 

8.9  mg  Sr/100  g  as  nitrate+9.5  mg 

K2SO4/100g  0.63 
Above  +  22  mg  Sr/100  g  as  nitrate 

+  50  mg  K2SO4/100  g  0.64,  0.58 

9  mg  Sr/100  g  as  nitrate  0.81 

Above  +  72  mg  KNO3/100  g  0.62,  0.63 

No  additions,  except  tracer  Sr  #  1.00,  1.01, 

43  mg  Sr  */100  g  as  Sr  *S04  0.90 


t  Each  entry  is  one  crop. 

aqueous  solution  approximately  10  micro- 
curies  of  strontium  90.  Four  earthen  pots 
were  used  for  trial  with  radish  seeds  for 
test  of  the  efficacy  of  the  purposeful  addi- 
tion of  S04=  and  K+  in  the  reduction  of 
plant  pick-up  of  the  radiostrontium. 

Pot  A  containing  370  g  of  the  contami- 
nated soil  mixture  was  prepared  as  fol- 
lows: Within  a  few  minutes  after  the 
addition  of  the  radiostrontium  to  the  soil, 
32  mg  of  ordinary  nonradioactive  stron- 
tium was  added  in  dilute  aqueous  solution 
as  nitrate.  The  soil  was  stirred  and  made 
into  a  thick  mud  by  further  addition  of 
water.  After  about  15  minutes,  35  mg  of 
K2SO4  was  added  in  dilute  aqueous  solu- 
tion and  stirred.  After  one  crop,  another 
81  mg  of  strontium  as  nitrate  and  200  mg 
of  K2SO4  were  added  in  the  same  manner. 

outside  in  the  open  during  the  summer  or 
by  exposing  to  a  bank  of  fluorescent  lights 
indoors  in  the  winter.  At  maturity  the 
plants  were  ashed  (after  careful  washing), 
the  ash  was  dissolved  in  dilute  hydro- 
chloric acid,  and  sodium  carbonate  solu- 
tion was  used  to  precipitate  the  insoluble 
hydroxides  and  carbonates,  which  were 
measured  for  strontium  90  content.  The 
results  are  presented  in  table  3. 

The  results  indicate  clearly  that  the  ad- 
dition of  sulfate  is  not  very  effective  as  a 
means  of  reducing  radiostrontium  pickup 
by  crops  grown  on  contaminated  soils.  Al- 
though additional  soluble  strontium  does 
seem  to  have  some  effect,  the  principal  one 
was  caused  by  potassium,  for  which,  at  as 
low  a  level  as  about  60  pounds  per  2  mil- 
lion pounds  of  soil  (or  about  30  pounds 



per  acre  for  normal  2-inch  depth  of  pene- 
tration of  water-soluble  fallout),  something 
like  a  40  per  cent  reduction  of  radiostron- 
tium  uptake  was  observed. 

Although  these  experiments  show  that 
radish  plants  on  certain  kinds  of  soil  cer- 
tainly can  utilize  the  strontium  in  stron- 
tium sulfate,  and  that  the  formation  of 
radiostrontium  sulfate  does  not  necessarily 
reduce  the  uptake  of  radiostrontium,  the 
positive  effect  of  potassium  is  established. 
It  is  possible  that  additional  fertilizers  or 
amendments  may  have  a  more  marked 
effect  than  either  of  the  two  investigated  in 
this  work. 

The  partial  retention  of  radiostrontium 
in  soils  may  involve  effects  other  than 
those  tested  here.  Certainly,  as  strontium 
remains  in  the  soil  it  is  very  likely  eventu- 
ally to  be  incorporated  into  large  crystals, 
where  it  will  become  physically  unavailable 
to  the  plants.  And  so  the  possibility  of 
chemical  aging,  taking  place  slowly  over 
several  years,  exists.  It  does  not  seem 
likely,  however,  that  this  process  will  take 
place  on  a  large  enough  scale  to  return 
heavily  contaminated  soil  to  a  useful  con- 
dition, and  further  work  needs  to  be  done 
on  methods  of  quick  beneficiation. 

Rain  Fallout 

During  the  past  year,  samples  of  rain 
have  been  collected  regularly  to  determine 
the  radioactive  fallout  content.  A  wash- 
tub  (2.5  square  feet  area)  is  placed  out  in 
the  open;  the  water  is  bottled  and  brought 
into  the  laboratory  for  processing.  Larger 
samples  have  also  been  collected  from  a 
roof.  A  rain  gauge  is  used  to  determine 
the  rainfall  in  each  specific  instance. 

It  has  been  well  established  that  radio- 
active fallout  is  mainly  carried  down  by 
rain  and  snow,  the  fraction  deposited  di- 
rectly being  minimal.  Therefore,  by  sam- 
pling the  rainfall  and  snowfall,  it  is  pos- 
sible to  determine  the  radioactive  fallout 
in  a  given  area.  The  purpose  of  the  work 
in  the  Laboratory  has  been  to  settle  a  par- 
ticular point  about  the  nature  of  the  at- 
mospheric storage  mechanism. 

In  the  model  proposed  by  the  author, 
the  fallout  material  introduced  into  the 
stratosphere  is  immediately  mixed  hori- 
zontally to  a  uniform  concentration  and 
has  a  residence  time  of  about  10  years 
there;  that  is,  about  10  per  cent  of  the  ma- 
terial resident  in  the  stratosphere  precipi- 
tates annually.  The  point  of  interest  is 
whether  this  simple  theory  can  actually 
explain  the  amount  of  radioactive  fallout 
observed  at  a  given  locality  and,  more  im- 
portant, its  composition  in  terms  of  the 
short-lived  isotopes;  in  other  words,  can  it 
explain  the  age  of  the  fallout,  the  length  of 
time  that  it  has  been  airborne  before  being 

According  to  this  simple  model,  strato- 
spheric material  is  airborne  for  about  10 
years  and  tropospheric  material  for  only 
1  month  on  the  average.  Therefore,  if 
during  periods  of  heavy  precipitation  heavy 
fallouts  were  observed  to  contain  relatively 
little  of  the  short-lived  isotopes,  it  would 
show  that  the  stratospheric  material  was 
coming  down  and  at  a  variable  rate  which 
could  not  be  maintained  throughout  the 
whole  year.  Some  meteorologists,  particu- 
larly Lester  Machta  of  the  U.  S.  Weather 
Bureau,  have  stated  that  meteorological 
considerations  of  stratospheric  wind  pat- 
terns have  led  them  to  the  conclusion  that 
heavier  fallout  should  be  observed  in  the 
general  latitude  of  40°  to  50°  N.  and  simi- 
larly in  the  southern  hemisphere  at  certain 
seasons  of  the  year.  Therefore,  they  pre- 
dict that  the  heavier  fallout  observed  in 
the  spring  and  summer  of  each  year  in  our 
latitudes  is  mainly  of  stratospheric  origin. 
The  author's  model  predicts  that  the  in- 
creased fallout  would  be  of  tropospheric 
origin,  and  of  younger  age  than  debris  of 
stratospheric  origin.  The  question  of 
which  model  is  correct  is  an  important 
one,  since  it  obviously  affects  the  amounts 
of  fallout  to  be  expected  in  the  populous 
latitudes  of  the  northern  hemisphere. 

The  procedure  for  settling  the  point  is 
straightforward.  Among  the  fission  prod- 
ucts is  barium  140,  which  has  a  half-life  of 
12.8  days  and  is  therefore  appropriate  for 

J£    *    «   *i    -   o 

i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — r2o"i — i — i — r 


N.      (fi      Ifi      Tf      IO      CM 



distinguishing  between  fallout  ages  of  1 
month  and  about  1  to  2  years.  The  radio- 
chemical procedure  for  barium  140  is  sim- 
ilar to  that  for  strontium  90,  and  both  are 
more  sensitive  and  reliable  than  that  for 

10  gives  the  data  obtained  so  far  in  the 
analysis  of  the  Geophysical  Laboratory 
samples  in  Washington,  D.  C.  They  are 
incomplete,  but  they  do  show  an  interest- 
ing  point:    that   during   the    months    of 

_i        10      — 



Fig.  10.  Rain  fallout  in  Washington,  D.  C.  Ba*,  La*,  and  Sr*  separated  chemically  and  measured 
2  weeks  after  collection. 

the  51-day  strontium  89  isotope,  which  is 
particularly  susceptible  to  errors  and  radio- 
active impurities.  Figure  9  shows  the  type 
of  variation  of  strontium  90  fallout  ob- 
served at  Pittsburgh.  Presumably,  the 
Washington,  D.  C,  samples,  when  fully 
analyzed,  will  give  a  similar  curve.  Figure 

February,  March,  and  April  1958  relatively 
little  young  fallout  came  to  Washington. 
Therefore,  these  samples  are  particularly 
important  for  determining  whether  the  rise 
that  has  normally  been  observed  in  the 
spring  in  the  past  years  will  occur  this 


Tritium  Hydrology 

Investigations  already  made  have  indi- 
cated the  usefulness  of  tritium  in  the  study 
of  such  problems  as  the  circulation  of 
ground  water,  the  circulation  of  water 
masses  in  the  ocean,  and  atmospheric  cir- 
culation, as  well  as  in  cosmic-ray  research. 
Another  application  of  tritium  to  hydrol- 
ogy and  geophysics  is  the  use  of  synthetic 
tritium  water  to  label  rather  extensive  areas 
for  hydrological  and  geophysical  purposes. 
Radioactive  water  could  be  distributed 
either  by  low-flying  airplanes,  by  drilling 
and  injection,  or  by  sprinkler  systems  over 
a  given  area  that  is  being  investigated  hy- 
drologically  or  geophysically.  This  type  of 
experiment  is  planned  in  collaboration 
with  Luna  B.  Leopold,  Chief  Hydraulic 
Engineer,  U.  S.  Geological  Survey.  Mr. 
Lee  Thatcher,  of  the  Survey,  is  making 
hydrologic  studies  with  cosmic-ray  and 
bomb  tritium. 

The  counter  and  apparatus  for  measur- 
ing tritium  in  electrolytically  enriched 
water  samples  have  been  installed  and  are 

functioning  well;  they  are  part  of  the  ap- 
paratus used  at  the  University  of  Chicago 
and  brought  to  the  Laboratory  some  time 

The  principal  problem  being  attacked 
at  the  present  with  the  equipment  is  the 
measurement  of  stratospheric  water  for 
tritium  content.  Stratospheric  water  sam- 
ples, collected  in  connection  with  a  balloon 
sampling  program  for  fallout  materials,  are 
being  measured  for  tritium  to  observe 
whether  the  bomb  tests  have  appreciably 
contaminated  the  stratosphere  with  tritium. 
It  is  expected  that  they  will  not  show  very 
serious  contamination,  because  the  bomb 
tritium  is  immediately  burned  to  water  and 
is  diluted  with  enormous  masses  of  ordi- 
nary tropospheric  moisture,  which,  when 
the  bomb  cloud  reaches  the  stratosphere, 
condenses  into  ice  crystals  in  the  cool  upper 
air  and  falls  back  into  the  troposphere  in 
a  matter  of  a  few  days.  This  is  in  sharp 
contrast  to  the  fission  products,  which  re- 
main airborne  for  years  when  the  cloud 
reaches  into  the  stratosphere. 



H.  S.  Yoder,  Jr. 

Most  magmas  contain  some  water,  and 
it  has  been  shown  that  water  is  an  impor- 
tant factor  in  magmatic  processes.  In  re- 
cent Year  Books,  for  example,  the  large 
shift  in  the  composition  of  the  "eutectic" 
of  simple  rock-forming  mineral  systems 
under  high  water  pressure  was  depicted 
(Yoder,  Year  Book  53,  p.  107;  Stewart, 
Year  Book  56,  p.  215) .  In  addition,  it  was 
shown  that  water  greatly  expands  the  re- 
gion in  which  crystals  are  in  equilibrium 
with  liquid  in  basalts  (Yoder  and  Tilley, 
Year  Book  55,  p.  150).  In  other  systems 
the  effect  of  water  on  the  liquid  phase  was 
elucidated  under  conditions  where  a  gas 
phase  is  prohibited  (Yoder,  Stewart,  and 
J.  R.  Smith,  Year  Book  55,  p.  208).  These 
and  other  effects  are  important  in  deter- 
mining the  course  of  events  in  a  crystal- 

lizing magma.  Above  all,  however,  the 
most  significant  role  of  water  is  the  great 
lowering  of  the  liquidus  temperature  of 
minerals  and  mineral  systems.  At  H20 
pressures  corresponding  to  depths  of  only 
12  miles,  the  lowering  of  the  liquidus  may 
be  100°  C  (CaMgSi206-H20)  or  as  great 
as  700°  C  (NaAlSi04-H20) !  Data  ob- 
tained to  date  by  various  workers  on  the 
melting  behavior  of  common  rock-forming 
silicates  under  high  water  pressure  are 
summarized  in  figure  11. 

In  addition  to  the  great  lowering  of  melt- 
ing temperatures  by  water,  the  effect  of 
water  on  polymorphism  and  incongruent 
melting  can  be  seen  from  the  figure.  In 
the  Si02-H20  and  NaAlSi04-H20  sys- 
tems the  high-temperature  polymorphs  are 
no  longer  stable  at  elevated  water  pres- 
sures. The  fields  of  cristobalite  and  tridy- 
mite  in  the  Si02-H20  system  are  succes- 
sively   suppressed    with   increasing   water 


pressure  so  that  quartz  melts  directly  to  a 
liquid  saturated  with  water.  In  a  similar 
way  the  field  of  carnegieite  in  the  NaAl- 
Si04-H20  system  is  suppressed  so  that  a 
form  of  nepheline  melts  directly  to  a 
water-saturated  liquid.  In  the  KAlSisOs- 
H20  system  the  incongruent  melting  be- 
havior of  high-sanidine  to  leucite  and  liq- 
uid is  suppressed  and  high-sanidine  melts 
directly  to  a  water-saturated  liquid.   This 

The  lowering  of  the  melting  tempera- 
tures of  these  common  rock-forming  min- 
erals with  increasing  water  pressure  also 
appears  to  support  the  idea  advanced  by 
Morey  (1922)  that  very  high  pressures  may 
develop  during  the  cooling  of  a  contained 
hydrous  magma.  If  certain  conditions  are 
met,  according  to  Morey,  the  hydrous 
magma  on  cooling  will  follow  a  univari- 
ant  curve  similar  to  those  in  figure  11.  As 

1200         I  1400 

Temperature,  °C 

Fig.  11.  Projection  of  univariant  liquidus  curves  in  portions  of  the  following  systems:  (1) 
Si02-H00:  Tuttle  and  England  (1955);  Yoder  (unpublished,  1955);  Stewart  (unpublished,  1957). 
(2)  CaAl2Si208-H20:  Yoder  (unpublished,  1953);  Stewart  (unpublished,  1957).  (3)  NaAlSi04- 
H20:  Yoder  (unpublished,  1958).  (4)  KAlSi308-H20:  Goranson  (1938);  Bowen  and  Tuttle 
(1950);  Yoder,  Stewart,  and  J.  R.  Smith  (unpublished,  1957).  (5)  NaAlSi308-H20:  Goranson 
(1938);  Bowen  and  Tuttle  (1950);  Yoder  (unpublished,  1953);  Yoder,  Stewart,  and  J.  R.  Smith 
(unpublished,  1957).   (6)  CaMgSi906-H20:  Yoder  (unpublished,  1953). 

suggests  that  the  effect  of  water  on  the 
melting  temperature  of  leucite  is  more 
drastic  than  on  that  of  sanidine. 

One  of  the  important  discoveries  in  this 
study  is  that  the  melting  temperatures  of 
the  rock-forming  silicates  in  the  presence 
of  water  are  well  within  those  believed  to 
exist  in  the  crust.  The  observations  sup- 
port the  view  that  magmas  exist  in  the 
crust,  which  on  cooling  would  consist 
solely  of  highly  refractory  minerals  such 
as  anorthite,  the  principal  constituent  of 

the  temperature  falls  the  pressure  of  gas  in 
equilibrium  with  crystals  and  liquid  rises 
sharply.  Morey  further  predicted  from  his 
analysis  of  salt  solutions  and  the  system 
K2Si03-Si02-H20  that  the  pressure  would 
rise  to  a  maximum  and  then  decrease  with 
further  cooling.  It  appears  from  the  data 
in  the  figure  that  this  maximum  pressure, 
if  it  obtains,  exceeds  that  compensated  by 
the  weight  and  strength  of  the  crustal 
rocks,  approximately  10,000  bars  at  the  base 
of  the  crust  35  kilometers  in  depth.  If  the 
melt  curves  continue  to  rise  to  high  pres- 


sures  with  cooling,  it  would  be  necessary 
to  conclude  that  all  contained  hydrous 
magmas  in  the  crust  must  eventually  break 
the  containing  strata  on  cooling  and  in- 
trude the  country  rock  or  extrude  onto  the 
surface  o£  the  earth.  Geologic  evidence 
indicates,  however,  that  many  magmas  be- 
lieved to  have  contained  H2O  cool  with- 
out greatly  deforming  their  chamber  or 
breaking  out  to  the  surface.  Because  of 
the  discrepancy  between  conclusions  drawn 
from  laboratory  and  field  observations,  let 
us  examine  more  closely  the  conditions  re- 
quired by  Morey's  theory. 

It  is  assumed  by  Morey  that  the  walls  of 
the  magma  chamber  are  essentially  rigid 
and  impermeable.  But  rocks  are  known 
to  vary  in  permeability,  and  some  undergo 
deformation  readily.  It  is  possible,  there- 
fore, that  the  gas  pressure  may  be  relieved 
by  diffusion  of  some  of  the  gas  through 
the  walls  of  the  chamber,  by  the  formation 
of  a  hydrous  phase,  or  by  enlargement  of 
the  chamber  itself.  Perhaps  the  most  im- 
portant condition  to  be  met  for  pressure 
generation  is  that  the  magma  must  reach 
a  univariant  condition  in  the  early  stages 
of  cooling.  That  is,  the  magma  can  have 
only  one  degree  of  freedom  as  prescribed 
for  the  univariant  curves  of  the  simple 
binary  systems  shown  in  figure  11.  If  the 
magma  is  not  univariant,  the  vapor  pres- 
sure could  be  relieved  by  changing  the 
relative  proportion  of  crystals,  liquid,  and 
vapor.  Recent  experiments  on  basalts  and 
granites  suggest  that  the  condition  of  uni- 
variancy  is  met  early  in  the  cooling  of  a 
relatively  dry  magma:  the  principal  phases 
in  both  these  major  magma  types  appear 
together  within  a  very  narrow  temperature 
interval  when  only  several  per  cent  of 
crystals  are  present.  With  increasing  water 
content,  however,  the  principal  phases  ap- 
pear together  within  a  much  larger  tem- 
perature interval,  and  the  condition  of 
univariancy  is  probably  met  only  in  the 
final  stages  of  cooling.  The  restrictions  of 
a  rigid,  impermeable  magma  chamber  and 
the  condition  of  univariancy  do  not  appear 
to  be  easily  met  in  a  cooling  complex 

Two  other  factors  must  be  considered. 
It  is  possible  that  the  melt  curves  are  in- 
terrupted by  critical  phenomena  or  by 
phase  changes.  Critical  phenomena  have 
not  been  expected  in  magmas,  according 
to  Morey,  because  of  the  large  number  of 
components  and  the  concentration  of  the 
more  volatile  components  in  the  residual 
liquid.  The  less  complex  magmas,  how- 
ever, may  exhibit  critical  phenomena  in 
the  light  of  the  data  on  common  rock- 
forming  minerals.  When  the  critical  point 
is  reached,  the  magma  loses  a  phase  and 
becomes  divariant.  The  pressure  may 
then  be  compensated  by  changes  in  the 
relative  proportions  of  crystals  and  gas. 
Phase  changes  may  be  anticipated  in  mag- 
mas containing  analcite,  nepheline,  and 
quartz,  which  are  known  to  have  high- 
pressure  polymorphs.  The  formation  of 
an  additional  phase  will  decrease  the  de- 
grees of  freedom.  These  factors  will  in- 
fluence pressure  generation. 

The  explosive  volcanoes  attest  to  the 
fact  that  the  special  conditions  for  pressure 
generation  are  met  in  magmas  of  a  wide 
range  of  composition,  yet  it  is  not  ex- 
pected from  the  arguments  stated  above 
that  the  conditions  will  necessarily  be  met 
in  the  majority  of  hydrous  magmas  solely 
by  cooling.  External  forces,  causing  a  rise 
of  the  magma  to  a  lower  pressure  region, 
for  example,  may  be  the  principal  agent 
attending  the  explosive  release  of  water 
in  magmas. 

A.  C.  Turnoc\  and  H.  P.  Eugster 

Of  the  group  of  layered  silicates  (micas, 
chlorites,  clay  minerals)  the  chlorites  have 
received  the  least  experimental  attention. 
Yoder's  work  on  clinochlore  (1952)  is  the 
only  quantitative  information  available  on 
the  stability  of  a  chlorite.  Yoder  also  was 
first  to  point  out  the  polymorphic  relations 
between  the  7  A  phases  (kaolinite  types) 
and  the  14  A  phases  (true  chlorites). 
Since  most  natural  chlorites  contain  con- 
siderable amounts  of  iron,  it  was  thought 
advisable  to  study  the  phase  relations  of 


iron-rich  chlorites  as  well.  Daphnite, 
5FeO  •  AI2O3  •  3Si02  •  4H20,  the  iron  analog 
o£  clinochlore,  was  chosen  as  an  example. 
A  7  A  daphnite  was  synthesized  from  a 
variety  of  starting  materials  at  tempera- 
tures as  low  as  400°  C.  Even  at  pressures 
as  high  as  5000  bars  it  has  not  been  pos- 

+  quartz  at  intermediate  Po2,  and  mag- 
netite+mullite+ quartz  at  high  Po2.  Ged- 
rite  was  formed  in  several  runs,  but  was 
shown  to  be  a  metastable  product.  Iron 
cordierite  was  also  encountered  between 
600°  and  650°  C.  Its  position  in  the  equi- 
librium diagram  is  not  yet  established. 







20  40  60 

Composition  of  mofic  phases  in 




Fig.  12.  Phase  relations  of  biotites  on  the  join  phlogopite-annite  at  a  constant  pressure  (Ptot= 
Ph2o  +  Ph2  =  2000  bars).  Ph,  phlogopite;  Bi,  biotite;  H,  hematite;  M,  magnetite;  Fe,  iron;  Sa,  sani- 
dine;  Ol,  olivine;  Lc,  leucite;  Ks,  kalsilite;  Ph(H,Sa),  locus  of  all  phlogopites  coexisting  with  H  +  Sa, 
and  so  on.  Only  surfaces  representing  micas  coexisting  with  other  phases  have  been  drawn  (I  to  V). 
The  surfaces  conjugate  to  surfaces  I  to  V  have  been  omitted. 

sible  to  convert  this  form  wholly  to  the 
14  A  polymorph.  Po2  was  controlled  in  all 
experiments  employing  the  usual  oxygen 
buffers.  At  2000  bars  water  pressure  the 
highest  temperatures  at  which  daphnite 
was  synthesized  for  a  series  of  oxygen  pres- 
sures are  as  follows:  525°  C  (hematite- 
magnetite  buffer),  600°  C  (quartz-mag- 
netite-fayalite  buffer),  650°  C  (magnetite- 
wiistite  buffer).  The  following  high-tem- 
perature assemblages  were  found  on  the 
daphnite  composition:  fayalite  +  hercynite 
+  quartz  at  low  Po2,  magnetite  4-  hercynite 

The  mica  biotite  is  one  of  the  most  com- 
mon iron-containing  minerals.  The  phase 
relations  of  two  important  end  members 
of  the  biotite  group  have  already  been 
clarified:  phlogopite,  the  magnesian  mica 
(Yoder  and  Eugster,  1954) ;  and  annite, 
which  contains  ferrous  iron  (Eugster,  Year 
Book  56) .  A  third  end  member,  a  ferrous- 
ferric  biotite,  is  now  under  investigation. 
Unquestionably  the  most  important  sub- 
stitution in  the  biotite  group  is  that  of  Fe+2 
for  Mg+2.    It   is   accompanied   by    major 



changes  in  physical  properties  as  well  as 

Since  annite  is  the  best-known  iron-rich 
end  member  of  the  biotites,  it  will  be 
most  fruitful  to  consider  the  phase  rela- 
tions on  the  join  phlogopite-annite.  Such 
an    analysis,    with    appropriate    modifica- 

positional  variations  can  be  treated  most 
successfully  by  plotting  Fe/(Fe  +  Mg) 
ratios  of  the  phases  concerned. 

Figure  12  shows  a  three-dimensional  sec- 
tion with  Ph2o  fixed  at  2000  bars  (Ptot= 
Ph2o  +  Ph2  =  2000  bars)  for  the  join  phlog- 
opite  (left)-annite  (right).  The  curves  (a), 



10  '  10"  10" 

Partial  pressure  of  -*— 
oxygen  in  bars 


40  60  80  100 

Composition  of  mafic  phases  'r>F7^.  -ratios 

10  10 

— ►  Partial  pressure  of 
oxygen  in  bars 

Fig.  13.  T-X  projection  of  the  phase  relations  on  the  join  phlogopite-annite  at  a  total  pressure 
of  2000  bars  (Ptot  =  Ph2o -}- Ph2)  projected  along  the  Po2  axis.  Po2-T  data  for  the  end  members 
phlogopite  (left)  and  annite  (right)  have  been  rotated  into  the  same  plane.  Projections  of  the 
intersections  of  conjugate  surfaces  onto  the  T-X  plane  have  been  labeled  for  Mg-Fe  phases  only, 
omitting  feldspars  and  feldspathoids.  Data  for  biotites  at  an  oxygen  pressure  of  a  Ni  +  NiO  buffer 
are  dashed  and  labeled  [Ni/NiO].  Abbreviations  as  in  figure  12. 

tions,  is  equally  applicable  to  all  Mg-Fe 
solid  solutions  and  can  be  carried  out  using 
the  data  on  the  end  members  only. 

Phase  Relations  of  Hydrous  Silicates  with 
Intermediate  Mg/Fe  Ratios 

H.  P.  Eugster  and  D.  R.  Wones 

All  iron-bearing  members  of  Mg-Fe 
solid  solutions  are  affected  by  changes  in 
the  partial  pressure  of  oxygen.  Conse- 
quently Po2,  Ph2o,  Ptot,  and  T  must  be 
considered  as  independent  variables.  Com- 

(c),  and  (e)  on  the  right  and  left  sides  of 
the  block  diagram  represent  the  reactions 
iron  +  oxygen  ^  wiistite  (a),  wiistite  +  oxy- 
gen ^  magnetite  (c),  and  magnetite + 
oxygen  ^  hematite  (e).  Their  position  is 
essentially  that  given  by  Darken  and  Gurry 
(1945)  with  appropriate  corrections  for 
total  pressure.  The  right  side  panel  gives 
the  data  for  annite  (see  Year  Book  56,  p. 
162,  fig.  11),  whereas  phlogopite  on  the 
left  is  stable  up  to  a  vertical  line  going 
through  points  A\  C,  and  E'  at  1085°  C. 



Between  these  panels  a  volume  is  repre- 
sented within  which  biotites  are  the  stable 
phases.  This  volume  is  bounded  by  the 
four  curved  surfaces  I,  II,  III,  and  IV.  I  is 
defined  by  FEE'F,  II  by  ECCE\  III  by 
CAA'C,  and  IV  by  GAA'G'. 

Biotites  lying  on  these  surfaces  can  co- 
exist with  four  different  assemblages  as 
follows:  I,  biotites  coexisting  with  hema- 
tite+sanidine  [Bi(H,Sa)];  II,  biotites  co- 
existing with  magnetite  +  sanidine  [Bi(M, 
Sa)];  III,  biotites  coexisting  with  olivine 
+  leucite  +  kalsilite    [Bi(01,Lc,Ks)];    IV, 

In  order  to  construct  figure  13  the  curves 
EE',  CC,  and  AA  of  figure  12  have  been 
projected  onto  the  base.  Simultaneously 
the  left  and  right  sides  of  the  block  dia- 
gram have  been  rotated  into  the  same 
plane,  giving  the  data  for  phlogopite  in  the 
left  panel  and  those  for  annite  in  the  right 
one.  The  Fe/(Fe  +  Mg)  ratios  of  the 
phases  that  coexist  with  biotites  along  the 
above  curves  have  been  added  in  figure  13. 
Data  for  the  Ni/NiO1  buffered  runs  have 
also  been  plotted.  Figure  13  is  useful  for 
presenting  experimental  results,  but  very 


Wt    %  Annite 

Fig.  14.   Plot  of  index  of  refraction  and  ^(060)  value  versus  composition  for  biotites  on  the  join 
phlogopite-annite  synthesized  at  several  temperatures  and  oxygen  pressures  of  a  Ni-NiO  buffer. 

biotites  coexisting  with  iron  +  sanidine 

Figure  13  indicates  that  biotites  can  co- 
exist with  hematite  only  along  the  hema- 
tite-magnetite boundary,  but  not  within 
the  field  of  stability  of  hematite  proper. 
This  was  demonstrated  experimentally  for 
Mg-rich  biotites  (80  weight  per  cent  phlog- 
opite), employing  a  Cu/Cu20  buffer, 
which  has  a  slightly  higher  oxygen  pres- 
sure than  magnetite/hematite  at  650°  C. 

On  surface  II,  which  represents  biotites 
coexisting  with  magnetites,  data  deter- 
mined with  a  Ni/NiO  buffer  are  shown. 
The  Poo-T  curve  for  Ni/NiO  lies  well 
within  the  magnetite  field. 

often  figure  12  must  be  consulted  to  eluci- 
date specific  spatial  relations. 

The  Phlogopite- Annite  Join 
D.  R.  Wones 

The  composition  of  biotites  on  the 
phlogopite-annite  join  coexisting  with  (1) 
hematite  +  magnetite  +  sanidine  +  vapor, 
(2)  magnetite  +  sanidine  +  vapor,  and  (3) 
olivine  +  leucite  +  kalsilite  +  vapor  at  fixed 
temperatures,  total  pressure,  and  Po2  is  be- 
ing studied.  The  results  will  determine 
the  location  of  the  surfaces  I,  II,  III,  and 
IV  of  figure  12. 

Oxygen  buffers  are  used  in  all  runs. 
Loss  of  iron  into  the  platinum  crucibles 


is  minimized  by  placing  a  silver  capsule 
within  the  platinum  tube  and  surrounding 
it  with  the  same  material  as  the  charge. 

Compositions  of  the  equilibrium  biotites 
are  determined  by  measuring  the  indices 
of  refraction  and  the  position  of  the  (060) 
reflection.  Standards  were  synthesized  un- 
der a  variety  of  conditions.  Some  experi- 
ments yield  small  amounts  (1  per  cent)  of 
metastable  sanidine,  magnetite,  and  pyrox- 
ene. Figure  14  shows  index  of  refraction 
and  ^(060)  in  A.  U.  plotted  as  a  function 
of  composition  of  biotites  synthesized  at 
600°,  700°,  and  800°  C,  30,000  psi  total  pres- 
sure, using  Po2  of  a  Ni-NiO  bufTer.  In- 
dices of  refraction  appear  to  be  functions 
of  Po2  as  well  as  Fe/Mg  ratios  with  efTects 
most  pronounced  in  the  iron-rich  members. 

Phase  boundaries  were  established  by 
forming  equilibrium  assemblages  at  a 
given  Ptot  (  =  Ph2o  +  Ph2),  temperature, 
and  Po2  from  a  homogeneous  biotite  as 
well  as  from  a  mixture  of  K20'4Si02, 
Si02,  Y-AI2O3,  MgO,  Fe,  and  H20  at  dif- 
ferent bulk  compositions.  Runs  of  different 
durations  demonstrated  that  5  days  are 
necessary  to  achieve  equilibrium.  In  figure 
14  it  may  be  noted  that  biotites  of  80  and  90 
per  cent  annite  (by  weight)  react  at  800° 
C  to  form  a  biotite  of  the  composition  60 
per  cent  annite  +  sanidine  +  magnetite  + 
vapor.  Line  HH'  in  figures  12  and  13  is 
the  result  of  these  experiments  and  defines 
surface  II  at  the  particular  sets  of  T  and 
Po2  used.  Corresponding  experiments  are 
being  carried  out  using  buffers  Cu  +  Cu20, 
Fe203  +  Fe304,  quartz  +  f ayalite  +  magnet- 
ite, Fe304  +  FeO,  and  FeO  +  Fe. 

Ferrous-Ferric  Biotites 
D.  R.  Wones 

Further  studies  of  the  substitution  of  Fe 
in  micas  have  been  made  by  synthesizing 
a  mica  of  the  composition  K206FeO* 
Fe203,6Si02.  This  aluminum-free  biotite 
was  originally  described  by  Veres,  Meren- 
kova,  and  Ostrovski  (1955).  Its  stability 
field  is  quite  different  from  that  of  the 
aluminous  analog,  annite.   The  following 

sequence  of  mineral  assemblages  at  30,000 
psi  total  (water)  pressure  for  a  series  of 
oxygen  pressures  was  found:  (1)  Po2  of  a 
hematite  +  magnetite  buffer:  660°  C,  fer- 
rous-ferric biotite;  670°  C,  iron-sanidine 
+  hematite  +  magnetite;  680°  C,  melt  + 
hematite  +  magnetite.  (2)  Po2  of  a  quartz 
+  magnetite +  f  ayalite  buffer:  795°  C,  fer- 
rous-ferric biotite;  800°  C,  melt  +  magnet- 
ite. (3)  Po.,  of  a  wustite  +  magnetite  buf- 
fer: 680°  C,  ferrous-ferric  biotite;  750°  C, 
melt +  fayalite  + magnetite.  (4)  Po2  of  an 
iron  +  wustite  buffer:  700°  C,  ferrous-ferric 
biotite;  750°  C,  fayalite  +  melt. 

At  850°  C,  30,000  psi  total  (water)  pres- 
sure, and  the  Po2  of  an  iron  +  wustite 
buffer,  these  biotites  recrystallize  before 
melting,  forming  crystals  as  large  as  a  mil- 
limeter in  diameter.  Single  crystals  pre- 
pared under  these  conditions  are  being 
used  in  crystal-structure  analysis  by  G. 
Donnay  and  co-workers. 

W.  Schreyer  and  H.  S.  Yoder,  Jr. 

Cordierite  is  often  a  constituent  of  ex- 
trusive as  well  as  intrusive  igneous  rocks 
of  a  wide  range  of  composition,  but  the 
most  common  occurrence  is  in  meta- 
morphic  contact  aureole  rocks  such  as  spot- 
ted slates  and  hornfelses.  The  mineral  is 
also  encountered  in  some  regionally  meta- 
morphosed rocks  such  as  the  cordierite- 
anthophyllite  schists,  cordierite-sillimanite 
gneisses,  granulites,  and  charnockites. 
Knowledge  of  the  stability  of  cordierite 
and  its  polymorphs  contributes,  therefore, 
to  an  understanding  of  a  wide  range  of 
rock  types. 

Cordierite  is  usually  assumed  to  have 
the  composition  (Mg,Fe)2Al3(AlSi5)Oi8, 
but  most  analyses  of  natural  cordierites 
show  up  to  3  per  cent  water  and  minor 
amounts  of  alkalies.  It  was  necessary, 
therefore,  to  study  the  system  Mg2Al3- 
(AlSi5)Oi8-Fe2Al3(AlSi5)Oi8  both  dry 
and  in  the  presence  of  water.  The  prob- 
lems concerning  its  polymorphism,  compo- 
sition, and  lower  stability  limits  were  in- 


vestigated  in  the  magnesium  end  member 


Previous  experimental  work  indicated 
that  at  least  three  forms  of  Mg2Al3(AlSi5)- 
Ois  existed:  a  high-temperature  form  syn- 
thesized in  the  dry  way,  now  called  india- 
lite  ( =  a-cordierite)  ;  a  low-temperature 
form  synthesized  hydrothermally,  called 
cordierite  (  =  (3-cordierite) ;  and  a  meta- 
stable  form,  called  u-cordierite.  The  inver- 
sion between  cordierite  and  indialite  was 
believed  to  be  about  830°  C  at  1000  bars 
water  pressure  (Yoder,  1952)  extending  to 
about  600°  C  at  200  bars  water  pressure 
(Iiyama,  1958).  These  polymorphs  were 
distinguished  on  the  basis  of  refractive  in- 
dices. Recent  X-ray  measurements  on  nat- 
ural material  confirmed  the  existence  of 
a  high-temperature  indialite  and  a  low- 
temperature  cordierite,  and  demonstrated 
that  there  were  structural  states  intermedi- 
ate between  the  two  forms  (Miyashiro, 

The  present  hydrothermal  studies  at 
2000  and  5000  bars  water  pressure  confirm 
the  existence  of  intermediate  states  in  the 
vicinity  of  the  alleged  inversion  on  the 
basis  of  indices  of  refraction.  It  is  not 
known  as  yet  whether  these  states  are 
time  dependent,  mainly  because  the  optical 
properties  and  X-ray  diffraction  properties 
do  not  appear  to  change  concomitantly. 
That  is,  a  change  in  the  refractive  indices 
may  take  place  without  a  significant 
change  in  the  X-ray  diffraction  pattern, 
and  vice  versa.  It  can  be  said  with  cer- 
tainty that  a  low-temperature  form  of  cor- 
dierite identical  to  that  from  Albany 
County,  Wyoming,  and  Guilford,  Con- 
necticut, has  been  synthesized.  A  re-exam- 
ination of  cordierites  previously  prepared 
hydrothermally  (Yoder,  1952)  indicated 
they  were  similar  to  but  not  identical  with 
the  low-temperature  form. 


The  shape  of  the  curve  representing  the 
beginning  of  the  change  from  cordierite 

to  some  intermediate  state  is  similar  to 
that  of  a  decomposition  curve  of  a  hydrous 
mineral.  Experiments  are  now  under  way 
to  determine  whether  the  low-temperature 
form  contains  water,  and  to  correlate  the 
water  loss  of  natural  cordierites  with  any 
changes  in  optical  and  X-ray  properties. 

In  addition  to  the  question  of  the  water 
content,  there  is  also  some  doubt  as  to  the 
ratio  of  the  nonvolatile  components.  The 
cordierites  synthesized  in  the  wet  way  al- 
ways exhibit  minute  inclusions  which  look 
like  tiny  spinel  crystals.  Assuming  that 
they  are  spinels,  this  fact  might  be  attrib- 
uted to  loss  of  silica  to  the  vapor.  Further 
investigations  will  be  necessary  to  indicate 
the  true  nature  of  these  inclusions,  which 
might  be  liquid  as  well.  All  our  runs  on 
materials  of  the  cordierite  composition 
above  about  800°  C,  however,  indicated 
clearly  a  loss  of  silica  as  well  as  some  mag- 
nesia to  the  vapor.  At  these  elevated  tem- 
peratures, but  well  below  the  incongruent 
melting  of  cordierite,  we  obtained  minor 
amounts  of  spinel  (outside  the  cordierite 
crystals),  sapphirine,  and  corundum  in  ad- 
dition to  cordierite.  The  quenched  vapor 
consisted  partly  of  balls  up  to  0.5  mm  in 
diameter  of  a  highly  siliceous  glass  with  a 
very  low  refractive  index. 

Lower  Stability  Limits 

Magnesium  cordierite  is  known  to  be 
stable  up  to  the  solidus,  where  it  breaks 
down  to  mullite  and  liquid.  Its  lower  sta- 
bility limit,  however,  has  only  been  exam- 
ined in  a  cursory  way.  Both  natural  and 
synthetic  materials  of  the  magnesium  cor- 
dierite composition  were  subjected  to  a 
wide  range  of  water  pressures  and  temper- 
atures to  fix  the  lower  stability  limit  of 
cordierite.  Below  about  500°  C,  2000  bars, 
and  550°  C,  5000  bars,  the  stable  assem- 
blage is  chlorite  (amesite)  +  pyrophyllite. 
In  the  vicinity  of  400°  C  pyrophyllite  gives 
way  to  a  magnesium-bearing  montmoril- 
lonite.  These  magnesian  montmorillonites 
have  a  wide  range  of  solid  solution,  and 
possibly  a  montmorillonite  solid  solution 



of  the  cordierite  composition  itself  exists  at 
the  lowest  temperatures.  On  the  basis  of 
the  experiments  performed  it  is  believed 
that  cordierite  is  produced  in  a  sediment 
of  the  requisite  composition  while  under- 
going metamorphism  in  the  following 
way.  The  original  sediment  probably  con- 
tains a  magnesian  montmorillonite  solid 

semblage.  The  relation  of  cordierite  to 
chlorite,  pyrophyllite,  montmorillonite,  and 
a  hypothetical  hydrous  cordierite,  2MgO 
2Al203,5Si02-H20,  is  shown  in  figure  15. 
The  water-deficient  region  is  stippled.  In 
this  region  cordierite  may  be  stable  at  tem- 
peratures as  low  as  that  on  the  earth's 


T  =  400°  C 
P=  2000  bors 



2MgO-AI203-Si02  •     Cordierite 

Weight  per  cent 



Fig.  15.  Assemblages  stable  at  approximately  400°  C  and  2000  bars  in  the  section  2MgO'Al203- 
Si02-Al203'4Si02-H20.  The  solid  solution  of  magnesian  montmorillonite  has  been  neglected.  The 
triangular  point  is  the  hypothetical  composition  2MgO#  2 Al2Os  -5Si02'H20.  The  portions  of  the 
triangle  bordered  by  dashed  lines  contain  additional  phases  in  MgO-Al203-Si02-H20  not  repre- 
sented in  this  section.   The  stippled  area  is  the  water-deficient  region. 

solution  and  possibly  an  aluminous  ser- 
pentine. The  latter  inverts  to  a  chlorite 
and  the  montmorillonite  becomes  less  mag- 
nesium rich  with  increasing  temperature. 
Then  the  montmorillonite,  essentially  free 
of  magnesium,  inverts  to  pyrophyllite.  At 
some  higher  temperature  the  chlorite  and 
pyrophyllite  react  to  form  cordierite.  Such 
may  be  the  growth  sequence  leading  to  a 
cordierite-bearing  spotted  slate. 

Future  studies  of  cordierite  will  include 
investigation  of  its  behavior  in  the  water- 
deficient  region  with  water  contents  less 
than  that  in  the  chlorite-pyrophyllite  as- 

W.  Schreyer  and  /.  F.  Schairer 

The  basic  structure  of  cordierites  con- 
tains Mg,  Al,  Si,  and  O,  and  it  possesses 
channels  which  may  under  appropriate 
conditions  be  filled  in  whole  or  in  part 
with  water  or  alkalies.  In  order  to  ascer- 
tain the  possibilities  of  variations  in  the 
composition  of  anhydrous  alkali-free  cor- 
dierites, we  are  at  present  re-examining 
parts  of  the  system  MgO-Al203-Si02.  In 
their  study  of  the  system  MgO-Al203- 
Si02,  Rankin  and  Merwin  (1918)  indi- 
cated that  there  might  be  solid  solution  in 



cordierite  between  the  limits  2MgO* 
2Al203-5Si02     and     MgO  •  A1203  •  3Si02. 

We  have  prepared  a  series  of  melts  on  the 
line  spinel  (MgO'Al203) -silica  with  47, 
48,  50,  51.35  (2:2:5),  52.5,  53.5,  54.5,  55.87 
(1:1:3),  56.5,  58,  60,  65.95,  67.40,  69.84, 
and  73.24  weight  per  cent  silica.  These 
melts  have  been  crystallized  and  are  being 
subjected  to  thermal,  optical,  and  X-ray 
studies.  It  is  already  apparent  that  there 
is  no  variation  in  the  composition  of  cordi- 
erite between  2:2:5  and  spinel.  The  com- 
position 50  per  cent  Si02  when  completely 
crystalline  consists  of  cordierite  with  an 
amount  of  spinel  easily  detected  under  the 
microscope.  Cristobalite  is  present  with 
cordierite  and  can  also  easily  be  detected 
under  the  microscope  in  the  completely 
crystallized  compositions  with  56.5  per  cent 
silica  or  more.  Studies  of  compositions  be- 
tween 2:2:5  and  1:1:3  are  now  in  prog- 
ress. If  the  composition  of  cordierite  is 
variable  owing  to  solid  solution,  crystal- 
lization paths  of  the  points  lying  in  the 
field  of  cordierite  should  be  curved  rather 
than  straight.  Nine  compositions  that 
should  lie  in  the  field  of  cordierite  have 
been  prepared,  and  a  study  of  their  crys- 
tallization paths  is  now  in  progress. 

In  order  to  study  the  course  of  crystal- 
lization of  the  cordierite  polymorphs,  glass 
of  pure  2:2:5  composition  was  subjected 
at  atmospheric  pressure  to  various  temper- 
atures for  various  lengths  of  time.  No  crys- 
tallization was  observed  at  500°  and  700° 
C  after  2  months.  At  800°  C  after  22  days 
the  sample  showed  no  sign  of  devitrifica- 
tion when  examined  optically  but  yielded  a 
faint  peak  in  the  X-ray  pattern.  After  58 
days  the  rims  of  the  glass  fragments  were 
birefringent  and  the  X-ray  pattern  was 
similar  to  that  of  quartz,  but  with  the 
peaks  definitely  shifted  to  lower  20  angles. 
At  900°  C  after  1  day  the  glass  had  crys- 
tallized to  the  same  quartzlike  structure, 
which  after  longer  heating  gradually  dis- 
appeared and  gave  rise  to  a  cordierite 
structure.  At  the  temperatures  1000°, 
1100°,  1200°,  and  1300°  C  the  glass  readily 

crystallized  to  cordierite.  According  to  its 
thermal  behavior  and  optical  properties, 
the  metastable  quartzlike  structure  ob- 
tained at  800°  and  900°  C  corresponds  to 
the  so-called  u-form  of  cordierite  described 
by  Rankin  and  Merwin  (1918).  In  a  re- 
cent study  Karkhanavala  and  Hummel 
(1953)  also  obtained  the  u-form  and  be- 
lieved it  to  have  a  structure  similar  to  that 
of  (3-spodumene,  LiAlSi2Os.  Our  own  X- 
ray  studies,  however,  suggest  that  its 
structure  is  more  closely  related  to  that  of 
eucryptite,  LiAlSi04.  Winkler  (1948)  has 
synthesized  this  mineral,  and  on  the  basis 
of  single-crystal  studies  indicated  that  it 
had  a  high-quartz  structure  with  the  Li 
ions  filling  a  hollow  spiral  in  the  lattice. 
In  a  study  of  the  lithium  metasilicate- 
spodumene-silica  system  Roy  and  Osborn 
(1949)  encountered  an  apparently  similar 
metastable  quartzlike  phase,  which  can 
take  various  amounts  of  Li  into  solid  solu- 
tion. Since  the  Mg  ion  has  essentially  the 
same  size  as  the  Li  ion,  it  is  likely  that  our 
silica  phase  contains  Mg  in  solid  solution. 
It  is  not  certain,  however,  that  its  compo- 
sition is  the  same  as  that  of  cordierite,  in 
other  words,  that  the  glass  has  completely 
crystallized  to  this  phase. 

The  X-ray  patterns  of  all  the  cordierites 
made  in  the  temperature  range  between 
900°  and  1250°  C  showed  a  single  sharp 
peak  between  29°  and  30°  20.  According 
to  Miyashiro  (1957),  therefore,  they  are 
indialites,  the  so-called  high-temperature 
form  of  cordierite.  For  most  of  the  high- 
melting  compositions  examined,  runs  of 
5  days  or  longer  at  temperatures  between 
1420°  and  1450°  C,  however,  yielded  prod- 
ucts exhibiting  at  least  two  peaks  in  this 
range,  in  this  respect  corresponding  closely 
to  most  of  the  natural  cordierites.  Either 
type  of  pattern  or  an  intermediate  stage 
may  result  from  a  run  of  less  than  5  days 
at  this  temperature.  The  one-  and  multi- 
peak  phases  do  not  differ  in  refractive 
indices.  The  effect  of  slow  cooling  and 
of  quenching  at  various  rates  from  a  series 
of  temperatures  is  now  under  investiga- 



tion.  Cordierites  crystallized  from  the 
lower-melting  compositions  in  the  cor- 
dierite  stability  field  may  represent  the 
multi-peak  form  at  temperatures  as  low  as 
1350°  C. 

W.  G.  Ernst 

The  most  important  alkali  amphiboles 
belong  to  the  glaucophane-riebeckite  series, 

been  reported.  Extensive  solid  solution 
exists  among  the  other  three  end  members. 
Glaucophane  and  riebeckite  are  found 
principally  in  low-grade  metamorphic 
rocks;  some  iron-rich  riebeckites  occur  in 
alkalic  igneous  bodies.  The  presence  of 
sodic  amphiboles  in  rocks  of  diverse  com- 
position has  led  to  much  speculation  about 
their  mode  of  origin.  Laboratory  investi- 
gation has  been  undertaken  in  an  attempt 



>Na2  Fe+3+ 

Fe2+  +  Si8°22(0H)2 



Fe+++for   Al- 

>Na2Mg3  Al£  Si 


°Na2Mg3  Fe+++Sig  0^  (0H)2 

©Crystalline  schists        O  Low- grade  metasediments      •  Igneous  rocks 
Fig.  16.    Chemical  variation  in  natural  members  of  the  glaucophane-riebeckite  series. 

which  has  four  end  members:  glauco- 
phane, oNa2Mg3Al2Si8022(OH)2;  ferro- 
glaucophane,  °  NasFe^ALSisC^  (OH)  2 ; 
magnesioriebeckite,  °  Na2Mg3Fe2+++Si8022- 
(OH)2;  and  riebeckite,  °Na2Fe3++Fe2+f+- 
Sis022(OH)2.  (The  symbol  °  represents 
a  vacant  position  in  the  structure.) 

Chemical  analyses  of  natural  glauco- 
phanes  and  riebeckites  are  plotted  in  fig- 
ure 16.  No  amphibole  closely  approaching 
the  composition  of  ferroglaucophane  has 

to  evaluate  the  physical  and  chemical 
parameters  that  govern  the  stability  of 
these  amphiboles. 

Magn  esioriebec\ite 

X-ray  diffractometer  patterns  of  syn- 
thetic magnesioriebeckite  crystallized  over 
a  wide  Po2-^vaPor- T  range  are  mutually  in- 
distinguishable, and  are  similar  to  that  of 
natural  magnesioriebeckite  from  Cocha- 
bamba,    Bolivia    (USNM    4980).     Calcu- 


lated  lattice  parameters  for  the  synthetic 
amphibole  are  somewhat  different  from 
those  determined  by  Whittaker  (1949)  for 
another  sample  of  the  Cochabamba  ma- 
terial (see  table  4). 

TABLE  4.     Unit  Cell  Dimensions  of  Synthetic 
and  Natural   Magnesioriebeckite 


Magnesioriebeckite  * 

a  =  10.04  ±  0.01  A 
£=18.02  ±0.02  A 
c=  5.28  ±0.01  A 
5  =  72°  01' ±03' 

a—  9.89  A 
£=17.95  A 
c=   5.31  A 

5  =  72!/20 

E.  J.  W.  Whittaker  (1949). 

acterizes  magnesioriebeckite  grown  at  high 
partial  pressure  of  oxygen;  at  fixed  partial 
oxygen  pressure,  magnesioriebeckite  crys- 
tallized at  high  temperature  has  a  low  re- 
fractive index.  Equilibrium  has  been  dem- 
onstrated by  cycling  synthetic  magnesio- 
riebeckite at  both  higher  and  lower  Po2; 
refractive  indices  are  independent  of  start- 
ing material  and  duration  of  run.  Refrac- 
tive indices  are  assumed  to  be  a  measure 
of  the  oxidation  state  of  the  iron  in  mag- 
nesioriebeckite:  the  lower  the  index  of  re- 
fraction, the  smaller  the  Fe203/FeO  ratio. 
Electrostatic  neutrality  is  probably  main- 
tained (1)  through  replacement  of  oxygen 
by   hydroxyl,    (2)    through   partial   occu- 










1                           1 

1.657   ^ 

rn^-  — 

652                  1.656 










-  — — —   ""           "" 



_ "CJi.640 


. ^"1  640 











Ol  633 
1                            1                           1 






800  820  840  860  880 

Temperature   in  °C 


Fig.  17.  Variation  in  Nz  of  magnesioriebeckite  with  Po2  and  T.  The  apparently  anomalous 
refractive  index  of  the  amphibole  crystallized  at  901°  C  results  from  the  fact  that  it  formed  within 
the  melting  interval  of  magnesioriebeckite  (described  last  year)  and  is  more  magnesian  than  the 
phase  stable  below  the  magnesioriebeckite  solidus  surface. 

The  variation  in  optical  properties  of 
synthetic  magnesioriebeckite  with  Po2  and 
T  is  illustrated  in  figure  17.  At  a  given 
temperature,  high  index  of  refraction  char- 

pancy  of  the  vacant  position  by  cations 
(H3O+?),  or   (3)    through  minor  solid 
solution  with  griinerite  or  arfvedsonite. 
The  influence  of  partial  oxygen  pressure 


on  the  stability  field  of  magnesioriebeckite 
is  shown  in  figure  18.  At  a  specified  total 
pressure,  increased  Po2  elevates  the  temper- 
ature at  which  magnesioriebeckite  breaks 
down.  All  condensed  phases  stable  above 
the  amphibole  field  contain  both  MgO  and 

gen  pressure  declines,  causing  all  mag- 
nesium-bearing phases  to  become  enriched 
in  ferrous  iron.  The  orthopyroxene  stable 
adjacent  to  the  magnesioriebeckite  field 
changes  composition  from  En98  with  a 
hematite-magnetite  buffer  to  En79  with  a 

-5  — 

-10  — 

-15  — 


1                1 

1                                     ' 


/    Hematite    +    Olivine    + 
Orthopyroxene    +   Liquid 

/                     +   Vapor 




/        Magnesioferrite    +    Olivine    + 

/            Orthopyroxene     +     Liquid 


/                                     +    Vapor 

oNo2Mg3Fe+2*+Si8022   (0H)2 


_.  „  -^y  ^^            1 

^ — ■"""      Magnesiowustite    + 

— -^0"^    Olivine  +  Orthopyroxene  +  Liquid 

^®^^     / 

+   Vapor             ^ — 

^^             J_^ 

1                                      1 

1                        1 


900  950 

Temperature   in   °C 


Fig.  18.  Po2-T  diagram  for  the  composition  Na20'3MgO'Fe203'8Si02  with  excess  water  at  2000 
bars  vapor  pressure.  Curves  (1),  (2),  (3),  and  (5)  are  defined  by  the  oxygen  buffers  hematite- 
magnetite,   magnetite  +  silica-fayalite,,   magnetite-wiistite,    and    wustite-iron,    respectively. 

FeO,  with  the  exception  of  hematite  and 
possibly  acmite.  The  ratio  MgO/FeO 
among  coexistent  phases  is  highest  in  the 
lowest-melting  crystalline  phase,  lower  in 
the  more  refractory  minerals,  and  lowest  in 
the  liquid.  At  fixed  temperature  and  total 
pressure,  the  MgO/FeO  ratio  of  the  bulk 
composition  decreases  as  the  partial  oxy- 

magnetite  + silica-fayalite  buffer,  and  to 
En73  with  a  magnetite-wiistite  buffer;  co- 
existent olivine  over  the  same  Po2  interval 
changes  from  Fo95  to  Fo73  and  to  Foes. 
Magnesioferrite  decreases  in  amount  (as  a 
greater  portion  of  the  ferric  iron  is  re- 
duced) and  approaches  the  composition 
Fe304  with  a  magnetite-wiistite  buffer. 



The  effect  of  the  partial  pressure  of 
oxygen  on  the  composition  and  stability 
relations  of  minerals  encountered  in  this 
system  illustrates  the  petrologic  importance 
of  this  variable. 

An  intermediate  member  of  the  mag- 
nesioriebeckite-riebeckite  series  has  been 
found  as  an  authigenic  mineral  in  the 
Green  River  shale  of  the  Colorado  Pla- 
teau, indicating  that  the  magnesioriebeck- 
ite  stability  field  extends  to  low  tempera- 
tures and  pressures.  The  rarity  of  this 
mineral  is  in  part  due  to  the  fact  that  few 
rocks  contain  Na20  in  excess  of  AI2O3. 
However,  magnesioriebeckite  has  been 
found  in  a  few  crystalline  schists  whose 
bulk  compositions  show  no  excess  of  soda 
over  alumina.  In  this  case,  the  presence 
of  sodic  amphibole  suggests  special  phys- 
ical conditions;  the  nature  of  these  condi- 
tions will  be  discussed  in  the  next  section. 


The  preliminary  P-T  stability  field  of 
°Na2Mg3Al2Si8022(OH)2  up  to  2000  bars 
vapor  pressure  is  presented  in  figure  19. 
The  slope  of  the  breakdown  curve  of  this 
amphibole  is  unusually  steep. 

Forsterite,  enstatite,  albite,  glaucophane, 
liquid,  and  vapor  coexist  at  an  invariant 
point  at  867°  C  and  1500  bars.  With  pres- 
sures greater  than  1500  bars,  glaucophane 
melts  incongruently  to  forsterite,  enstatite, 
liquid,  and  vapor.  The  high-temperature 
assemblage  adjoining  the  glaucophane  field 
at  pressures  lower  than  1500  bars  consists 
of  forsterite,  enstatite,  albite,  and  vapor. 

At  966°  C  and  500  bars,  the  melting 
curve  of  albite  and  the  incongruent  melt- 
ing curve  of  enstatite  intersect.  Below  500 
bars,  the  liquidus  curve  for  enstatite  coin- 
cides with  solidus  curves  for  enstatite  and 
albite;  above  500  bars,  the  liquidus  curve 
for  albite  coincides  with  solidus  curves  for 
albite  and  enstatite.  The  locus  of  the  inter- 
section is  a  point  on  a  univariant  curve 
along  which  five  phases  (forsterite,  ensta- 
tite, albite,  liquid,  and  vapor)  are  stable. 
The  two  liquidus   (and  solidus)   surfaces 

intersect  for  a  range  of  bulk  compositions. 

Results  of  the  experimental  investiga- 
tion indicate  that  glaucophane  is  not  itself 
a  high-pressure  mineral.  This  fact  is  of 
considerable  petrologic  significance  since 
many  geologists  consider  glaucophane  the 
characteristic  mineral  of  a  low-tempera- 
ture, high-pressure  metamorphic  facies. 
Their  conclusion  is  based  on  the  evidence 
that  (1)  glaucophane  schists  are  more 
dense  than  parent  sediments  and  lavas,  and 
typical  metamorphic  rocks;  (2)  many 
glaucophane  schists  are  chemically  equiva- 
lent to  green  schists  and  amphibolites;  (3) 
glaucophane  schists  are  developed  on  a 
regional  scale  in  some  areas;  and  (4)  jade- 
ite,  experimentally  demonstrated  to  be  a 
high-pressure  mineral,  has  been  found  in 
certain  glaucophane  schists.  Other  workers 
have  cited  the  unusual  bulk  compositions 
of  many  glaucophane  schists,  local  develop- 
ment of  glaucophane  schists,  often  adja- 
cent serpentinites,  and  the  intimate  asso- 
ciation of  glaucophane  schists  with  green 
schists  and  amphibolites  as  indication  that 
glaucophane  schists  owe  their  production 
to   peculiar  chemical  conditions. 

In  an  attempt  to  integrate  field  and  lab- 
oratory data,  glaucophane  schist  localities 
of  the  California  coast  ranges  were  visited 
during  July  and  August  1957.  Irregularly 
distributed  schist  bodies  have  developed 
locally  in  the  severely  folded  and  faulted 
Franciscan  formation,  a  series  of  arkosic 
subgraywackes  with  intercalated  argilla- 
ceous cherts,  basaltic  and  spilitic  lavas,  and 
serpentine  intrusives;  small  masses  of  ec- 
logite  are  present  rarely. 

The  Tres  Pinos  Creek  area  (San  Benito 
quadrangle)  was  examined  in  some  detail. 
Here  glaucophane-  and  riebeckite-bearing 
schists  have  developed  from  graywackes, 
cherts,  and  basic  extrusives  with  no  ap- 
parent relation  to  serpentinite.  Gradational 
contacts  between  sheared  Franciscan  for- 
mation and  glaucophane  schist  indicate 
that  the  initial  stage  of  metamorphism  in- 
volved granulation  and  the  production  of 
lawsonite,  minor  glaucophane,  and  stilp- 



nomelane;  in  highly  recrystallized  schists, 
lawsonite  has  been  replaced  by  aggregates 
of  epidote  or  clinozoisite.  Except  for  the 
relative  proportions,  metagraywackes  and 
metabasalts  display  similar  mineralogy, 
consisting  of  glaucophane-crossite  +  law- 
sonite or  clinozoisite-epidote  +  stilpnome- 

these  glaucophane  schists  probably  formed 
at  low  temperatures.  The  restriction  of 
the  metamorphic  rocks  to  zones  of  shear- 
ing may  be  an  indication  that  unusual 
pressures  were  required  for  the  formation 
of  these  alkali  amphibole-bearing  rocks 
and/or  that  solutions  migrating  along  fa- 


Forstente  t  Enstatite  +  Liquid  +  Vapor 






Temperature   in  °C 



Fig.  19.    Preliminary   PvaDOr-^   diagram    for    the   composition    Na20'3MgO-Al203-8Si02    with 
excess  water. 

lane +  muscovite  + quartz  (+ garnet)  + 
ores.  Metacherts  contain  crossite-riebeckite 
+  sodic  pyroxene  +  stilpnomelane  +  quartz 
(+ garnet?)  +ores.  The  different  min- 
eralogy of  the  metacherts  is  due  to  the  low 
CaO  and  AI2O3  content.  Albite  is  ab- 
sent in  metamorphic  rocks  of  the  area  in- 
vestigated (except  as  a  vein  mineral). 
Rapid  lateral  gradation  into  unmetamor- 
phosed  Franciscan  formation  suggests  that 

vorable  channelways  fluxed  the  relatively 
sluggish  reaction  involving  conversion  of 
granulated  sediments,  flows,  and  cherts  to 
a  stable  low-grade  metamorphic  assem- 

Returning  to  the  general  problem,  it 
may  be  observed  that  metamorphic  rocks 
of  various  compositions  generally  contain 
both  amphibole  and  plagioclase.  The  par- 
tition of  Na20  and  CaO  among  these  and 


other  phases  depends  on  the  physical  and 
chemical  conditions  accompanying  the 
metamorphism;  commonly  sodium  is  con- 
centrated in  feldspar  while  amphibole  and 
other  ferromagnesian  minerals  contain  the 
calcium.  Where  special  bulk  compositions 
are  deficient  in  CaO,  as  in  the  experi- 
mental investigation,  glaucophane  can 
exist  over  a  wide  P-T  range.  Rock  com- 
positions with  soda  in  excess  of  alumina 
would  favor  the  production  of  intermedi- 
ate members  of  the  glaucophane-riebeck- 
ite  series,  since  they  contain  more  Na20 
than  AI2O3.  The  instability  of  sodic  pla- 
gioclase  in  rocks  of  normal  chemical  com- 
position at  high  pressure  would  also  pro- 
mote the  formation  of  sodic  amphibole  or 
sodic  pyroxene,  or  both. 

In  conclusion  it  may  be  stated  that,  al- 
though some  glaucophane-bearing  rocks 
are  formed  under  common  metamorphic 
temperatures  and  pressures  in  response  to 
hypersodic  chemical  conditions,  others  are 
rocks  of  normal  bulk  compositions  that 
have  been  subjected  to  relatively  high  total 


Investigation  of  0Na2Fe3++Fe2+++Si8022- 
(OH)2  is  under  way.  To  pressures  in  ex- 
cess of  2000  bars,  riebeckite  breaks  down 
to  fayalite,  magnetite,  quartz,  acmite,  and 
vapor,  using  a  magnetite  +  silica-fayalite 
buffer.  Preliminary  data  at  2000  bars  indi- 
cate that  the  stability  limit  of  riebeckite 
lies  between  600°  and  700°  C. 


E.  G.  Zies  and  F.  Chayes 

The  pseudoleucites  rank  high  among 
the  familiar  but  little-understood  mineral 
associations  characteristic  of  alkaline  rocks. 
The  mineral  leucite  occurs  primarily  in 
volcanic  rocks;  it  is  found  occasionally  in 
dike  rocks  or  very  shallow  intrusives 
clearly  associated  with  volcanics,  but  has 
never  been  described  from  a  truly  plutonic 
environment.  From  a  number  of  alkaline 
areas     mineral     aggregates     are     known 

which  exhibit — sometimes  poorly,  some- 
times with  remarkable  sharpness — the 
crystal  form  of  leucite,  but  contain  none  of 
that  mineral.  These  pseudoleucites  are 
usually  mixtures  of  sanidine  or  orthoclase 
with  one  or  more  of  the  feldspathoids  or 
zeolites.  Nepheline  is  perhaps  the  com- 
monest, but  sodalite,  cancrinite,  and  anal- 
cite  are  all  well  known. 

The  leucite-like  shape  was  long  ago 
taken  as  an  indication  that  at  some  stage 
of  their  history  these  aggregates  had  in 
fact  been  leucite.  As  far  as  is  known,  how- 
ever, natural  leucite  is  always  poor  in  Na, 
a  constituent  ordinarily  abundant  in 
pseudoleucite.  Conversion  of  an  initially 
Na-poor  leucite  into  an  Na-rich  aggregate 
of  nepheline  and  orthoclase  was  regarded 
by  Bowen  as  an  expectable  consequence  of 
crystal  fractionation  from  melts  of  appro- 
priate composition.  He  made  it  the  basis 
of  an  ingenious  theory  about  the  origin  of 
alkaline  rocks  and  their  relation  to  biotite 

Firm  quantitative  information  about  the 
chemical  and  mineralogical  composition 
of  pseudoleucites  is  still  very  scarce.  The 
pseudoleucite  in  a  porphyritic  tinguaite 
from  the  Bearpaw  Mountains,  Montana, 
seemed  to  afford  an  ideal  opportunity  for 
combined  chemical  and  modal  analysis. 
It  was  collected  by  one  of  us,  from  a  dike 
that  outcrops  along  the  fire  road  to  Elk 
Peak,  during  a  brief  visit  to  the  area  under 
the  guidance  of  R.  Schmidt,  of  the  U.  S. 
Geological  Survey. 

Our  specimen  contains  numerous  white 
subangular  nodules,  vaguely  suggestive  of 
leucite  in  outline,  consisting  of  inter- 
growths  and  mixtures  of  nepheline  and 
sanidine  together  with  a  little  aegerine 
and  an  occasional  grain  of  other  minerals. 
No  analcime  or  sodalite  has  been  noted; 
the  nepheline  is  virtually  unaltered;  and 
the  occasional  fibrous  alteration  (sericite 
or  cancrinite)  on  sanidine  is  so  scarce  that 
it  has  not  been  possible  to  identify  it 

The  white  patches  are  mostly  between 


x/i  and  1  cm  in  maximum  dimension,  but 
they  do  not  liberate  readily  from  the  ma- 
trix of  the  rock  until  crushed  to  a  size  at 
which  hand  sorting  is  impracticable.  Rock 
fragments  were  therefore  crushed  to  pass 
a  no.  7  cloth,  and  a  concentrate  was  ob- 
tained by  electromagnet  from  the  fraction 
retained  on  a  no.  9  cloth.  The  bulk  analy- 
sis of  this  concentrate  and  the  analyses  of 
portions  of  it  soluble  and  insoluble  in 
boiling  1  : 5  HCl  are  shown  in  table  5. 

TABLE  5.  Chemical  Analyses  of  a  Pseudoleu- 
cite  from  Elk  Peak,  Bearpaw  Mountains, 












































Table  6  gives  the  results  of  modal  analyses 
of  two  types;  columns  A  and  B  are  the 
average  modes  of  six  stained  and  five  un- 
stained thin  sections  cut  from  chips,  parts 
of  which  were  used  to  obtain  the  magnetic 
concentrate.  Column  C  is  the  average  of 
modes  made  on  five  microsamples  of  the 
ground,  sized,  and  purified  concentrate 
actually  used  for  the  chemical  analysis. 

The  purification  procedure  is  consider- 
ably less  extreme,  particularly  as  regards 
washing  and  desliming,  than  that  often 
resorted  to  in  sample  preparation  of  this 
type.  It  has  nevertheless  led  to  a  shift  in 
the  ratio  of  nepheline  to  feldspar  so  ex- 
treme that  comparison  of  chemical  and 
modal  results  would  have  been  unintelligi- 
ble or  misleading  if  carried  through  di- 
rectly from  bulk  analysis  to  thin-section 

Materials  reprecipitated  from  the  acid 
solution  amounted  to  20.1  per  cent  of  the 
initial  sample,  and  from  SiC>2,  Na20,  and 
K20  entries  in  table  5  the  calculated 
weight  per  cent  of  the  acid-soluble  portion 
is  18.8  per  cent.  The  sharp  discrepancy 
between  either  of  these  values  and  the 
nepheline  entry  (or  the  sum  nepheline  + 
acmite)  in  column  A  of  table  6  led  to  the 
analyses  that  yielded  column  B.  The 
agreement  between  these  two  columns 
stimulated  much  speculation  about  the 
"actual"  composition  of  the  minerals,  spec- 

TABLE  6.     Average  Modes  of  a  Pseudoleucite 
from  Elk  Peak,  Bearpaw  Mountains,  Montana 















j  „ 




A.  Mean  of  112 


in  6  stai 

ned  thin 


B.  Mean  of  111 

aggregates  in  5  unsta: 

ined  thin 


C.  Mean  of  5  microsamples 

from  final  powder 

used  for  chemical  analysis. 

ulation  that  was  brought  to  an  abrupt  halt 
by  the  analysis  given  in  column  C,  which 
shows  that  the  modal  content  of  nepheline 
has  indeed  been  drastically  reduced  by  the 
sample  preparation.  We  conclude  that  the 
nepheline  and  sanidine  of  the  pseudoleu- 
cite have  essentially  the  compositions 
shown  in  columns  2  and  3  of  table  5,  but 
that  the  amounts  of  these  two  minerals 
actually  present  in  the  pseudoleucite  are 
better  estimated  by  columns  A  and  B  of 
table  6.  It  is  interesting  to  note  that  a 
nepheline  rather  rich  in  K2O  but  in  al- 
most exact  balance  with  regard  to  Si02 
occurs,  in  intimate  intergrowth,  and  ap- 
parently in  equilibrium,  with  a  sanidine 
unusually  poor  in  NasO.  A  coarse-grained 
nepheline  syenite  consisting  largely  of 
nepheline  and  feldspar  with  compositions 
like  those  in  our  pseudoleucite  has  been 
described  from  Assynt  by  Tilley.  He  has 
also   described   a   Brazilian   pseudoleucite 


which  must  be  very  similar,  but  ours  ap- 
pears to  be  the  first  of  which  actual  analy- 
ses are  available  for  both  phases. 

The  details  of  this  study  will  be  pub- 
lished separately.  In  much  petrographic 
work  of  this  general  type  the  only  quanti- 
tative data  would  be  the  bulk  analysis  of 
the  concentrate.  More  rarely,  the  thin- 
section  mode  would  also  be  available.  The 
bulk  analysis  alone  would  lead  to  one 
description.  The  combination  of  thin- 
section  mode  and  bulk  analysis  would  lead 
to  a  very  different  one,  and  from  this 
combination  the  petrologist  might  be  en- 
couraged to  suppose  that  the  nepheline 
observed  under  the  microscope  was  ex- 
traordinarily siliceous.  The  analysis  of  the 
acid-soluble  fraction  dispels  this  fantasy 
but  leaves  unresolved  a  glaring  discrep- 
ancy between  petrographic  and  chemical 
estimates  of  the  mode.  The  discrepancy 
might  be  explained  by  supposing  that  the 
thin-section  modes  are  absurdly  faulty, 
but  comparison  of  column  C  with  column 
A  or  B  of  table  6  indicates  that  the  real 
difficulty  is  in  sample  preparation. 

P.  M.  Orville 

As  part  of  a  continuing  program  of 
applying  laboratory  data  to  petrologic 
problems,  a  study  is  being  made  of  co- 
existing plagioclase  and  alkali  feldspars 
from  pegmatites  in  the  southern  Black 
Hills,  South  Dakota;  the  Spruce  Pine  dis- 
trict, North  Carolina;  and  New  Hamp- 
shire. The  perthitic  texture  of  most  alkali 
feldspars  from  pegmatites  offers  convinc- 
ing evidence  that  movement  of  ions  within 
the  crystal  did  not  cease  at  the  moment  of 
final  crystallization.  The  scale  of  perthite 
exsolution  textures  suggests  that  the  limits 
within  which  unmixing  of  feldspar  phases 
and  migration  of  alkali  ions  have  taken 
place  is  of  the  order  of  a  few  millimeters. 
On  this  basis  it  seems  reasonable  to  as- 
sume that  the  central  portion  of  a  perthite 
crystal  many  inches  in  size  has  been  af- 
fected only  by  an  unmixing  process  taking 

place  within  its  own  volume  and  has 
neither  gained  nor  lost  material  with  re- 
spect to  its  surroundings. 

If  the  assumptions  can  be  made  that 
contiguous  plagioclase  and  perthite  crys- 
tals have  crystallized  in  equilibrium  with 
one  another  and  that  their  compositions 
have  not  changed  subsequent  to  the  time 
of  crystallization  as  single  phases,  then  the 
bulk  compositions  of  the  two  feldspars  can 
be  considered  to  represent  the  two  end 
points  of  a  tie  line  at  the  crystallization 
temperature  within  the  solidus  region  of 
the  ternary  feldspar  system.  The  orienta- 
tion of  the  tie  lines  and  the  position  of  the 
end  points  will  change  as  functions  of 
temperature  and,  to  a  lesser  extent,  pres- 
sure. Previous  reports  (Year  Book  56, 
p.  206;  55,  p.  190)  have  indicated  the  prog- 
ress made  in  determining  phase  relations 
within  the  synthetic  ternary  feldspar  sys- 
tem. More  data  are  needed  on  the  posi- 
tion of  the  tie  lines  in  the  subsolidus  re- 
gion before  the  composition  of  coexisting 
natural  feldspars  can  be  used  as  a  geo- 
thermometer,  but  it  should  be  possible,  on 
the  basis  of  data  now  at  hand,  to  obtain 
an  idea  of  the  relative  changes  in  tempera- 
ture during  the  formation  of  successive 
zones  of  a  zoned  pegmatite  at  an  assumed 
constant  pressure. 

The  composition  of  ten  pairs  of  con- 
tiguous plagioclase  and  microline  perthite 
crystals  from  pegmatites  in  the  Spruce 
Pine  district,  North  Carolina,  has  been 
plotted  in  figure  20.  Ab  and  Or  contents 
are  based  on  flame  photometer  determina- 
tions of  Na  and  K.  The  An  content  of 
the  contiguous  plagioclase  was  determined 
from  the  refractive  index  of  a  fused  sam- 
ple. The  bulk  An  content  of  perthite  is 
based  on  semiquantitative  determinations 
of  Ca  by  the  emission  spectrograph.  The 
orientations  of  the  tie  lines  and  the  posi- 
tion of  the  end  points  are  consistent  with 
data  for  the  ternary  feldspar  system  and 
indicate  final  crystallization  at  relatively 
low  temperature  or  high  pressure. 

The    natural    perthites,   for   most   pur- 


poses,  can  be  considered  a  ternary  system 
in  which  Ab  and  Or  are  major  constitu- 
ents and  An  a  minor  constituent,  generally 
present  in  amounts  less  than  2  per  cent  by 
weight.  Ba  and  Rb  may  also  be  present 
as  minor  constituents.  There  is  no  single 
technique  except  chemical  analysis  by 
which  the  ternary  components  can  be  di- 
rectly determined  with  adequate  precision. 
Flame    photometer    determination    of    K 

X  is  the  angular  distance  in  degrees  and 
Y  is  the  composition  in  terms  of  weight 
per  cent  Or.  The  corresponding  equation 
for  Fe  Ka  radiation  is  Y  =  167.21  -73.69X. 
This  curve  has  an  advantage  over  deter- 
minative curves  previously  proposed  in 
that  a  single  internal  standard  which  is 
easily  obtainable  in  pure  form  has  been 
used  over  the  entire  range,  and  composi- 
tion   of   the    feldspar   is    plotted    directly 


Weight    per  cent 

Fig.  20.    Plot  of  contiguous  feldspar  pairs   (plagioclase  and  microcline  perthite)  from  pegmatites 
in  Spruce  Pine  district,  North  Carolina. 

and  Na  permits  an  estimate  of  composi- 
tion in  terms  of  Ab  and  Or  content,  and 
the  (201)  X-ray  technique  of  Bowen  and 
Tuttle  gives  a  composition  in  terms  of 
the  Or  content. 

Figure  21  is  a  newly  established  de- 
terminative curve  for  the  synthetic  alkali 
feldspar  series  in  which  the  difference  in 
20  for  Cu  Ka  radiation  between  the  (201) 
peak  of  alkali  feldspar  and  the  (101)  peak 
of  KBrOs  is  plotted  directly  against  the 
composition  in  terms  of  weight  per  cent 
Or.  The  curve  is  a  straight  line,  the  equa- 
tion of  which  is  Y=  166.39  -92.31X,  where 

against  A 20  values.  In  the  original  X-ray 
determinative  curve  presented  by  Bowen 
and  Tuttle,  composition  was  plotted 
against  spacings  of  the  (201)  plane  in  the 
feldspar  crystals.  An  olivine  powder  with 
a  peak  at  20  of  approximately  23°  was 
taken  as  an  internal  standard,  and  the 
absolute  position  of  the  olivine  peak  was 
determined  by  calibration  against  a  quartz 
peak  having  a  20  value  of  20.85°.  The 
curve  of  Chayes  and  Robbins  (Year  Book 
53,  p.  135)  was  determined  only  for  the 
Or-rich  end  of  the  series  between  Orioo 
and  Or6o.  The  internal  standard  for  this 



curve  was   an  Amelia   albite  with    (201) 
peak  at  20  =  22.06°. 

KBr03  seems  well  suited  for  use  as  an 
internal  standard.  It  is  a  standard  chemi- 
cal reagent  and  obtainable  in  very  pure 
form.  The  position  of  the  (101)  peak 
(20  =  20.205°  ±0.010°)  does  not  differ  ap- 
preciably between  different  lots  of  reagent 
grade  material.  The  salt  is  only  slightly 
soluble  in  water  and  is  much  less  hygro- 
scopic than  most  alkali  halide  salts.  Smear 
mounts  prepared  with  KBrOs  as  an  in- 

preciably  from  Ab  to  An  within  the  syn- 
thetic plagioclase  series.  Therefore,  it  ap- 
pears likely  that,  within  the  ternary  feld- 
spar system,  the  position  of  the  (201)  peak 
is  primarily  a  function  of  the  weight  per 
cent  Or. 

The  curve  of  figure  21  can  be  used  di- 
rectly for  estimating  the  Or  content  of 
homogeneous  natural  sanidines  or  of  any 
natural  alkali  feldspar  that  has  been  ho- 
mogenized and  inverted  to  the  high-tem- 
perature monoclinic  form.   This  has  been 


T 1 

"I —           i           — r    -     —  i 

l —          I 



"       ^V 















cm  1.20 





1 .10 












i                i               i               i 

40  50  60 

Weight    per  cent  Or 3 

Fig.  21.     Difference  in  20  between  (201)  peak    of  synthetic  alkali  feldspar  and   (101)   peak  of 
KBr03  for  Cu  Ka  radiation  plotted  against  composition. 

ternal  standard  have  been  exposed  to  the 
atmosphere  for  periods  of  3  months  or 
more  without  noticeable  change  in  the 
position  or  intensity  of  the  KBrOs  peaks. 
The  position  of  the  (101)  peak  is  not 
appreciably  affected  by  a  variation  in  room 
temperature  of  20°  C. 

The  synthetic  alkali  feldspars  on  which 
the  curve  is  based  were  crystallized  from 
glasses  prepared  by  J.  F.  Schairer.  Crystal- 
lization was  carried  out  hydrothermally 
in  sealed  platinum  tubes  at  a  temperature 
of  800°  C  and  an  H20  pressure  of  1000 
bars  for  periods  of  5  to  7  days.  A  longer 
crystallization  period  does  not  produce 
significant  changes  in  the  (201)  spacing. 
The   (201)    spacing  does  not  change  ap- 

done  for  a  number  of  coarsely  perthitic 
microclines  by  heating  the  finely  ground 
material  in  sealed  platinum  tubes  at 
800°  C  with  1000  bars  H20  pressure  for 
25  days. 

In  table  7  a  comparison  is  made  between 
bulk  Or  compositions  determined  by  the 
(201)  X-ray  method  and  by  the  flame 
photometer.  The  samples  are  microcline 
perthites  from  pegmatites,  the  first  7  from 
Spruce  Pine,  North  Carolina,  the  last  7 
from  the  Black  Hills,  South  Dakota. 

The  possibility  that  this  curve  could 
also  be  used  to  estimate  the  composition 
of  triclinic  alkali  feldspars  has  been  in- 
vestigated. Dry  heat  treatment  of  most 
perthitic    microcline    at    1050°    C    for   48 



hours  results  in  complete  homogenization 
of  the  feldspar  without  inverting  it  to 
monoclinic  form,  although  the  triclinicity 
as  measured  by  the  difference  between  the 
(131)  and  (131)  peaks  decreases  some- 
what. (Many  laboratories  lacking  hydro- 
thermal  apparatus  have  high-temperature 
ovens  that  could  be  used  for  the  dry  ho- 
mogenization of  perthites.)  Preliminary 
results  based  on  X-ray  determination  of 
28  microcline  perthites  from  South  Da- 
kota and  North  Carolina  in  the  composi- 
tion ranges  70  to  96  weight  per  cent  Or 

TABLE  7.    Weight  Per  Cent  Or 
















































indicate  that  samples  homogenized  in  the 
dry  way  at  1050°  C  give  Or  percentages 
slightly  lower  than  the  same  samples 
homogenized  and  inverted  hydrothermally 
at  800°  C.  The  average  difference  between 
the  two  values  of  A  20  for  the  triclinic  and 
monoclinic  form  of  each  sample  is  0.015°, 
which  is  equivalent  to  1.5  weight  per 
cent  Or. 


Spinels  constitute  one  of  the  most  inter- 
esting groups  of  minerals.  They  show 
widespread  occurrences  as  accessory  con- 
stituents and  a  few  occurrences  of  great 
economic  importance.  Synthetic  forms  are 
widely  used  for  many  industrial  purposes. 
All  spinels  consist  essentially  of  a  single 
structural   framework   of   oxygens   which 

accommodates  a  great  variety  of  bi-  and 
trivalent  cations  with  the  general  formula 
R+-R2+304.  Geologically  the  most  impor- 
tant iron-bearing  members  are  magnetite, 
FeFe204;  hercynite,  FeALO*;  and  Fe- 
chromite,  FeCr^O^ 

Buddington  and  co-workers  have  demon- 
strated the  geologic  significance  of  solid 
solution  and  exsolution  phenomena  in 
natural  ilmeno-magnetites.  For  experi- 
mental reasons,  laboratory  investigations 
of  the  phase  relations  in  the  spinel  group 
have  been  restricted  to  high  temperatures 
and  atmospheric  pressure.  Yet  elucidation 
of  the  subsolidus  behavior  of  iron-rich 
spinels  at  moderate  temperatures  is  highly 
desirable  since  members  of  the  magnetite 
group  belong  to  the  most  common  of  ac- 
cessory minerals.  Hence  the  important 
magnetite-hercynite  system  (FeO*Fe203- 
FeO'ALOs)  was  chosen  for  investigation. 

Magnetite-Hercynite  Relations 
A.  C.  Turnoc\  and  H.  P.  Eugster 

Atlas  and  Sumida  (1958)  report  com- 
plete solid  solution  above  1000°  C.  Yet 
hercynite  is  known  to  have  exsolved  from 
many  magnetites  formed  at  high  tempera- 
tures, often  in  conjunction  with  exsolution 
of  ilmenite.  It  was  decided,  therefore,  to 
investigate  the  phase  relations  below 
1000°  C.  The  partial  pressure  of  oxygen 
was  controlled  by  using  a  quartz  +  mag- 
netite+fayalite  buffer  (see  Eugster,  1957). 
The  Po2-T  curve  for  this  assemblage  lies 
at  all  temperatures  within  the  field  of 
stability  of  magnetite.  Complete  solid 
solubility  was  found  above  870°  ±20°  C  at 
a  total  pressure  of  2000  bars.  The  cell  di- 
mensions of  the  series,  synthesized  at 
800°  C  and  2000  bars  (900°  C,  1000  bars 
for  Mt  50  He  50),  decrease  from  magnetite 
d  =  8.393  ±0.002  A.U.  to  hercynite  8.149± 
0.003.  The  curve  interplanar  spacing  vs. 
composition  is  not  a  straight  line  but 
shows  a  slightly  convex  upward  curvature 
on  a  mole  per  cent  scale,  according  to 
d = 8.393  -  0.00194X  -  0.5  X  10-5X2,  where 
X  is  mole  per  cent  hercynite.  This  agrees 


well  with  the  results  of  Atlas  and  Sumida 
(1958)  from  experiments  done  at  low  pres- 
sures in  a  helium  atmosphere. 

Preliminary  work  indicates  that  at  2000 
bars  the  top  of  the  solvus  lies  at  870°  ± 
20°  C,  and  that  the  solvus  itself  is  bounded 
by  the  following  two-phase  assemblages 
(in  weight  per  cent)  : 

800°  C  Mt  76  He  24  — Mt  25  He  75 

700°  Mt  85  He  15  — Mt  14  He  86 

600°  Mt  89  He  11— Mt  10  He  90 

500°  Mt  92  He    8  —  Mt    8  He  92 

It  appears  that  hercynite  is  not  stable  at 
higher  partial  pressures  of  oxygen,  such  as 
that  of  a  magnetite-hematite  buffer.  The 
hercynite  composition  is  represented  by 
the  assemblage  magnetitessH-corundumss. 
This  work  will  be  useful  in  the  interpre- 
tation of  rocks  containing  magnetites 
formed  in  an  aluminous  environment, 
such  as  magnetites  coexisting  with  hercy- 
nite, corundum,  and  sillimanite. 

Chester  TLinery  Deposits 
D.  R.   Wones 

The  emery  deposits  at  Chester,  Massa- 
chusetts, have  been  a  famous  mineral- 
collecting  locality  for  many  years.  In  spite 
of  the  widespread  interest  in  the  mineral- 
ogy of  the  area,  intensive  geological  and 
petrological  studies  have  not  been  made. 
The  current  efforts  have  been  undertaken 
to  outline  the  geologic  environment  and 
major  mineral  assemblages  of  these  de- 

The  emery  deposits  occur  as  veinlets  of 
corundum-magnetite-hematite  rock  with- 
in a  quartz-free  mica-chlorite  schist.  The 
schist  contains  accessory  tourmaline  and 
magnetite  with  corundum  and  zoisite. 
The  chlorite  is  the  variety  amesite,  as  de- 
termined by  X-ray  and  optical  analysis. 
The  mica  is  either  paragonite  or  mus- 
covite,  depending  on  location,  but  no  rock 
has  yet  been  found  containing  both  min- 
erals. Cross-cutting  the  schist  are  veins  of 
margarite,  chlorite,  and  diaspore. 

The  emery-bearing  schist  is  wholly  con- 
tained   within    the    Chester    amphibolite, 

near  the  east  border  of  the  amphibolite 
body.  The  amphibolite  varies  from  an 
epidote-amphibole  rock  to  an  epidote- 
oligoclase-amphibole  rock.  In  places  the 
amphibolite  and  emery  are  transected  by 
a  talc-serpentine  rock. 

Metasomatic  origins  have  been  proposed 
for  emery  deposits.  X-ray  and  optical 
properties  of  the  minerals,  many  of  which 
are  members  of  solid  solutions,  do  not  vary 
throughout  the  Chester  deposits,  indicat- 
ing homogeneous  compositions.  There- 
fore, either  metasomatic  activity  was  not  a 
major  factor  or  compositional  differences 
have  been  erased  by  subsequent  meta- 


/.  F.  Schairer  and  H.  S.  Yoder,  Jr. 

Two  years  ago  we  began  a  study  of  this 
important  quaternary  system  for  which 
phase-equilibrium  data  were  almost  com- 
pletely lacking.  Substantial  progress  was 
made  during  the  first  year  of  study,  and 
last  year  (Year  Book  56,  pp.  217-222,  figs. 
50-53)  we  presented  phase-equilibrium 
diagrams  for  the  joins  albite-cordierite- 
silica,  albite-forsterite-cordierite,  and  al- 
bite-magnesium  metasilicate-cordierite  and 
a  diagram  showing  the  relation  of  univari- 
ant  lines  to  ternary  invariant  points  in 
limiting  systems  and  to  seven  of  the  qua- 
ternary invariant  points.  We  showed  that 
in  the  two  large  volumes  albite-corun- 
dum-spinel-silica  and  albite-forsterite- 
spinel-silica  the  residual  liquid  during 
crystallization  proceeds  toward  a  similar 
goal,  a  soda-granite. 

During  the  past  year  we  have  vigorously 
pursued  the  study  of  the  quaternary  sys- 
tem, expanding  the  portion  under  investi- 
gation to  include  compositions  rich  in 
nepheline  which  lie  in  the  volume  nephe- 
line-forsterite-spinel-silica.  At  this  time 
we  present  the  phase-equilibrium  diagram 
for  the  system  nepheline-spinel-silica  as 
figure  22.  Open  circles  represent  the  com- 
positions of  compounds,  and  black   dots 


the  compositions  studied.  This  is  a  ternary 
system  within  the  quaternary  system 
Na20-MgO-Al203-Si02  except  for  those 
compositions  that  crystallize  mullite 
(3AI2O3 '  2Si'Oa)  at  some  temperature  dur- 
ing their  crystallization.  Even  these  be- 
come completely  ternary  in  their  behavior 

angle  albite-spinel-cordierite  become  com- 
pletely crystalline  at  the  temperature  of 
the  ternary  reaction  point  K,  and  the  last 
liquid  has  the  composition  K;  those  in  the 
triangle  albite-cordierite-silica  become 
completely  crystalline  at  the  temperature 
of  /,  and  the  last  liquid  has  the  composi- 

SPINEL  2135+20' 


A  =  /448±5° 
B  =  /470*/0° 
C  =  /438t5a 
H  =  /062t3<- 
I  =/045S/0° 
J  =  1470+10° 
K  =  t09S±S" 
L  =  /055t5" 
M = 1068 t5° 
N  =  1280+° 
0  =  l280tB 
P=A80L/r  1490 




40  L\  RITF    50  60      H 

///S5-  Na2QAI203.6Si02 


Fig.  22.   Phase-equilibrium  diagram  of  the  system  nepheline-spinel-silica. 


at  lower  temperatures   when  mullite  has 
disappeared  by  reaction  with  liquid. 

The  portion  albite-cordierite-silica  was 
shown  in  more  detail  in  last  year's  report 
(Year  Book  56,  p.  219,  fig.  50),  and  binary 
and  ternary  points  are  similarly  lettered 
in  figure  22.  All  compositions  in  the  tri- 
angle nepheline-spinel-albite  become  com- 
pletely crystalline  at  the  temperature  of 
the  ternary  eutectic  L,  and  the  last  liquid 
has  the  composition  L;  those  in  the  tri- 

tion  /.  Thus  we  see  from  the  positions  of 
L,  K,  and  /,  which  are  near  the  side  line 
nepheline-silica,  that  residual  liquids  from 
crystallization  are  poor  in  spinel  or  cordi- 
erite  or  both  and  rich  in  the  alkali  alu- 
minosilicates  albite  or  nepheline  or  both. 
We  also  note  that  those  compositions  on 
the  silica  side  of  albite-spinel  that  also  lie 
in  the  triangle  albite-spinel-cordierite  have 
the  residual  liquid  at  K,  which  lies  on  the 
nepheline  side  of  albite-spinel.   If  the  re- 


sidual  liquid  were  separated  from  early- 
formed  crystals,  it  would  give  a  nepheline- 
bearing  product  of  crystallization. 


/.  F.  Schairer  and  N.  Morimoto 

The  mutual  melting  relations  among 
these  three  important  groups  of  minerals 
are  of  primary  importance  in  unraveling 
the  crystallization  of  many  igneous  rocks, 
particularly  the  basalts.    Last  year  Yoder 

olivines,  pyroxenes,  and  feldspars.  The 
groundwork  for  this  study  was  laid  when 
Bo  wen  (1914)  studied  the  system  forster- 
ite-diopside-silica  at  temperatures  where  a 
liquid  phase  was  present.  This  system  de- 
picts the  relations  of  pyroxenes  between 
enstatite  (MgO*Si02)  and  diopside 
(CaO'MgO*2Si02)  to  the  magnesian  oli- 
vine forsterite  (2MgO*Si02)  and  to  silica. 
Recently  Boyd  and  Schairer  (Year  Book 
56,  pp.  223-225)  expanded  this  study  of 
Bowen's  to   include   the   subsolidus   rela- 





Fig.  23.    Diagram  showing  the  relations  between  the  olivine  forsterite,  pyroxenes  between  ensta- 
tite and  diopside,  the  soda  feldspar  albite,  and  silica. 

(Year  Book  56,  pp.  159-160)  discussed 
crystallization  in  some  of  the  systems  in- 
volved in  the  origin  of  basalts.  Because  of 
their  very  complexity  both  in  crystalline 
modifications  and  in  chemical  composition 
(see  Year  Book  56,  pp.  208-210)  the  rock- 
forming  pyroxenes  and  pyroxenoids  can 
yield  much  information  on  temperatures 
and  the  crystallization  processes  in  rock 

During  the  past  year  we  have  had  the 
opportunity  to  begin  an  important  in- 
vestigation of  the  melting  relations  in 
the  system  forsterite-diopside-silica-albite 
which  involves  the  three  mineral  groups 

tions  in  compositions  between  MgO*Si02 
and  CaOMg02Si02. 

The  relation  of  the  ternary  system  for- 
sterite-diopside-silica  to  the  system  forster- 
ite-diopside-silica-albite  is  shown  in  figure 
23.  The  ternary  system  just  named  is 
shown  as  an  equilateral  triangle  which  is 
the  base  of  a  regular  (equilateral)  tetra- 
hedron with  the  composition  albite  as  its 
apex.  The  three  faces  of  the  tetrahedron 
between  the  base  and  apex  have  been  laid 
flat  in  the  plane  of  the  base. 

In  addition  to  forsterite-diopside-silica, 
three  other  systems  between  this  base  and 



the  apex  albite  limit  the  system  forsterite- 
diopside-silica-albite : 

1.  One  of  these,  forsterite-albite-silica, 
has  been  studied  by  Greig  (whose  results 
have  not  yet  been  published),  and  last 
year  Schairer  (Year  Book  56,  pp.  217-222) 
gave  data  for  the  limiting  binary  system 
forsterite-albite  and  data  for  the  tie  line 

roxene,  and  soda-rich  plagioclase  are 
shown.  Attention  is  drawn  to  the  initial 
rise  in  liquidus  temperature  from  pure 
albite  toward  diopside,  and  then  a  falling 
of  liquidus  temperature  toward  the  point 
where  sodic  plagioclase  and  diopside  co- 
exist with  liquid  in  the  side  line  albite- 
diopside,  which  is  not  truly  binary.   The 

ALBITE   li/8±3° 


i/38°     PLAGIOCLASE 



40  SO  /600" 


M48°/4,2°\  J389°  D 1 0  PS  I DE 


Fig.  24.    Preliminary  phase-equilibrium  diagram  of  the  system  forsterite-albite-diopside. 

albite-magnesium  metasilicate  as  a  part  of 
his  data  in  the  system  Na20-MgO-Al203- 

2.  Diopside-albite-silica  is  a  part  of  the 
join  nepheline-diopside-silica  that  has 
been  studied  by  Schairer,  but  the  data 
have  not  yet  been  published. 

3.  During  the  past  year  we  have  made  a 
study  of  the  system  forsterite-albite-diop- 
side, the  preliminary  phase-equilibrium 
diagram  for  which  is  given  here  as  figure 
24.   The  fields  of  forsterite,  diopsidic  py- 

diopside  crystals  may  not  be  pure  diopside 
but  may  have  a  small  Al2Os  content. 

The  triangular  join  MgSiOs-diopside- 
albite  (not  shown  in  fig.  24)  is  now  under 
intensive  investigation.  So  far  we  have 
prepared  forty-six  separate  compositions, 
and  these  are  being  subjected  to  thermal, 
optical,  and  X-ray  studies.  Many  of  the 
melts  in  this  important  join  have  the 
olivine  forsterite  with  two  pyroxenes  in 
equilibrium  with  liquids  at  appropriate 


P.  H.  Abelson 

Attempts  to  extract  organic  chemicals 
from  Precambrian  rocks  result  in  disap- 
pointingly low  or  negligible  yields.  Shales 
and  slates  of  a  low  grade  of  metamor- 
phism,  such  as  the  Keeweenawan  None- 
such shale  and  the  Huronian  Rove  slate, 
have  been  examined. 

After  fine  grinding  in  ball  mills  for 
3  days,  the  samples  were  extracted  with  an 
ethyl  alcohol-benzene  mixture  for  a  week 
in  a  Soxhlet  apparatus.  Only  a  small 
trace  of  soluble  organic  substance  was 
found  in  the  Nonesuch  sample  and  noth- 
ing in  the  Rove  slate. 

It  is  possible  that  other  Precambrian 
shales  may  be  found  that  do  contain  ex- 
tractable  substances.  Indeed,  Swain  has 
reported  finding  small  amounts  of  organic 
compounds  in  some  old  rocks,  including  a 
granite.  This  latter  occurrence  is  certainly 

The  usual  methods  rarely  extract  more 
than  a  small  fraction  of  the  total  organic 
matter  present  in  sedimentary  rocks. 
Therefore,  in  approaching  so  difficult  a 
problem  as  the  Precambrian  sediments  it 
seemed  desirable  to  develop  improved  and 
essentially  nondestructive  methods  of  iso- 
lating the  organic  compounds  present. 

One  possible  method  is  hydrogenation. 
In  general,  this  process  does  not  break 
carbon-carbon  bonds  and  may  be  carried 
out  at  temperatures  below  those  causing 
excessive  thermal  degradation.  It  probably 

breaks  sulfur  cross-linking  between  hydro- 
carbon chains  and  converts  unsaturated 
aromatic  compounds  into  more  easily  ex- 
tracted naphthenes. 

Exploratory  experiments  have  been  car- 
ried out  using  Kolm  shale,  Sweden,  of 
Cambrian  age;  the  Vanini  shale,  Nevada, 
of  Ordovician  age;  and  the  Green  River 
shale,  Colorado,  of  Eocene  age.  The 
shales  were  ground  in  a  ball  mill  for  3 
days,  mixed  with  M0S2,  serving  as  a  cata- 
lyst, and  hydrogenated  at  375°  C  for  5 
hours  at  4000  psi  H2  pressure.  Spuriously 
high  yields  were  obtained  when  a  suspend- 
ing medium  such  as  tetralin  was  em- 
ployed, and  most  runs  were  therefore 
made  without  any  added  organic  solvent. 

The  solids  were  extracted  in  Soxhlet 
equipment  for  a  week.  Results  showed 
that  hydrogenation  increased  very  substan- 
tially the  yields  from  these  processed 
shales,  values  of  70  per  cent  of  the  total 
organic  matter  present  in  the  Kolm  being 
obtained.  In  comparison,  an  aliquot  of 
the  original  ground  Kolm,  extracted  with- 
out heating,  yielded  1.6  per  cent,  and  an- 
other sample,  extracted  after  heating  at 
375°  C  for  5  hours  in  a  nitrogen  atmos- 
phere, gave  only  an  8  per  cent  yield.  The 
hydrogenation  procedure  was  thus  very 
effective  in  rendering  organic  matter  more 
extractable,  and  the  approach  seems  suffi- 
ciently attractive  to  merit  further  investi- 
gation of  its  applicability  to  studies  of  or- 
ganic sediments  of  all  ages. 


During  this  past  year  systematic  labora- 
tory studies  of  the  relations  among  the 
more  common  sulfide-type  minerals  have 
added  greatly  to  our  understanding  of 
mineral  associations  found  in  nature. 

Investigations  of  the  subsolidus  relations 
in  the  Cu2S-S  part  of  the  Cu-S  system 
and  of  the  FeS-S  part  of  the  Fe-S  system 
have  been  completed.  The  results  have 
provided  the  necessary  basic  information 

for  systematic  exploration  in  a  number 
of  ternary  systems  involving  these  ele- 
ments. The  solidus  relations  in  the  system 
Cu-Fe-S  have  been  determined  between 
400°  and  800°  C.  This  knowledge,  in 
combination  with  the  data  we  have  ac- 
quired in  the  Cu-S  and  Fe-S  systems,  is 
crucial  to  the  work  now  under  way  on  the 
sulfur-deficient  regions  of  the  Cu-Fe-S 


New  studies  have  included  the  more 
sulfur-rich  parts  of  the  Fe-Zn-S  system. 
Thus  a  number  of  points  have  been  de- 
termined on  the  curve  relating  tempera- 
ture to  composition  of  sphalerite  formed 
in  equilibrium  with  pyrite  in  the  presence 
of  liquid  and  vapor. 

A  detailed  investigation  of  the  phase 
relations  in  the  Fe-As-S  system  is  nearly 
completed.  The  upper  stability  curve  of 
the  only  known  ternary  phase,  arsenopy- 
rite,  in  this  system  has  been  determined. 
The  studies  of  iron,  nickel,  and  cobalt 
arsenides,  on  which  a  preliminary  report 
was  given  last  year  (Year  Book  56),  have 
been  completed. 

Pyrrhotite  formed  in  equilibrium  with 
pyrite  varies  significantly  in  composition 
with  temperature.  The  experimentally  de- 
termined relationship  of  composition  to 
temperature  has  been  applied  to  a  number 
of  ore  deposits,  and  temperatures  of  forma- 
tion of  pyrrhotite-pyrite  assemblages  have 
been  estimated.  In  one  locality  where 
both  the  sphalerite-pyrrhotite  and  the  pyr- 
rhotite-pyrite assemblages  could  be  used 
as  temperature  indicators,  the  two  meth- 
ods gave  substantially  identical  results. 

Estimates  of  the  temperature  of  the 
formation  of  certain  sulfide  veins  and  ore 
bodies  by  the  sphalerite-pyrrhotite  method 
have  indicated  the  existence  of  strong 
temperature  gradients  during  formation 
of  the  deposits,  implying  that  such  ore 
solutions  cooled  rapidly  when  moving 
away  from  their  source,  and  that  they  may 
have  been  more  concentrated  than  is 
usually  thought. 

Equipment  for  systematic  studies  of  the 
solubilities  of  various  ore-forming  sulfides 
in  aqueous  solutions  at  pressures  up  to 
1500  psi  and  at  temperatures  up  to  200°  C 
has  been  designed  and  constructed,  and  is 
now  in  routine  operation.  Experimental 
results  on  the  solubility  of  sphalerite  in 
H2S-saturated  water  at  various  tempera- 
tures and  pressures  indicate  solubilities  six 
or  seven  orders  of  magnitude  higher  than 
those  found  for  ZnS  in  pure  water. 

G.  Kullerud 

The  compounds  in  this  system  are  chal- 
cocite  (Q12S),  digenite  (Cu9S5),  and 
covellite  (CuS).  Of  these  minerals  chal- 
cocite  is  most  often  encountered  in  ore 
deposits.  Covellite  is  less  important  as  an 
ore  mineral  than  chalcocite  but  is  never- 
theless a  common  mineral  in  many  copper 
ores.  The  third  mineral,  digenite,  greatly 
resembles  chalcocite  in  polished  sections 
and  therefore  has  often  been  incorrectly 
identified.  During  the  last  year  X-ray 
diffraction  studies  of  ore  specimens  from 
various  copper  sulfide  deposits  have  shown 
that  digenite,  which  possesses  a  very  char- 
acteristic X-ray  diffraction  pattern,  is  a 
much  more  common  mineral  than  earlier 
investigations  employing  polished  sections 
had  indicated. 

Chalcocite,  digenite,  and  covellite  may 
be  readily  synthesized  in  the  dry  way  in 
silica  tubes  by  mixing  copper  with  ap- 
propriate amounts  of  sulfur  and  heating 
the  mixtures.  The  synthesis  of  covellite 
was  described  in  last  year's  report,  and  the 
synthesis  of  digenite  was  described  in  a 
recent  paper  by  Donnay,  Donnay,  and 
Kullerud  (1958).  Chalcocite  forms  readily 
from  stoichiometric  mixtures  of  copper 
and  sulfur  even  at  low  temperatures.  At 
25°  C,  small  amounts  of  chalcocite  are 
formed  after  only  a  few  hours.  At  200°  C, 
the  reaction  between  copper  and  sulfur  is 
rapid;  after  1  day  no  unreacted  copper  or 
sulfur  can  be  seen  in  the  tubes.  At  500°  C 
and  higher  temperatures  the  reaction  is 
extremely  fast  and  is  completed  in  less 
than  1  minute. 

Figure  25  shows  the  relations  between 
chalcocite,  digenite,  and  covellite  in  the 
Cu-S  system.  In  the  diagram,  vapor  oc- 
curs with  every  phase  or  phase  assemblage, 
since  all  experiments  were  performed  in 
evacuated  and  sealed,  rigid  silica  glass 
tubes,  with  a  vapor  phase  always  present. 
It  is  seen  that  below  507°  C  covellite  is 
stable  in  the  presence  of  vapor  and  either 
digenite  or  a  sulfur-rich  liquid.  At  507°  ± 


3°  C  the  four  phases  digenite  +  covellite  + 
liquid  + vapor  are  all  stable,  and  the  vapor 
pressure  at  this  invariant  point  is  about 
900  mm  Hg. 

Numerous    experiments     were     under- 
taken  to   determine   whether   covellite   is 

The  contents  of  the  inner  tube  were 
ground  and  reheated  in  the  same  way 
until  no  further  change  in  weight  was  ob- 
served. (2)  CuS  with  small  amounts  of 
sulfur  was  heated  in  evacuated  tubes  hav- 
ing a  very  small  vapor  space.  This  method 







o,  700 









Liquid  +  Vapor 

Chalcocite  s.  s.  +  Liquid  +  Vapor 
Chalcocite  s.  s.  +  Vapor 

J-Chalcocite  s.s. 
^      +  Diqenite  +  Vapor 

Digenite  s.  s.  +  Liquid  +  Vapor 

Digenite  s.s.  +  Vapor 

Digenite  s.s.  +  Covellite 
+  Vapor 

L_l — JI I I I I I I I I I I l_ 

21    /22     23     24     25     26     27     28     29     30     31      32     33 

£S      CU9S5 




34     35     36 


Wt.  percent  S 

Fig.  25.    Relations  between  chalcocite,  digenite,  and  covellite  in  the  Cu-S  system. 

capable  of  taking  sulfur  or  digenite  into 
solid  solution.  The  maximum  sulfur  con- 
tent of  covellite  was  obtained  by  two 
methods.  (1)  A  weighed  amount  of  cop- 
per filings  was  heated  in  an  open  tube 
inside  a  larger  closed  tube  containing  ex- 
cess sulfur;  thus  liquid  and  vapor  were 
always  present.  The  amount  of  sulfur  that 
combined  with  the  copper  was  determined 
by  weighing  the  inner  tube  after  each  run. 

was  used  only  at  temperatures  below 
400°  C,  where  the  vapor  pressures  are  low. 

Minimum  sulfur  content  of  covellite 
was  obtained  by  mixing  small  amounts  of 
digenite  with  covellite  and  heating  the 
mixture  in  small  evacuated  silica  tubes. 

X-ray  diffraction  studies  of  phases  and 
measurements  to  detect  any  displacement 
of  reflections  caused  by  possible  composi- 
tional changes   of  covellite  indicate   that 


covellite  is  a  stoichiometric  compound. 
Studies  of  polished  sections  of  the  products 
verified  these  findings.  The  copper-to- 
sulfur  ratio  of  covellite  lies  within  the 
range  of  1±0.01. 

The  synthesis,  as  well  as  crystal  and 
twin  structure  of  digenite,  was  described 
by  Donnay,  Donnay,  and  Kullerud 
(1958).  Further  experiments  on  the  sta- 
bility of  this  compound  have  shown  that 
digenite,  which  decomposes  on  heating  to 
chalcocite  and  vapor,  is  stable  up  to  about 
925°  C  in  the  presence  of  excess  sulfur 
(liquid  +  vapor).  The  vapor  pressure  at 
this  invariant  point  where  the  four  phases 
chalcocite  +  digenite  +  liquid  +  vapor  all  are 
stable  has  not  been  measured.  However, 
it  must  be  less  than  about  65  atm,  which 
is  the  vapor  pressure  of  pure  sulfur  at 
925°  C  (West,  1950).  This  invariant  point 
is  the  origin  of  four  univariant  curves, 
CU2S  +  CU9S5  +  F,  CU9S5  +  L  +  F,  Cu2S  + 
CU9S5  +  L,  and  Cu2S  +  L  +  V,  as  shown 
schematically  in  figure  26.  In  the  three 
univariant  assemblages  Cu2S  +  Cu9S5  +  F, 
CU9S5  +  L+F,  and  Cu2S  +  L  +  V,  vapor 
exists  as  a  phase.  The  univariant  curves 
(1),  (2),  and  (4),  therefore,  can  be  de- 
termined by  experiments  employing  rigid 
silica  tubes,  provided  the  pressures  in  the 
tubes  can  be  measured.  The  fourth  uni- 
variant assemblage,  Cu2S  +  Cu9S5  +  L, 
does  not  involve  a  vapor  phase,  and,  there- 
fore, the  upper  stability  curve  (3)  in  fig- 
ure 26  can  be  determined  only  by  the  use 
of  collapsible  tubes  in  which  a  vapor  phase 
is  not  present.  Because  of  extensive  reac- 
tion between  copper  sulfide  and  gold  at 
the  temperatures  and  pressures  involved, 
this  curve  could  not  be  determined  by  the 
use  of  gold  tubing.  However,  it  is  likely 
that  the  upper  stability  curve  of  digenite, 
similar  to  the  stability  curves  of  covellite 
and  pyrite,  is  very  steep.  Thus,  even  under 
30,000  psi  of  sulfur  pressure,  the  break- 
down temperature  of  digenite  would  prob- 
ably not  exceed  950°  C. 

The  solubility  of  covellite  in  Cu9S5,  as 
well  as  that  of  chalcocite  in  digenite,  was 

studied  by  the  methods  discussed  above. 
It  was  found  that  a  digenite  containing 
about  22.3  weight  per  cent  sulfur  was 
stable  at  450°  C.  This  is  0.4  weight  per 
cent  in  excess  of  the  21.9  weight  per  cent 
sulfur  contained  in  stoichiometric  Cu9S5. 
The  solubility  of  chalcocite  in  digenite 
was  not  measurable  even  at  600°  C.  The 
solubility  of  digenite  in  chalcocite,  on  the 
other  hand,  is  readily  detected.  Thus,  at 
600°  C  a  chalcocite  containing  about  20.5 



J    1 

*io  1 

CO     1 


p>  1 



<->      1 

*       1 




CM    f 


5  J 


e>*       s                      \ 



Temperature  °C >■ 

Fig.  26.  Curves  showing  schematically  invari- 
ant point  where  the  four  phases  Cu2S  +  Cu9S5 
+  L  +  V  are  stable. 

weight  per  cent  sulfur  exists  in  equilib- 
rium with  digenite.  (Cu2S  contains  20.15 
weight  per  cent  S.)  Estimates  based  on 
numerous  runs  indicate  that  the  chalco- 
cite solid  solution  series  in  equilibrium 
with  digenite  at  the  invariant  point  (about 
925°  C)  may  extend  to  21.0  weight  per 
cent  sulfur. 

The  melting  relations  of  chalcocite  as 
shown  in  figure  25  are  those  determined 
by  Jensen  (1947).  The  dashed  liquidus 
curves  have  not  been  determined. 

Heating  experiments  were  performed 
on  covellite  and  digenite  in  closed  evacu- 
ated tubes  by  the  method  described  by 
Kracek  (1946).  Covellite  showed  no  phase 


changes  between  25°  and  490°  C,  where 
breakdown  occurred.  Runs  on  digenite 
indicated  an  inversion  at  65°  C,  but  no 
other  effects  of  heat  between  65°  and 
650°  C  were  noted. 

R.   G.  Arnold 

During  the  past  year  experimental  work 
on  the  pyrrhotite-pyrite  relations  has  been 
completed,  and  experiments  to  determine 
the  location  of  the  curve  representing  the 
composition  of  liquids  that  can  coexist 
with  pyrrhotite  and  vapor  have  been  ini- 
tiated. The  applicability  of  the  experi- 
mental ^(102) -value/composition  relation 
to  natural  pyrrhotites  has  been  investi- 
gated over  a  limited  composition  range. 
Temperatures  of  formation  of  a  number 
of  coexisting  pyrrhotite-pyrite  assemblages 
have  been  estimated. 

The  curve  representing  the  composition 
of  pyrrhotite  formed  in  equilibrium  with 
pyrite  at  the  pressure  of  the  coexisting 
vapor  has  been  extended  from  650°  C  to 
the  incongruent  melting  temperature  of 
pyrite  at  743°  C.  The  extended  portion  of 
this  curve  was  found  to  be  a  smooth  con- 
tinuation of  the  curve  previously  deter- 
mined up  to  650°  C  presented  in  last 
year's  report  (Year  Book  56,  p.  192,  fig. 
24).  Thus  the  earlier  evidence  for  a  pyr- 
rhotite inversion  at  670 ±5°  C  has  dis- 

To  determine  this  curve  above  650°  C 
the  procedure  was  as  follows.  Charges 
consisting  of  approximately  2.5  g  of  pel- 
leted sulfide  (pyrrhotite  and  pyrite)  of 
known  bulk  composition  were  fitted  snugly 
into  the  bottom  of  sealed  silica  glass  tubes. 
Silica  glass  rods  and  silica  glass  powder 
were  used  to  reduce  the  vapor  space  to  a 
minimum.  All  charges  were  heated  to  755° 
C  for  several  hours  to  decompose  any  py- 
rite present,  then  held  at  specified  temper- 
atures for  84  hours  or  longer.  Microscopic 
examination  of  the  products  permitted  dis- 
tinction of  stable  pyrite  from  exsolved  py- 
rite and  from  pyrite  formed  by  reaction 

between  pyrrhotite  and  vapor  during 
quenching.  Rapidly  exsolved  pyrite  was 
dispersed  throughout  the  pyrrhotite  par- 
ent in  grains  or  plates  about  6  u  or  less 
in  thickness.  Pyrite  rimming  the  pyrrho- 
tite grains  was  formed  by  the  reaction  of 
pyrrhotite  with  vapor  on  quenching  and 
occurred  as  grains  about  the  same  size  as 
exsolved  pyrite.  Stable  pyrite,  however, 
was  present  in  much  larger  grains  (—20 
u)  situated  at  or  near  the  borders  of  pyr- 
rhotite grains.  The  curve  representing  the 
composition  of  pyrrhotite  on  the  pyrrho- 
tite-pyrite solvus  was  bracketed  by  points 
representing  the  bulk  composition  of 
charges  in  which  stable  pyrite  was  present 
or  absent. 

The  initial  compositions  of  the  charges 
were  corrected  for  loss  of  sulfur  to  the 
vapor  by  means  of  the  familiar  relation 
PV  =  nRT.  The  molecular  weight  of  sul- 
fur was  calculated  from  the  data  summar- 
ized by  West  (1950).  The  vapor  pressures 
were  estimated  from  the  extrapolated  data 
of  Allen  and  Lombard,  De  Rudder,  D'Or, 
Raeder,  Juza  and  Biltz,  and  Rosenqvist. 

The  resulting  solvus  curve  is  shown  in 
figure  27.  The  composition  attained  by 
extrapolating  the  curve  to  743°  C  is  44.9 
atomic  per  cent  Fe.  The  uncertainty  in 
these  compositions  is  about  ±0.1  atomic 
per  cent  Fe.  The  position  of  the  curve  be- 
low 650°  C  was  checked  at  600°  and  400° 
C  by  the  sintering  and  microscopic  method 
described  above. 

The  lamellar  phase  described  and  shown 
in  last  year's  report  (Year  Book  56,  fig.  25) 
was  repeatedly  identified  in  pyrrhotites 
rapidly  quenched  from  above  666°  C  and 
initially  more  sulfur-rich  than  45.65  atomic 
per  cent  Fe.  X-ray  powder  photographs 
of  pyrrhotite  containing  about  20  volume 
per  cent  lamellae  gave  a  series  of  reflections 
that  could  not  be  attributed  to  either  pyrite, 
marcasite,  the  parent  hexagonal  pyrrhotite, 
monoclinic  pyrrhotite,  smythite  (FesS4),  or 
kansite  (Fe9S8).  The  five  most  prominent 
reflections  attributed  to  the  lamellar  phase 
gave  the  following  d  values:   5.648,  5.346 


4.999,  2.562,  1.980  A.  The  symmetry  of  the 
crystal  structure  and  the  field  of  stability 
of  this  lamellar  phase  have  not  been  de- 

The   determinative    curve   relating   the 
d(l02)  values  of  pyrrhotite  solid  solutions 

their  d(\02)  values)  of  pyrrhotites  from 
charges  whose  compositions  were  more 
sulfur-rich  than  46.0  atomic  per  cent  Fe. 
The  subsequent  detection  of  exsolved  py- 
rite  in  these  pyrrhotites,  however,  indicates 
that  the  true  limit  of  solid  solution  must 




g.    500 



Pyrrhotite   solid 




Pyrrhotite  +  Liquid 

Pyrrhotite     +    Pyrite 

O  Determined  by  the   X-roy 

A,A    Determined  by  the  sintering 
and  microscopic  method 


47.00  46.00  4  5.00 

-< —   Atomic  %  Fe 



Fig.  27.  A  portion  of  the  FeS-S  equilibrium  diagram  as  it  would  look  at  a  pressure  of  about 
10  bars.  Circles  represent  pyrrhotite  compositions  determined  by  the  X-ray  method.  Open  and 
hatched  triangles  represent  the  bulk  compositions  of  runs  containing  the  condensed  phases  pyrrho- 
tite, and  pyrrhotite  and  pyrite,  respectively.  Data  points  have  been  projected  on  this  isobaric 

synthesized  at  800°  C  to  their  iron  contents 
was  presented  in  Year  Book  55  (p.  177). 
The  composition  (46.0  atomic  per  cent  Fe) 
designated  as  the  most  sulfur-rich  solid 
solution  attainable  at  this  temperature  is 
believed  to  be  incorrect.  This  limit  of  solid 
solution  was  based  on  the  similarity  of  the 
iron  contents  (judged  by  the  constancy  of 

lie  somewhat  to  the  sulfur  side  of  the  re- 
ported limit  of  solid  solution.  Pyrrhotite 
solid  solutions  less  sulfur-rich  than  46.0 
atomic  per  cent  Fe  contained  no  detectable 
pyrite;  hence,  the  determinative  curve, 
based  on  the  measurement  of  these  pyrrho- 
tites, is  believed  to  be  correct. 
The  d(  102)  -value/composition   relation 


has  been  studied  further  by  measuring  the 
d(102)  values  of  pyrrhotite  solid  solutions 
synthesized  at  600°  and  400°  C.  The  agree- 
ment of  these  data  with  those  obtained 
from  preparations  synthesized  at  800°  C 
indicates  that  the  d{  102)  -value/composi- 
tion relation  is  independent  of  the  temper- 
ature at  which  the  preparations  are  syn- 

The  possibility  of  applying  the  ^(102)- 
value/composition  relation  to  natural  pyr- 
rhotites  (containing  up  to  0.4  per  cent 
combined  cobalt  and  nickel  in  solid  solu- 
tion) has  been  investigated  over  the  com- 
position range  46.5  to  47.7  atomic  per  cent 
Fe.  The  iron  contents  of  six  pyrrhotites 
determined  by  chemical  analysis  were 
within  ±0.1  atomic  per  cent  Fe  of  the 
iron  content  as  indicated  by  the  X-ray 

The  temperatures  of  formation  of  a 
number  of  pyrrhotite-pyrite  pairs  (con- 
taining —  0.15  weight  per  cent  combined 
nickel  and  cobalt)  from  mineral  deposits 
were  estimated  at  arbitrarily  assumed  total 
pressures  of  1000  and  2000  bars,  approx- 
imately the  pressures  exerted  by  2.5  and  5 
miles  of  rock  cover,  respectively.  The  iron 
contents  of  all  but  one  pyrrhotite,  which 
was  determined  by  chemical  analysis,  were 
determined  by  the  X-ray  method.  The 
temperatures  of  formation  were  estimated 
using  the  solvus  curve  determined  at  1000 
and  2000  bars  total  pressure  which  was 
presented  in  last  year's  report  (fig.  26,  p. 
194).  Table  8  gives  the  iron  contents  of 
each  pyrrhotite,  their  estimated  tempera- 
tures of  formation,  and  a  brief  statement 
of  the  type  of  deposit  involved.  The  iron 
content  of  sphalerite  coexisting  with  pyr- 
rhotite and  pyrite  in  the  Heath  Steele  Mine 
indicated  a  temperature  of  formation  of 
about  550°  C  at  2000  bars  using  Kullerud's 
(1953)  data  for  the  FeS-ZnS  system.  Pyr- 
rhotite formed  in  equilibrium  with  pyrite 
in  the  same  specimen  also  gives  550°  C 
(see  table  8). 

Figure  28  shows  the  distribution  of  the 
iron    contents   of   61    natural    pyrrhotites 

plotted  against  the  number  of  measure- 
ments in  each  %  atomic  per  cent  Fe  range. 
This  compilation  includes  the  results  of 
chemical  analysis  obtained  from  the  liter- 
ature and  from  the  present  study,  as  well 
as  the  X-ray  analysis  of  material  contain- 
ing —  0.4  per  cent  combined  cobalt  and 
nickel.  Chemical  analysis  was  recalculated 
to  100  per  cent  where  required.  All  but 
one  of  these  pyrrhotites,  a  troilite,  are  of 
terrestrial  origin  and  come  from  base- 
metal  deposits,  pegmatites,  schists,  and 
basic  igneous  rocks.  The  iron  contents  of 
these  pyrrhotites  fall  into  two  distinct 
groups,  designated  A  and  B,  the  majority 
of  measurements  being  in  group  B.  The 
pyrrhotites  of  group  B  were  generally  as- 
sociated with  pyrite  or  chalcopyrite  or 
both,  whereas  only  one  pyrrhotite  of  group 
A  showed  a  trace  of  any  other  sulfide. 

The  curve  representing  the  composition 
of  sulfur-rich  liquids  that  can  coexist  in 
equilibrium  with  pyrrhotite  solid  solution 
and  vapor  above  743°  C  is  being  deter- 
mined. Charges  consist  of  sulfur  and  a 
single  grain  of  iron  (converted  to  pyrrho- 
tite above  743°  C)  sealed  in  evacuated 
bent  silica  tubes.  A  single  grain  was  used 
in  preference  to  a  powder  to  decrease  the 
difficulty  of  separating  the  solid  phase  from 
the  sulfur-rich  phase  in  the  completed 
run.  Each  tube  was  rotated  at  1  rpm  in 
a  vertical  plane  in  the  furnace  for  up  to  6 
days  to  facilitate  intimate  contact  between 
the  single  pyrrhotite  lump  and  all  por- 
tions of  the  sulfur-rich  liquid.  Before  re- 
tracting a  completed  run  for  slow  air 
quenching,  each  tube  was  tipped  so  that 
the  single  pyrrhotite  lump  remained  in 
the  short  arm  of  the  bent  tube  while  the 
liquid  drained  into  the  long  arm.  After 
quenching,  the  contents  of  the  long  arm 
consisted  largely  of  a  sulfur-rich  crystal- 
line phase  (containing  small  sulfide  crys- 
tals precipitated  during  cooling)  repre- 
senting original  liquid,  which  was  capped 
by  a  virtually  pure  sulfur  phase  represent- 
ing condensed  vapor.  The  two  phases 
were  separated,  and  the  sulfide-containing 

TABLE  8.    Estimated  Temperature  of  Formation  of  a  Number  of  Ore  Deposits  Containing 

Pyrrhotite  and  Pyrite. 

No.  and 







P=  P  = 
1000  2000 
bars      bars 

Geological  Setting 

1  Lucky  Strike  Mine,  Colo. 

2  Bicroft  Uranium  Mine,  Ont. 

3  Heath  Steele  Mine,  N.  B. 

4  Aldermac  Mine,   Que. 

5  East  Sullivan  Mine,  Que. 

6  Clearwater  Brook,  N.  B. 

7  Highland  Surprise  Mine,  Ida. 

8  Brunswick  Mine,  N.  B. 

9  Burra  Burra  Mine,  Tenn. 

46.40  560       610  Replacement  of  crystalline  limestones 

46.51 1         540      585  Vein   in  uraniferous   pegmatite  cutting 

biotites  paragneiss 
46.62  515      550  Replacement  of  acid  and   intermediate 

volcanic  tuffs  and  flows 
46.66  510       540  Replacement  of  Keewatin  volcanics 

46.64  510      540  Massive  sulfide  replacement  of  Precam- 

brian  volcanics 
46.72  490      520  Replacement  of  siliceous   metasedimen- 

tary  rocks 
47.00  -         420  -   425  -       Replacement  of  fractured  zone  in  sili- 
46.60  515       555  ceous  argillite  and  quartzite 

46.78  455      495  Massive  sulfide  replacement  of  siliceous 

metasedimentary  rocks 
46.86  440      470  Massive  sulfide  replacement  of  crystal- 

line limestones 

*  The  uncertainty  in  these  compositions  determined  by  the  X-ray  method  is  not  known.  However, 
a  comparison  of  chemical  analysis  and  X-ray  analysis  in  6  instances  described  in  the  text  indicates 
agreement  to  ±0.1  atomic  per  cent  Fe.  Until  more  information  is  obtained  the  uncertainty  in  these 
compositions  is  taken  to  be  ±0.1  atomic  per  cent  Fe. 

t  Determined  by  chemical  analysis.  X-ray  analysis  gave  46.58  atomic  per  cent  Fe.  Analyst,  M.  K. 
Carron.  Job  3723. 

Source  of  Material 

No.  1     A.  H.  Koschmann,  U.  S.  G.  S. 

No.  2     P.    K.    Cunningham-Dunlop,    Princeton 

No.  4     #3432  McGill  University  collection. 
No.  5    J.  R.  Assad,  McGill  University. 

No.  6    W.  Petruk,  McGill  University. 

No.  7    R.    G.    Coleman   and    V.   C.   Fryklund, 


No.  9    C.  S.  Ross',  U.  S.  G.  S. 



Method  of  composition  determination 
jj         "1    Chemical  analysis 
1    ^7]    X-ray  analysis 


48.50  48.00  47.50 

•< —  Atomic  %  Fe 

4  7.00 


Fig.  28.   Diagram  showing  the  distribution  of  iron  content  of  61  pyrrhotites.   The  height  of  each 
column  indicates  the  number  of  each  kind  of  measurement. 



phase  was  analyzed  for  iron.  The  solu- 
bility of  iron  in  sulfur  liquid  even  at 
895°  C  was  found  to  be  ~0.1  weight  per 
cent  Fe. 

BETWEEN  400°  AND  800°  C 

E.  H.  Roseboom,  Jr.,  and  G.  Kullerud 

Of  great  significance  to  the  mining  in- 
dustry is  the  system  Cu-Fe-S,  which  con- 
tains most  of  the  sulfides  found  in  a 
large  number  of  ore  deposits.  Although 
numerous  workers  have  investigated  vari- 
ous parts  of  the  system,  the  phase  rela- 
tions in  the  sulfur-rich  portion  have  re- 
ceived little  attention.  Recently  Kullerud 
has  studied  the  sulfur-rich  part  of  the  sys- 
tem Cu-S,  and  Arnold,  as  well  as  Kullerud 
and  Yoder,  has  investigated  the  Fe-S  join. 
The  study  most  closely  related  to  the 
present  work  is  that  of  Merwin  and  Lom- 
bard (1937). 

In  the  system  Cu-Fe-S,  a  sulfur-rich 
liquid  phase  persists  down  to  a  tempera- 
ture probably  only  slightly  lower  than  the 
melting  point  of  sulfur  (114.5°  C  for  pure 
S).  The  present  work  is  an  investigation 
of  the  solid  phases  in  equilibrium  with 
this  liquid  plus  a  vapor  phase  between 
400°  and  800°  C. 

Mixtures  of  synthetic  phases  or  of  cop- 
per and  iron  metals  were  heated  in  sealed 
evacuated  silica  glass  tubes  containing 
enough  sulfur  to  provide  a  liquid  as  well 
as  a  vapor  phase.  Above  550°  C  equilib- 
rium was  attained  in  a  few  hours.  Above 
565°  C,  mixtures  of  copper,  iron,  and  sul- 
fur were  used  as  starting  materials;  below 
565°  C,  mixtures  of  covellite,  pyrite,  and 
sulfur.  The  temperatures  of  invariant 
points  below  565°  (fig.  29;  points  1,  2,  and 
3)  were  located  by  taking  two  samples,  one 
that  had  previously  been  annealed  to  pro- 
duce the  stable  assemblage  above  the  in- 
variant point  under  investigation  and  an- 
other that  had  been  similarly  annealed 
below  that  invariant  point,  and  heating 
them  in  separate  tubes  side  by  side.  Thus, 
equilibrium  was  approached  from  both 
higher  and  lower  temperatures.   Near  the 

invariant  point  (fig.  29,  point  1)  at  434°  C, 
runs  of  1  month  produced  only  about  5 
per  cent  reaction,  but  the  direction  was 
clearly  established  from  the  pairs  of  runs. 
The  solid  phases  were  identified  by  means 
of  polished  sections  and  by  X-ray  dififrac- 
tometer  methods. 

The  results  of  these  experiments  can- 
not be  depicted  on  a  simple  isothermal  or 
isobaric  section  because  the  vapor  pressure 
increases  as  the  temperature  is  raised.  Al- 
though the  experimental  work  as  described 
is  simple,  the  P-T  conditions  in  the  part 
of  the  system  studied  are  rather  complex. 
In  order  to  relate  the  portion  of  the  system 
studied  to  other  experimental  work  we 
shall  first  consider  a  P-T  projection  of  a 
P-T-X  diagram  covering  a  wider  range  of 
conditions  than  those  actually  studied 
(fig.  29). 

The  curve  extending  steeply  downward 
from  point  4  is  the  dissociation  curve  for 
covellite  going  to  digenite  plus  a  sulfur- 
rich  vapor  in  the  binary  system  Cu-S. 
Similarly  the  curve  extending  steeply 
downward  from  point  7  is  the  dissociation 
curve  for  pyrite  going  to  pyrrhotite  plus 
a  sulfur-rich  vapor  in  the  binary  system 
Fe-S.  In  each  case,  as  the  dissociation 
curve  is  crossed  at  constant  pressure  going 
toward  a  higher  temperature,  a  solid  phase 
dissociates  into  a  solid  phase  lower  in  sul- 
fur plus  a  sulfur-rich  vapor.  As  the  re- 
actions occurring  on  these  curves  each 
involve  three  phases  in  two-component 
systems,  the  dissociation  curves  represent 
univariant  conditions. 

The  curves  extending  steeply  downward 
from  points  1,  2,  3,  5,  and  6  are  univariant 
curves  for  three  solids  plus  a  vapor  in  the 
ternary  system  Cu-Fe-S.  They  are  analo- 
gous to  the  dissociation  curves  in  the 
binary  systems  except  that  one  additional 
solid  phase  is  always  present.  They  are  of 
two  general  types:  (1)  A+B< — >C+V, 
(2)  C< — >A  +  B+V.  A,  B,  and  C  are 
sulfides,  and  V  is  a  sulfur-rich  vapor.  The 
curves  from  points  1,  2,  5,  and  6  are  of  the 
first  type,  and  the  curve  from  point  3  is 
the  second  type. 


If  any  one  of  these  univariant  curves  is 
followed  to  higher  temperatures  and  pres- 
sures, it  is  seen  to  terminate  at  a  point 
where  the  vapor  phase  begins  to  condense, 
and  thus  a  fifth  phase,  a  sulfur-rich  liquid, 
appears.  The  temperature  and  pressure  at 
which  this  occurs  represent  an  invariant 
point  on  a  P-T  projection,  because  there  are 
now  five  phases  in  a  three-component  sys- 
tem and  there  are  no  degrees  of  freedom 

vertical  curves  represent  the  same  reac- 
tions as  the  curves  extending  steeply  down- 
ward from  the  invariant  points,  except 
that  the  sulfur-rich  vapor  has  been  re- 
placed by  a  sulfur-rich  liquid  in  the  sys- 

We  have  already  discussed  two  of  the 
five  univariant  curves.  The  remaining 
three  do  not  involve  the  formation  or  de- 
composition   of    a    solid    phase,    but    are 

100,000  mm 

10,000  mm 


ct    1,000 

4  55  mm* 

100  mr 

10  mm 








Fig.  29.  Pressure-temperature  projection  of  a  part  of  the  system  Cu-Fe-S.  The  points  on  the 
horizontal  dashed  lines  are  from  Merwin  and  Lombard's  455  mm  isobaric  section,  cv,  covellite;  py, 
pyrite;  dg,  digenite;  po,  pyrrhotite;  cp,  chalcopyrite;  SL,  sulfur  liquid;  SF,  sulfur  vapor. 

remaining.  The  ternary  invariant  points 
are  numbers  1,  2,  3,  5,  and  6  in  figure  29. 
Each  of  these  invariant  points  represents 
the  point  of  origin  of  five  univariant 
curves,4  four  of  which  involve  a  vapor, 
and  a  fifth  which  does  not.  From  each 
invariant  point  the  last  is  shown  as  a 
nearly  vertical  dashed  line  extending 
toward  regions  of  higher  pressures.  These 

4  Three  of  the  univariant  curves  emanating 
from  each  of  points  1,  2,  and  3  are  probably 
nearly  coincident  with  the  curve  labeled  SL-SF 
and  are  not  shown  as  separate  curves.  These  are 
the  curves  representing  the  P-T  conditions  of  the 
present  work. 

simply  curves  along  which  both  a  sulfur- 
rich  liquid  and  a  vapor  are  in  equilibrium 
with  two  solid  phases.  These  curves  con- 
nect the  various  invariant  points,  and  the 
pairs  of  solids  that  are  stable  together  with 
liquid  and  vapor  along  them  are  indicated 
in  figures  30  and  31.  On  the  high-pressure, 
low-temperature  side  of  these  curves,  the 
two  solids  are  stable  with  a  sulfur-rich 
liquid;  on  the  low-pressure,  high-tempera- 
ture side,  the  same  two  solids  are  stable 
with  a  sulfur-rich  vapor.  Some  of  these 
solid-solid-liquid-vapor  curves  approxi- 
mately coincide  with  one  another  and  with 


Fig.  30.  The  solidus  (in  the  presence  of  a 
vapor)  in  the  system  Cu-Fe-S  between  400°  and 
800°  C.  Same  abbreviations  as  preceding  figure 
except:  D,  digenite;  Cc,  chalcocite. 

the  liquid-vapor  curve  for  sulfur  at  low 
temperatures.  For  example,  running  from 
point  1  toward  points  2  and  3  there  are 
actually  two  curves  of  this  type.  One 
represents  the  vapor  pressure  over  a  sulfur- 

rich  liquid,  covellite,  and  "Cu5FeS6."  The 
other  represents  the  vapor  pressure  over 
liquid,  "CusFeSe,"  and  pyrite.  Over  this 
P-T  range  it  is  unlikely  that  sufficient  cop- 
per or  iron  goes  into  the  liquid  or  vapor 
phase  to  cause  the  vapor  composition,  and 
consequently  the  vapor  pressure,  to  be  ap- 
preciably different  from  those  of  sulfur. 
Consequently,  these  two  curves  coincide 
with  one  another  and  also  with  the  liquid- 
vapor  curve  for  pure  sulfur. 

The  line  labeled  Sl-Sf  is  the  liquid- 
vapor  curve  or  vaporization  curve  for  pure 
sulfur.  At  low  temperatures  this  curve 
coincides  with  the  solid-solid-liquid-vapor 
curves  because  the  liquid  and  vapor  phases 
in  equilibrium  with  the  solid  phases  are 
very  close  in  composition  to  pure  sulfur. 
At  higher  temperatures,  sufficient  amounts 
of  iron  or  copper  or  both  may  be  dissolved 
in  the  liquid  and  the  vapor  phases  to 
cause  the  solid-solid-liquid-vapor  curves 
to  shift  appreciably  from  the  sulfur  va- 
porization curve.  For  this  reason  the 
solid-solid-liquid-vapor  curves  are  shown 
as  having  lower  vapor  pressures  than  the 
sulfur  vaporization  at  temperatures  above 
500°  C. 


<u     600 



po    / 

cp  +  po 


\              cp 


cp  +  py 


dg  (bn) 






v     . 

dg  -  py 


•          \     •      •  s** 

"Cu5  Fe  S6  "   +  py 



'     »' 

»        r\/i    • 


:'fCu5Fe  S6" 

cv  +  py 
I             1             1             1             1             I             1             I             I 



80  90 


Fig.  31.    The  solidus  in  the  system  Cu-Fe-S  as  seen  from  the  S  corner,  i.e.  projected  onto  the 
Fe-Cu  side  of  figure  30. 


Experiments  conducted  by  Rosenqvist 
(personal  communication,  1957)  and  later 
by  Arnold  (1958)  on  the  solubility  of  iron 
in  liquid  sulfur  indicate  small  solubilities 
(<0.1  per  cent  at  800°  C),  suggesting 
that  the  pressure  difference  between  the 
total  vapor  pressure  over  pure  sulfur  and 
over  sulfur  saturated  with  iron  might  be 
small.  Extrapolation  by  Kullerud  and 
Yoder  (1957)  of  the  known  data  on  dis- 
sociation of  pyrite  to  pyrrhotite  and  vapor 
to  the  invariant  point  at  743°  C,  where 
pyrrhotite,  pyrite,  liquid,  and  vapor  all 
are  stable,  gives  a  total  pressure  of  about 
10  atm.  On  the  other  hand,  the  vapor 
pressure  over  pure  sulfur  at  743°  C  is 
given  by  West  (1950)  as  about  21  atm. 

Merwin  and  Lombard  constructed  a 
"constant  pressure"  section  at  455  mm  and 
made  various  measurements  of  dissocia- 
tion pressures.  They  encountered  on  their 
section  the  various  dissociation  curves  at 
the  temperatures  indicated  in  figure  29, 
on  the  line  at  455  mm. 

The  present  work  is  not  a  constant-pres- 
sure section  but  follows  along  the  solid- 
solid-liquid-vapor  curves  and  passes 
through  all  the  invariant  points  shown. 
With  respect  to  the  solid  phases  only,  such 
a  section  is  bound  to  resemble  Merwin  and 
Lombard's  because  their  section  crosses  the 
univariant  curves  that  originate  at  these 
invariant  points. 

Figure  30  shows  the  solid  phases  that 
exist  in  equilibrium  with  the  sulfur-rich 
liquid  and  vapor  in  the  system.  The 
liquid  and  vapor  are  not  shown,  but  are 
understood  as  lying  near  the  sulfur  corner 
and  being  stable  with  all  the  phases  shown. 
Thus,  below  434°  C,  pyrite  and  covellite 
are  stable  with  the  liquid  and  vapor.  At 
434°  C  (invariant  point  1  in  fig.  29),  py- 
rite, covellite,  and  CusFeSe  are  stable  to- 
gether. Immediately  above  434°  C,  pyrite 
and  CusFeSe  are  stable,  as  are  Cu5FeS6 
and  covellite.  The  latter  pair  are  stable 
together  up  to  482°  C  (invariant  point  2, 
fig.  29).  At  this  temperature,  a  digenite 
with  about  7  per  cent  iron  substituting 
for    copper    is    also    stable.    Immediately 

above  this  temperature,  a  digenite  lower 
in  iron  is  stable  with  covellite  and  a  dige- 
nite higher  in  iron  is  stable  with  CusFeSe. 
The  higher-iron  digenite  becomes  richer 
in  iron  with  rising  temperature  until  at 
501°  C  the  CusFeSe  breaks  down.  Above 
501  °  C,  the  iron-rich  digenite  is  stable  with 
pyrite.  It  becomes  richer  in  iron  with  in- 
creasing temperature.  At  about  550°  C, 
there  is  a  complete  solid  solution  between 
bornite  and  digenite. 

The  low-iron  digenite  that  became  sta- 
ble with  covellite  at  482°  C  becomes  lower 
in  iron  with  rising  temperature  until  it 
becomes  pure  digenite  at  507°  C  (invari- 
ant point  4,  fig.  29),  which  is  an  invariant 
point  in  the  binary  system  Cu-S.  At 
568°  C  the  digenite-bornite  solid  solution 
is  stable  with  pyrite  and  with  a  copper- 
rich  chalcopyrite  (invariant  point  5,  fig. 
29).  Above  568°  C  the  digenite-bornite 
solid  solution  is  stable  with  a  copper-rich 
chalcopyrite,  and  a  chalcopyrite  which  be- 
comes progressively  richer  in  iron  with 
rising  temperature  is  stable  with  pyrite. 
At  739°  C  the  latter  pair  are  stable  with 
a  pyrrhotite  containing  about  5  per  cent 
copper  substituting  for  iron  (invariant 
point  6,  fig.  29).  Immediately  above 
739°  C  chalcopyrite  (of  about  cubanite 
composition)  is  stable  with  the  copper- 
bearing  pyrrhotite  and  a  copper-bearing 
pyrrhotite  is  stable  with  pyrite.  This  last 
pyrrhotite  becomes  lower  in  copper  with 
rising  temperature,  until  at  743°  C  (in- 
variant point  7,  fig.  29)  pyrite  breaks  down 
and  pure  iron  pyrrhotite  becomes  stable. 
Above  this  temperature  three  solid  solu- 
tions are  stable:  a  copper-bearing  pyrrho- 
tite, a  chalcopyrite  solid  solution  contain- 
ing Cu  and  Fe  in  ratios  variable  from 
about  Cu3Fe7  to  Cu6Fe4,  and  a  digenite- 
bornite  solid  solution.  The  copper-rich 
pyrrhotite  is  stable  with  the  iron-rich  chal- 
copyrite, and  the  copper-rich  chalcopyrite 
is  stable  with  the  iron-rich  digenite.  It 
should  be  remembered  that  all  the  above 
solid  phases  are  stable  with  both  a  sulfur- 
rich  liquid  and  a  sulfur-rich  vapor. 

Figure  31  is  a  projection  of  the  surface 


shown  in  figure  30  as  seen  from  the  sulfur 
corner.  On  this  diagram  are  plotted  the 
Cu :  Fe  ratios  of  the  samples  and  the  tem- 
peratures at  which  they  were  heated.  The 
various  regions  indicate  what  sulfide  phases 
were  encountered  in  the  runs,  all  of  which 
contained  a  sulfur-rich  liquid  and  a  vapor. 
The  sulfur  contents  of  the  various  phases 
could  not  be  determined,  as  they  could  not 
be  separated  from  the  liquid  and  con- 
densed vapor;  moreover,  on  the  copper- 
rich  side  some  covellite  or  "CusFeSe"  or 
both  frequently  form  on  quenching  from 
reaction  between  the  digenite-bornite 
phase  and  liquid  or  vapor.  It  seems  likely, 
however,  that  the  sulfur  contents  are 
about  the  same  as  those  determined  by 
Merwin  and  Lombard. 

This  study  indicates  the  maximum  tem- 
peratures at  which  some  of  these  mineral 
assemblages  can  exist  in  the  presence  of 
a  sulfur-rich  vapor.  Thus,  covellite  and 
pyrite  cannot  exist  together  above  434°  C, 
although  covellite  alone  can  exist  up  to 
507°  C  and  pyrite  alone  up  to  743°  C.  A 
list  of  such  assemblages  and  correspond- 
ing maximum  temperatures  in  the  pres- 
ence of  sulfur-rich  vapor  follows:  covel- 
lite +  pyrite,  434°  C;  covellite +  "Cu5FeSG," 
482°  C;  pyrite +  "Cu.3FeS6,"  501°  C;  born- 
ite  +  pyrite,  568°  C;  chalcopyrite  + pyrite, 
739°  C. 

With  no  vapor  present,  and  under  con- 
fining pressures  greater  than  those  existing 
at  the  various  invariant  points,  these  as- 
semblages would  break  down  at  higher 
temperatures  than  those  listed  to  one  or 
two  solids  plus  a  sulfur-rich  liquid.  Kul- 
lerud's  work  on  dissociation  of  covellite,  as 
well  as  the  work  by  Kullerud  and  Yoder 
on  dissociation  of  pyrite,  at  high  pressures 
where  vapor  was  absent  showed  that  the 
stability  fields  of  these  two  minerals  were 
increased  by  about  10°  to  15°  C  per  1000 
bars  of  pressure,  and  changes  of  a  similar 
order  of  magnitude  might  be  expected  for 
other  dissociations  of  sulfides  or  assem- 
blages of  sulfides. 

The  study  also  showed  that  when  liquid 
and  vapor  are  present  digenite  and  bornite 

form  a  complete  solid  solution  above  about 
550°  C.  The  first  digenite  to  appear  with 
rising  temperature  was  one  nearly  midway 
in  the  series  digenite-bornite;  it  was  ob- 
served at  482°  C.  This  same  phase  ap- 
peared in  Merwin  and  Lombard's  work  at 
472°  C  at  455  mm  but  was  identified  as 
chalcocite  from  its  appearance  under  the 
microscope.  Thus,  the  crest  of  any  solvus 
between  digenite  and  bornite  is  probably 
below  this  temperature.  The  existence  of 
such  a  solvus  is  often  indicated  in  ore  de- 
posits where  digenite  and  bornite  occur  to- 
gether as  separate  phases.  The  assemblage 
chalcocite-pyrite  has  been  described  from 
many  localities.  Such  an  assemblage  is 
incompatible  with  the  tie  line  between 
digenite  and  bornite  found  under  the  con- 
ditions of  the  present  study  but  may  be 
stable  in  parts  of  the  system  that  have  not 
been  investigated. 

The  phase  Cu5FeS6,  first  described  by 
Merwin  and  Lombard,  had  a  Cu:Fe  ratio 
of  17:3  under  the  P-T  conditions  of  the 
present  study.  This  phase  has  never  been 
definitely  identified  in  any  natural  de- 
posits, although  its  strong  pleochroism  and 
intense  anisotropism  make  it  very  easy  to 
observe  in  polished  sections.  The  assem- 
blage covellite  +  pyrite  (which  would  oc- 
cur at  lower  temperatures  and  higher 
vapor  pressures)  and  the  assemblage  born- 
ite (digenite)  +  pyrite  (which  would  oc- 
cur at  higher  temperatures  and  lower  va- 
por pressures)  are  both  known  in  ore 
deposits.  Merwin  and  Lombard  heated 
"Cu5FeS6"  for  5  days  at  300°  and  200°  C, 
8  days  at  100°  C,  and  15  days  at  87°  C, 
and  it  showed  no  sign  of  breaking  down. 
The  failure  of  this  phase  to  appear  in 
nature  remains  to  be  explained. 

The  phase  labeled  chalcopyrite  in  figure 
31  extends  over  a  wide  Cu:Fe  range  (at 
800°  C  from  about  Cu58Fe42  to  Cu3oFe7o), 
beyond  the  ratio  of  1:2  of  the  mineral 
cubanite.  The  X-ray  pattern  obtained  for 
this  composition  range  is  that  of  chalcopy- 
rite. Thus,  since  there  is  a  continuous 
solid  solution  between  chalcopyrite  and 
cubanite  composition,  under  these  condi- 


tions  cubanite  must  invert  to  the  chalcopy- 
rite-type  structure  at  some  temperature  be- 
low that  at  which  the  solid  solution  series 
is  complete.  From  figure  31  it  is  noted 
that  under  the  conditions  of  these  experi- 
ments cubanite  must  possess  the  chalcopy- 
rite  structure  at  least  above  720°  C. 

P.  B.  Barton,  Jr.,5  and  G.  Kullerud 

The  composition  of  mix-crystals  in  the 
FeS-ZnS  binary  system  (Kullerud,  1953) 
has  been  extensively  used  in  geological 
thermometry.    For  this  purpose  the  im- 

sphalerite  (  + vapor)  field  (extending  to 
point  a  in  fig.  32#),  which  contains  no 
measurable  excess  of  sulfur.  Pyrite  is  es- 
sentially stoichiometric  (Kullerud  and  Yo- 
der,  1958),  and  the  liquid  field  is  almost 
pure  sulfur  (less  than  0.1  per  cent  Fe  +  Zn 
can  dissolve  in  sulfur  at  this  temperature). 
The  tie  lines  of  the  three-phase  fields  Po  + 
Sp  +  F,  Py  +  Sp+F,  and  Sp  +  L+V  rep- 
resent lines  of  equal  vapor  pressure,  and 
the  vapor  pressure  across  each  of  these 
fields  increases  toward  the  more  sulfur- 
rich  compositions. 
Geologically,  we   are  limited   to   those 



to  <*) 

Fig.  32.  Phase  relations  in  the  FeS-ZnS-S  part  of  the  Fe-Zn-S  system  (a)  at  650°  C,  (Jb)  at  850°  C. 

portant  part  of  the  binary  join  is  the  solvus 
curve  that  defines  the  iron  content  of 
sphalerite  saturated  with  FeS  as  a  func- 
tion of  temperature.  The  natural  iron  sul- 
fides (pyrite  and  pyrrhotite)  commonly 
found  with  sphalerite  are  richer  in  sulfur 
than  stoichiometric  FeS,  and,  therefore,  the 
more  sulfur-rich  part  of  the  Fe-Zn-S 
ternary  system  must  be  considered. 

Figure  32a  shows  the  FeS-ZnS-S  part 
of  the  ternary  Fe-Zn-S  system  at  650°  C. 
Two  of  the  solid  solution  series,  shown  by 
heavy  lines,  on  the  bounding  joins  are  for 
all  practical  purposes  binary:  the  pyrrho- 
tite (  + vapor)  field,  which  contains  essen- 
tially no  zinc  (Kullerud,  1953),  and  the 

5  U.  S.  Geological  Survey,  publication  approved 
by  Director. 

fields  that  do  not  contain  liquid  sulfur, 
because  sulfur  rarely,  if  ever,  accompanies 
sphalerite  as  a  hydrothermal  mineral.  The 
experiments  so  far  have  been  designed  to 
investigate  the  two  univariant  fields  con- 
taining the  assemblages  Py  +  Po  +  Sp  +  F 
and  Py  +  Sp  +  L+F.  At  650°  C  a  dif- 
ference between  the  composition  of  sphaler- 
ite in  equilibrium  with  pyrrhotite  and  that 
in  equilibrium  with  stoichiometric  FeS 
was  found  to  exist.  Figure  33  shows  the 
variation  in  the  iron  content  of  sphalerite 
as  a  function  of  the  phase  assemblage  and 
temperature.  Because  the  composition  of 
the  (Fe,Zn)S  mix-crystals  falls  essentially 
on  the  FeS-ZnS  binary  join,  the  experi- 
mental points  from  the  ternary  part  of  the 
system  have  been  projected  on  the  plane 



of  the  FeS-ZnS  binary.  The  new  experi- 
mental points  are  shown  by  a  bar  indicat- 
ing the  uncertainty  in  the  composition  of 
the  sphalerite.6  In  figure  33  the  curve 
marked  FeS  +  Sp  +  V  is  Kullerud's  original 
FeS-ZnS  solvus  curve.  The  curve  marked 
Sp  +  Py+L+F  gives  the  composition  of 
sphalerite  in  equilibrium  with  pyrite  as 
well  as  liquid  and  vapor.  Pure  pyrite 
melts  incongruently  at  743°  C  (Kullerud 

between  500°  and  600°  C.  Kullerud's  curve 
is  thus  applicable  to  any  pyrrhotite-sphal- 
erite  or  pyrrhotite-pyrite-sphalerite  assem- 
blage below  this  temperature. 

The  central  area  in  figure  33  is  that  of 
Py  +  Sp+F.  It  is  seen  that  a  sphalerite 
which  contains  10  mole  per  cent  FeS  and 
which  formed  in  equilibrium  with  pyrite 
could  have  been  deposited  at  any  tempera- 
ture  between  340°    and   680°    C.    Points 





I  500 






Fig.  33.   Variation  in  iron  content  of  sphalerite  as  a  function  of  phase  assemblage  and  temperature. 

and  Yoder,  1957) .  Above  this  temperature 
sphalerite  exists  in  equilibrium  with  pyr- 
rhotite,  liquid,  and  vapor  as  shown  in 
figure  32b.  The  Sp  +  Py  +  Po  +  F  curve 
represents  the  composition  of  sphalerite  in 
equilibrium  with  pyrite,  pyrrhotite,  and 
vapor.  It  is  seen  that  the  curves  Sp  +  FeS 
+  F  and  Sp  +  Py  +  Po+F  diverge  above 
600°  C  and  for  purposes  of  practical  meas- 
urement   coincide    at    some    temperature 

6  A  new  curve  relating  the  iron  content  of 
synthetic  sphalerites  to  their  cell  dimensions  was 
established  by  use  of  the  Norelco  X-ray  diffrac- 
tometer  with  Fe  radiation  and  Mn  filter,  using 
CaF2  as  internal  standard. 

within  this  area  are  determined  by  the 
sulfur  pressure  over  the  assemblage  at  the 
time  of  its  formation.  The  iron  content 
of  sphalerites  deposited  in  equilibrium 
with  pyrite  where  no  pyrrhotite  is  pres- 
ent, therefore,  gives  a  minimum  tempera- 
ture of  formation  of  the  sphalerite-pyrite 
assemblage,  when  the  binary  FeS-ZnS 
solvus  curve  is  used  as  a  temperature  indi- 
cator. Similarly,  a  "maximum"  tempera- 
ture is  obtained  if  the  curve  is  used  that 
relates  composition  of  sphalerite  existing 
in  equilibrium  with  pyrite  and  liquid  (  + 
vapor).  In  order  to  specify  at  which  tem- 
peratures,  between    these    limits,   natural 



sphalerite-pyritc  assemblages  were  depos- 
ited, the  vapor  pressure  of  sulfur  during 
formation  as  well  as  the  relations  between 
iron  content  of  sphalerite  and  sulfur  pres- 
sure must  be  known.  Whereas  the  partial 
vapor  pressure  during  deposition  of  sphal- 
erite-pyrite  assemblages  can  often  be  es- 
timated from  coexisting  sulfides,  the  solvus 
curves  relating  composition  of  sphalerite 
to  sulfur  pressure  are  being  studied  experi- 

common.  In  addition,  the  system  contains 
the  phases  FeAs  and  Fe2As,  of  which  no 
natural  occurrences  are  known. 

Knowledge  of  the  phase  relations  among 
minerals  containing  arsenic  is  applicable 
to  a  number  of  important  ore  deposits. 
Hence  a  systematic  study  of  these  arsenides 
was  initiated  early  this  year  and  is  now 
nearly  completed.  The  work  was  done  in 
evacuated,  sealed,  silica  glass  tubes,  where 
a  vapor  was  always  present.    In  all  runs 

per   cent 

Fig.  34.    Phase  relations  in  the  Fe-As-S  system  at   600°  C.    Vapor  pressures   are  those  of  the 
system.   FeAsS  composition  is  marked  by  the  small  triangle  at  center. 

L.  A.  Clark1 

Minerals  of  this  system  occur  in  most 
known  sulfide  deposits.  The  common 
minerals  of  the  group  are  pyrite  (FeS2), 
pyrrhotite  (Fei-a?S),  and  arsenopyrite 
(FeAsS).  Loellingite  (FeAs2),  realgar 
(AsS),    and    orpiment    (As2S3)    are    less 

7  Fellow  of  the  National  Research  Council  of 

the  vapor  volume  was  reduced  to  such  an 
extent  that  the  change  in  bulk  composi- 
tion of  the  condensed  phases  due  to  its 
presence  was  not  more  than  0.25  weight 
per  cent;  usually  it  was  much  less.  Vapor 
pressure  varies  with  changes  in  the  coex- 
isting phases,  and  is  constant  in  each  four- 
phase  assemblage  at  constant  temperature. 
It  increases  from  a  few  millimeters  in  the 
Fe-rich  assemblages  to  a  few  atmospheres 
along  the  S-As  side  of  the  diagram  (fig. 


34)  as  estimated  from  known  vapor  pres- 
sures of  Fe,  S,  As,  and  Fe-S  assemblages. 
The  S-As  minerals  orpiment  and  realgar 
melt  at  slightly  above  300°  C,  and  so  at 
600°  C  there  is  a  continuous  liquid  field 
extending  along  the  S-As  binary  from  100 
to  22.8  ±0.2  weight  per  cent  sulfur.  Since 
less  than  0.1  per  cent  iron  is  soluble  in  the 
liquid    at    this    temperature,    the    ternary 

component.  The  solubility  of  sulfur  in 
loellingite  is  greater  than  3  weight  per 
cent  at  700°  C,  but  appears  to  be  less  than 
1  per  cent  at  600°  C.  At  the  latter  tem- 
perature several  months  are  required  to 
attain  equilibrium.  Work  has  not  been 
completed  on  the  composition  of  synthetic 
arsenopyrite.  There  are  three  indications 
that   it  is   As-rich,   with   an   approximate 

(a)    T<  480±  5°C 


(c)  T=670±  5  to  743  +  3°C 

(d)    T>  743  +  3°C 

Fig.  35.    Assemblages  stable  at  various  temperatures. 

liquid  field  is  too  narrow  to  plot  as  more 
than  a  line  in  figure  34. 

The  eight  univariant  four-phase  assem- 
blages at  600°  C  are  shown  in  figure  34. 
The  limit  of  arsenic  solid  solution  in  iron 
at  this  temperature  is  approximately  3  to 
7  per  cent.  Each  boundary  limit  of  all  the 
four-phase  regions  has  been  determined  at 
two  places  to  within  l/2  per  cent. 

At  600°  C  none  of  the  binary  phases  can 
take  into  solid  solution  more  than  a  frac- 
tion of  1  per  cent  of  the  respective  ternary 

composition  FeAsi.iSo.9i  (1)  from  the  re- 
sults of  three  silica  tube-in-tube  runs  the 
arsenopyrite  composition  was  calculated, 
the  only  assumption  being  that  there  is  a 
1:2  ratio  of  Fe  to  As  +  S;  (2)  the  exten- 
sions of  tie  lines  intersect  at  this  approxi- 
mate composition;  (3)  disregarding  the 
symmetry  difference  in  the  calculation,8  the 
cell   volume   of   synthetic   arsenopyrite   is 

8  For  a  discussion  of  arsenopyrite  symmetry  see 
the  Crystallography  section  of  this  report,  page 


approximately  1  per  cent  larger  than  in  the 
only  two  natural  specimens  measured  to 
date,  from  Freiberg,  East  Germany,  and 
Llallagua,  Bolivia.  Unfortunately,  chemi- 
cal analyses  of  these  arsenopyrites  are  not 
yet  available. 

Arsenopyrite  free  of  pyrrhotite  inclu- 
sions could  be  synthesized  only  from  bulk 
compositions  lying  within  the  arsenopy- 
rite-arsenic-liquid-vapor  region.  The  start- 
ing materials  were  synthetic  FeAs-2  and  a 
quenched  liquid  saturated  with  arsenic  at 
600°  C.  Near  the  end  of  a  3-week  heating 
period  at  600°  C  a  temperature  gradient  of 
1°  or  2°  along  the  tube  was  sufficient  to 
drive  the  excess  liquid  +  arsenic  to  one  end, 
leaving  pure  arsenopyrite  in  the  other  end. 

The  range  of  stability  of  some  of  the 
ternary  assemblages  is  shown  in  figure  35. 
Figure  35£  is  a  slightly  simplified  version 
of  figure  34.  These  assemblages  are  stable 
throughout  a  range  of  temperature  from 
480°  ±5°  C  to  near  670°  C.  The  minerals 
pyrite  and  arsenopyrite,  which  form  a  com- 
mon assemblage  in  nature,  can  coexist  in 
equilibrium  with  each  other,  in  the  pres- 
ence of  a  vapor,  only  below  480°  ±5°  C,  as 
shown  in  figure  35<z.  Although  the  tem- 
perature at  which  pyrite  and  arsenopyrite 
react  to  form  pyrrhotite  and  liquid  is  sub- 
ject to  a  pressure  effect  depending  princi- 
pally on  volume  change,  it  is  unlikely  to 
be  much  above  510°  C  within  the  range 
of  geologically  important  pressures.  The 
reverse  reaction  is  extremely  sluggish;  it 
is  well  started  but  far  from  complete  in 
runs  heated  50  days  at  450°  and  470°  C. 

In  the  presence  of  a  saturated  vapor, 
arsenopyrite  is  unstable  above  approxi- 
mately 670°  C.  Above  this  temperature, 
and  up  to  743°  ±3°  C,  the  assemblages 
shown  in  figure  35c  are  stable.  The  pyrite 
breakdown  temperature  was  found  to  be 
unchanged  by  the  presence  of  arsenic,  and 
it  may  therefore  be  assumed  that  very 
little  arsenic  is  soluble  in  pyrite.  A  step 
intermediate  between  figures  35£  and  35c 
has  been  omitted;  the  reaction  FeAsi.iSo.9 
+  As^FeAs2  +  L  takes  place  at  a  tempera- 
ture slightly  lower  than  that  of  the  arseno- 
pyrite breakdown.  The  univariant  assem- 

blages arsenopyrite-loellingite-liquid-va- 
por,  and  loellingite-arsenic-liquid-vapor 
are  then  stable  within  a  very  narrow  tem- 
perature range. 

Above  743°  ±3°  C,  probably  well  be- 
yond the  region  of  geological  interest,  only 
the  five  univariant  assemblages  shown  in 
figure  35^  remain.  FeAs  and  Fe2As  are 
stable  throughout  the  entire  temperature 
range,  to  well  above  900°  C.  These  com- 
pounds are  optically  very  similar  to  loel- 





Arsenopyrite               t 




+  * 









Invariant  temperature' 

670±5°C\     J 

1                   ^i 

1                              1 

500  600  700  800 

Temperature  °C 
Fig.  36.   Upper  stability  curve  of  arsenopyrite, 
FeAsnso.9  ^  Fei-tfS  +  FeAs2  +  L. 

lingite  and  arsenopyrite.  Possibly  they  oc- 
cur in  nature  and  have  not  been  recog- 

Figure  36  shows  the  upper  stability  curve 
of  arsenopyrite,  determined  with  samples 
in  welded  collapsible  gold  tubes,  within 
which,  under  an  applied  external  pressure, 
no  vapor  could  form.  This  is  a  univariant 
curve  which  defines  the  effect  of  pressure 
on  the  temperature  at  which  the  break- 
down reaction  FeAsi.iSo.8— >Fei-»S+FeAs2 
+  L  can  proceed,  provided  that  no  gold 
enters  the  phases.  This  curve,  when  pro- 
jected down  to  a  pressure  of  a  few  at- 
mospheres, yields  670°  ±5°  C  as  the  in- 
variant temperature  at  which  the  four 
phases  coexist  with  vapor.    This  point  is 



also  the  origin  of  four  other  univariant 
curves,  each  defining  a  reaction  involving 
a  vapor  phase. 

The  upper  stability  curve  sets  a  maxi- 
mum temperature  for  the  occurrence  of 
arsenopyrite  in  nature,  but  this  compound 
can  form  at  any  temperature  below  670° 
±5°  C,  and  its  presence  in  nature  cannot 
be  used  to  demonstrate  high  temperature 
of  formation,  as  has  often  been  claimed. 
Similarly,  if  equilibrium  between  naturally 
occurring  pyrite  and  arsenopyrite  can  be 
demonstrated,  480°  ±5°  C  is  the  upper 
limit  of  formation  for  this  mineral  pair. 
Equilibrium  assemblages  formed  at  tem- 
peratures above  480°  C  may  contain  either, 
but  not  both,  of  these  minerals. 

THE  CoAs2-NiAs2-FeAs2-As  SYSTEM 
E.  H.  Roseboom,  Jr. 

In  last  year's  report  the  preliminary  re- 
sults of  a  reconnaissance  in  the  system 
CoAs2-NiAs2-FeAs2-As  at  800°  C  were 
presented.  The  work  described  here  is  a 
continuation  of  that  study. 

The  arsenides  in  this  system  that  oc- 
cur in  natural  deposits  are  skutterudite 
[(Co,Ni,Fe)As3-#],  the  polymorphs  ram- 
melsbergite  and  pararammelsbergite 
(NiAs2),  loellingite  (FeAs2),  and  safflor- 
ite  (Coi,Fei)As2.  Analyses  of  these  min- 
erals indicate  varying  amounts  of  Co,  Ni, 
and  Fe  substituting  for  one  another.  The 
compositions  of  skutterudites  were  dis- 
cussed previously.  The  compositions  of 
diarsenides  will  be  considered  below. 

The  diarsenides  in  the  Co-Fe  and  Fe-Ni 
series  were  made  by  heating  mixtures  of 
Co,  Ni,  Fe,  and  As  metals  of  the  desired 
composition  in  sealed,  evacuated  silica  glass 
tubes  at  800°  C  for  60  hours.  The  diar- 
senides in  the  Co-Ni  series  were  made  by 
heating  mixtures  of  the  cobalt  diarsenide 
phase  and  rammelsbergite.  When  native 
metals  were  used  for  this  series,  cobalt 
skutterudite  and  niccolite  (NiAs)  formed 
together  with  a  Co-Ni  diarsenide.  After 
several  regrindings  and  reheatings,  the  co- 
balt skutterudite  and  niccolite  would  first 

diminish  and  then  disappear,  leaving  the 
Co-Ni  diarsenide.  When  end-member 
phases  were  used  some  skutterudite  and 
niccolite  formed,  but  they  disappeared 
after  one  or  two  regrindings. 

Figure  37  shows  the  cell  dimensions  as 
calculated  from  the  111,  210,  101,  and  120 
X-ray  diffraction  peaks  in  the  Fe-Co,  Co- 
Ni,  and  Ni-Fe  diarsenides.  The  orienta- 
tion is  the  revised  one  suggested  by  Buer- 
ger (1937)  for  minerals  of  the  marcasite 
structure.  The  first  two  series  form  com- 
plete solid  solutions.  The  Ni-Fe  series  is 
interrupted  by  a  two-phase  field  in  which 
a  Ni94Fe6  diarsenide  is  stable  with  a 
Ni7oFe3o  diarsenide.  The  X-ray  diffrac- 
tion patterns  of  samples  with  composi- 
tions lying  in  this  two-phase  field  contain 
the  peaks  of  both  the  above  diarsenides, 
but  the  relative  intensities  vary  with  the 
composition.  A  sample  originally  heated  at 
800°  C  was  divided  and  reheated  at  dif- 
ferent temperatures.  At  850°  C  after  7 
days,  the  two  coexisting  diarsenides  had  d 
values  corresponding  to  Ni92Fe8  and 
Ni75Fe25.  At  750°  C  after  18  days  the  two 
diarsenides  had  d  values  corresponding  to 
Ni94Fe6  and  NierFess.  At  700°  and  600°  C 
after  3l/2  months  the  peaks  of  the  iron-rich 
diarsenide  were  too  weak  to  measure 
and  the  nickel-rich  diarsenide  was  about 
NigsFe*  in  both  cases.  Thus,  there  ap- 
pears to  be  a  highly  asymmetrical  solvus 
limiting  the  extent  of  solid  solution  in  this 

This  solvus  extends  into  the  ternary 
compositions  to  at  least  NicFeiCoi  at  800° 
C  because  a  sample  of  that  composition 
produced  two  diarsenide  phases  with  d 
values  similar  to  those  in  the  Ni-Fe  series. 
A  sample  with  equal  amounts  of  Co,  Ni, 
and  Fe  and  three  samples  along  the  70 
per  cent  Fe,  30  per  cent  (Co  +  Ni)  line 
gave  homogeneous  diarsenides. 

The  cobalt-rich  diarsenide  phase  is  not 
orthorhombic  like  rammelsbergite  and 
loellingite.  The  101  and  111  peaks,  when 
traced  through  the  solid  solution  series 
toward  CoAs2,  split  into  two  peaks  of 
approximately  equal  intensity,  suggesting 



that  the  cobalt-rich  diarsenides  are  mono- 

One  should  be  cautious  about  determin- 
ing the  composition  of  natural  diarsenides 
from  the  curves  on  figure  37,  because  small 
amounts  of  other  atoms  substituting  for 
Co,  Ni,  Fe,  or  As  may  affect  the  cell  di- 
mensions. Five  unanalyzed  safflorites  were 
X-rayed,  and  their  d  values  indicated  corn- 

Other  substitutions  may  cause  changes 
in  the  cell  dimensions.  Neumann,  Heier, 
and  Hartley  (1955),  who  measured  the  d 
values  of  several  natural  loellingites  con- 
taining sulfur,  found  that  increasing  sulfur 
content  causes  the  difference  between  the 
120  and  101  d  values  and  between  the  210 
and  111  d  values  to  decrease.  Sulfur  atoms 
substituting  for  about  9  per  cent  of  the 


Fig.  37.   The  relationships  between  composition  and  cell  dimensions  in  the  diarsenides  of  Co,  Fe, 
and  Ni. 

positions  ranging  from  about  Co45Fe55  to 
Cos5Fei5,  according  to  figure  37. 

The  experimental  work  revealed  no  sys- 
tematic change  in  the  cell  dimensions  with 
arsenic  content  of  loellingite.  The  111,  210, 
101,  and  120  d  values  of  synthetic  loel- 
lingites with  the  compositions  FeAs2.oo, 
FeAsi.99,  FeAsi.98,  and  a  loellingite  with 
a  trace  of  FeAs  in  a  sample  with  the  com- 
position FeAsi.97  were  measured,  using 
quartz  as  an  internal  standard.  The  widest 
range  of  values  for  a  single  d  value  was 
0.0019  A,  and  there  were  no  systematic 
changes  with  composition. 

arsenic  atoms  in  the  most  sulfur-rich  loel- 
lingite resulted  in  a  difference  of  0.04  A 
between  the  first  pair  of  d  values,  and  be- 
tween the  second  pair  a  difference  of  0.05 
A.  These  values  for  the  differences  cor- 
respond approximately  to  the  substitution 
of  30  per  cent  of  the  Fe  by  Co  or  about  15 
per  cent  of  the  Fe  by  Ni.  Although  the 
differences  are  the  same  in  these  cases,  the 
absolute  values  of  the  d  values  may  be 
different  for  the  different  substitutions. 
L.  A.  Clark  is  studying  the  sulfur  content 
of  loellingites  in  the  system  Fe-As-S. 
Figure  37  also  shows  the  cell  volumes  of 


the  Co,  Ni,  and  Fe  diarsenides.  They  in- 
crease in  the  order  FeAs2-CoAs2-NiAs2. 
The  change  in  cell  volume  is  essentially 
linear.  Of  the  unit  cell  dimensions,  only 
the  c  axis  follows  this  order  of  increasing 
size.  The  a  and  b  axes  do  just  the  opposite, 
decreasing  in  the  order  FeAs2-CoAs2- 
NiAs2.  Buerger  (1937)  and  Rosenqvist 
(1953)  have  recognized  two  distinct  divi- 
sions of  compounds  with  the  marcasite 
structure,  a  marcasite  group  and  a  loel- 
lingite  group.  The  marcasite  group  has  a 
relatively  longer  c  axis  and  a  larger  c/b 
ratio  than  the  loellingite  group,  suggest- 
ing a  difference  in  the  bonding  between 
the  atoms.  The  Ni-Co  and  Co-Fe  series 
thus  bridge  the  gap  between  these  two 
orthorhombic  groups  with  a  monoclinic 
phase,  cobalt-rich  diarsenide.  In  the  Ni-Fe 
series  the  two  groups  are  separated  by  a 
two-phase  region. 

G.  Kullerud  and  H.  S.  Yoder,  Jr. 

The  majority  of  sulfide  ores  are  gen- 
erally held  to  have  been  deposited  from 
aqueous  solutions.  Numerous  experiments 
show  that  the  solubility  of  the  sulfides  in 
water  is  very  small  at  low  temperatures 
and  pressures.  The  solubility  curve  leads, 
in  a  continuous  or  discontinuous  fashion, 
to  the  liquidus  curve  extending  to  the  de- 
composition of  the  sulfide  itself.  Thus,  in 
the  case  of  ore  deposits  associated  with 
magmas,  it  is  conceivable  that  the  sulfides 
are  deposited  from  a  concentrated  solution 
as  suggested  by  Spurr. 

Accordingly,  we  have  made  preliminary 
investigations  of  the  influence  of  water  at 
high  temperature  and  pressure  on  the  melt- 
ing points  of  several  sulfides.  It  was  con- 
venient to  study  these  sulfides  in  both 
rigid  silica  glass  tubes  and  collapsible  gold 
tubes.  Runs  up  to  950°  C  and  2000  bars 
H20  pressure  with  galena  (PbS)  in  gold 
tubes  failed  to  produce  any  melting.  It 
was  observed  that  galena  at  this  tempera- 
ture and  pressure  was  transported  through 
the  wall  of  the  gold  tube  along  intergranu- 

lar  boundaries  and  deposited  on  the  ex- 
terior. The  melting  point  of  PbS  was 
determined  as  1130°  ±5°  C  (Kracek,  1952) 
under  its  own  vapor  pressure  in  silica 
tubes.  Thus  our  experiments  set  a  limit 
on  the  possible  lowering  of  the  melting  of 
PbS  under  these  pressures. 

The  melting  point  of  acanthite,  Ag2S, 
was  found  to  be  837°  ±5°  C  under  its  own 
vapor  pressure  in  silica  tubes.  Melting  of 
acanthite  occurred  at  considerably  lower 
temperatures  when  the  charge  was  held 
in  an  open  gold  tube  inside  a  sealed  silica 
glass  tube,  presumably  because  of  reac- 
tions taking  place  between  the  gold  tubing 
and  Ag2S. 

Since  it  has  been  shown  that  gold  does 
not  measurably  influence  the  stability  rela- 
tions of  the  Fe-S  system  above  300°  C 
(Kullerud  and  Yoder,  Year  Book  55,  p. 
181),  experiments  with  pyrite,  FeS2,  and 
water  were  attempted  in  gold  tubes.  The 
breakdown  temperature  of  pyrite  is  low- 
ered by  the  presence  of  water.  Preliminary 
experiments  indicate  that  the  lowering  is 
in  excess  of  30°  C  at  2000  bars  H20  pres- 
sure. Pyrrhotite  and  a  sulfur-rich  gas  co- 
exist with  pyrite  in  its  stability  field. 
Only  pyrrhotite  and  a  sulfur-rich  gas  were 
observed  when  pyrite  was  completely 

H.  L.  Barnes 

Theoretical.  Available  results  of  labora- 
tory studies  of  mineral  assemblages  and 
their  vapor  pressures  under  anhydrous  con- 
ditions can  be  applied  thermodynamically 
to  place  useful  limitations  on  the  composi- 
tion of  ore  solutions.  Kullerud  (Year 
Book  56)  determined  the  phase  relations 
in  the  Fe-S-O  system,  which  contains  the 
almost  ubiquitous  mineral  assemblage  py- 
rite (FeS2),  pyrrhotite  (Fei-^S),  and  mag- 
netite (Fe304).  The  common  occurrence 
of  these  minerals  in  a  variety  of  types  of 
ore  deposits  makes  conclusions  based  on 
this  assemblage  of  widespread  significance. 
These  minerals,  when  coexisting  in  equi- 


librium  at  any  given  temperature,  fix  the 
partial  pressures  of  sulfur  (Pstot)  and  oxy- 
gen (Po2).  Kullerud  and  Yoder  (1958) 
have  summarized  the  vapor-pressure  data 
for  the  pyrite-pyrrhotite  assemblage.  Muan 
(1958)  has  summarized  the  data  and  given 
T  versus  Po2  curves  for  the  univariant  as- 
semblages wustite  (FeO)  and  magnetite -f- 
vapor,  iron  and  magnetite  +  vapor,  and 
hematite  (Fe2Os)  and  magnetite  +  vapor. 
These  curves  define,  at  any  specific  tem- 
perature, the  range  of  Po2  that  can  occur 
with  magnetite  as  a  stable  phase.  The  di- 
rect measurement  of  these  vapor  pressures 
has  been  possible  only  at  high  temperatures, 
but  the  curves  may  be  extrapolated  below 
the  critical  temperature  of  water  (374°  C) 
with  sufficient  accuracy  for  the  present 
purpose.  The  vapor  over  the  pyrite-pyr- 
rhotite pair  is  predominantly  sulfur  vapor 
composed  of  the  molecules  S2,  S6,  S8,  etc. 
Because  the  distribution  of  the  molecular 
species  is  not  known  under  these  condi- 
tions, the  range  of  possible  values  for  Ps2 
used  in  the  extrapolation  from  the  meas- 
ured vapor  pressures  includes  this  uncer- 

The  minimum  temperature  at  which  the 
pyrite-pyrrhotite-magnetite  assemblage  is 
formed  in  ore  deposits  is  not  known,  but 
ore  bodies  where  sphalerite  occurred  in 
coexistence  with  this  assemblage  have  been 
estimated  to  have  formed  at  temperatures 
as  low  as  about  200°  C.  The  pyrite-- 
pyrrhotite +  magnetite  assemblage  is  prob- 
ably not  formed  much  below  this  tempera- 
ture; the  partial  pressures  for  this  assem- 
blage are  shown  in  figures  38A,  39,  and 
40  only  to  illustrate  the  changes  in  the 
position  of  these  vapor  pressure  fields  with 
T  and  concentration. 

The  ionic  character  of  the  ore  solution 
is  a  function  of  the  Rwe  predominant  vari- 
ables: temperature,  T;  total  pressure,  Ptot; 
total  concentration  of  sulfur-containing 
ions,  (Stot);  Po2;  and  pH.  Bulk  composi- 
tion, which  includes  the  variables  (Stot), 
Po2,  and  pYL,  is,  by  itself,  inadequate  for 
describing  the  ionic  species  in  the  aqueous 

phase.  Thermodynamic  constants  can  be 
used  to  calculate  the  activities  of  major 
sulfur-containing  aqueous  ions  as  functions 
of  Po2  and  pH  if  (Stot)  (including  solids), 
as  well  as  T  and  Ptot,  is  constant.  The  re- 
sults of  such  calculations,  which  are  of 
classical  thermodynamic  types,  are  illus- 
trated graphically  in  figure  38.  The  areas 
shown  represent  regions  of  predominance 
of  each  of  the  major  ionic  species  in  the 
system.  The  lines  limiting  each  of  these 
areas  indicate  equal  activities  of  the  pre- 
dominant ions  of  the  adjacent  areas.  Ac- 
tivity equals  concentration  within  one  or- 
der of  magnitude  for  concentrations  up  to 
about  0.1  mole/liter;  therefore,  for  clarity, 
concentration  will  be  used  here  instead  of 

Although  Ptot  is  taken  as  1  atm  at  25° 
C  in  the  calculations,  inert  pressures  of 
several  hundred  atmospheres  are  not  ex- 
pected to  change  the  ionic  stabilities  be- 
yond the  limit  of  error  in  the  calculations 
(<10°-5).  If  Ptot  is  neglected,  a  series  of 
these  diagrams  for  two  or  more  values  of 
(Stot)  at  several  temperatures  is  necessary 
to  outline  the  ionic  behavior  in  the  range 
of  interest  to  this  study. 

The  effect  of  changing  (Stot)  by  a  fac- 
tor of  100  is  illustrated  by  comparison  of 
figures  38^4  and  39.  Figures  38^4,  40,  and 
41  show  the  change  in  ionic  stabilities  and 
partial  pressures  with  increasing  tempera- 
ture calculated  from  the  van  't  Hofif  equa- 
tion. Although  there  are  no  applicable 
experimental  data  on  ions  in  this  system 
above  25°  C,  other  studies  (Harned  and 
Owen,  1943)  indicate  that  AH  can  be 
assumed  constant  as  a  first  approximation 
in  this  equation.  The  error  in  this  assump- 
tion increases  with  temperature,  and  at 
250°  C  in  figure  41  only  a  rough  approxi- 
mation can  be  given  of  the  areas  of 
predominance  of  the  various  ions.  Never- 
theless, the  limits  placed  on  the  partial 
pressures  in  the  aqueous  system  by  the 
solid  phases  of  the  Fe-S-O  system  in  turn 
narrow  considerably  the  aqueous  region 
of  interest. 

Fig.  3SA 

-10     -20  0 

Log  (o, ) 

Fig.  38£ 


]          1          1          1 


1      1      1      1 

1      1      1 


1           1 

s*  _ 

__HST . 








£  -20 

'      = 


J    -30 

— ^- 


-4 "1            1             1 

---  ^7» |o"90 

1       1       1       1 

1           1           1 


I  2         3         4         5         6         7         8         9         10        II         12        13 


Fig.  38C 



Fig.  41 

Fig.  38.  St  =  0.1  mole/liter;  T  =  25°  C.  A. 
Fields  of  stabilities.  B.  Activities  of  minor  ions 
at  constant  pH.  C.  Activities  of  minor  ions  at 
constant  Po2. 

Figs.  38-41.  Fields  of  stabilities  of  predominant  ions  with  partial  pressures  of  S2  (contoured  in 
atmospheres)  in  equilibrium  with  water  at  fixed  temperature  and  total  concentration  of  sulfur- 
containing  ions.    See  text  for  discussion  of  crosshatched  areas. 

Fig.  39.    ST  =  0.001  mole/liter;  T  =  25°  C. 
Fig.  40.    ST  =  0.1  mole/liter;  T  =  100°  C. 
Fig.  41.    ST  =  0.1  mole/liter;  T  =  250°  C. 



The  pH  boundaries  of  this  region  can- 
not be  calculated  at  present,  but  it  can  be 
assumed  that  during  ore  transport  the  pW 
will  not  deviate  from  neutral  by  more  than 
2  /?H  units.  Variations  beyond  these  limits 
are  unlikely  because  of  the  buffering  ac- 
tion of  wall  rocks  on  the  ore  solutions. 
Furthermore,  measurements  of  the  pH 
in  pertinent  hot  springs,  and  fluid  inclu- 
sions in  ore  minerals,  lie  within  these 

The  ore-transporting  region  shown  in 
figures  3SA  and  39-41  corresponds  to  the 
aqueous  conditions  where  the  strongly 
complexing  ions  polysulfide  (Sa?=)  (fig. 
38C)  and  thiosulfate  (S203=)  (fig.  38B) 
reach  maximum  concentrations.  These  are 
the  most  probable  complexing  ions  for 
the  metals  in  the  ore  solution,  as  shown 
by  independent  evidence  from  zoning 
(Barnes,  1956)  and  by  the  maximum  solu- 
bilities of  metallic  complexes  in  such  solu- 
tions. The  three  lines  of  evidence  outline 
an  aqueous  region  in  which  the  ionic  char- 
acter of  the  sulfur-containing  species  needs 
systematic  experimental  study. 

In  figures  38^4  and  39-41  the  assemblage 
pyrite-pyrrhotite-magnetite  can  coexist 
with  water  at  each  (Stot)  only  where  the 
partial  pressure  fields  overlap  as  shown  by 
cross  hatching.  The  lack  of  overlap  of  the 
partial  pressure  fields  fixes  an  upper  limit 
to  (Stot)  in  the  aqueous  phase.  As  seen 
by  comparison  of  figures  38^4  and  39,  in- 
creasing (Stot)  raises  Ps2  and  spreads  the 
contours  away  from  the  sulfur  field.  In 
order  to  have  appreciable  overlap  of  these 
fields,  in  figure  41,  (Stot)  must  be  less  than 
1.0  mole/liter  at  250°  C.  In  figure  40,  (Stot) 
must  be  less  than  0.01  mole/liter  at  100° 
C.  An  ore-transporting  solution  capable 
of  depositing  the  assemblage  pyrite-pyr- 
rhotite-magnetite (hence  in  equilibrium 
with  these  minerals)  cannot  have  a  total 
concentration  of  ions  containing  only  S, 
H,  and  O  greater  than  the  above  limits  at 
each  temperature.  Compounds  or  ions  con- 
taining both  metals  and  sulfur  are  excluded 
from  this  maximum. 

When  the  stabilities  and  vapor  pressures 
of  other  assemblages  of  ore  minerals  are 
accurately  determined,  it  will  be  possible 
to  ascertain  the  specific  composition  of 
sulfur-containing  ions  in  the  ore  solution 
at  any  predetermined  temperature  for  indi- 
vidual ore  deposits. 

Experimental.  Apparatus  (fig.  42)  and 
experimental  technique  have  been  de- 
veloped for  measuring  solubilities  under 
the  temperature  and  pressure  conditions 
prevailing  during  the  formation  of  many 
ore  deposits.  This  apparatus  is  now  in 
use  in  a  study  of  the  solubility  of  sphalerite 
(ZnS)  in  H2S-saturated  water. 

The  solubility  in  each  sample  in  this 
system  is  determined  polarographically. 
Preliminary  results  show  that  solubilities 
near  100°  C  and  500  psi  exceed  10  mg/1 
ZnS,  which  is  more  than  106  times  greater 
than  the  calculated  solubility  product  in 
pure  water  at  this  temperature,  but  at  25° 
C  and  about  100  psi  the  solubility  is  not 
detectable  (<1  mg/1  ZnS)  with  the  pres- 
ent analytical  technique.  The  solubility 
is  controlled  by  the  formation  of  complex 
ions  probably  of  the  type  ZnS*xH2S.  The 
magnitude  of  the  solubility  indicates  a  sur- 
prisingly great  stability  for  the  simple  sul- 
fide complexes  of  zinc,  which,  however, 
are  probably  not  predominant  during  ore 
transport.  Only  by  a  fundamental  under- 
standing of  the  chemistry  of  this  rela- 
tively simple  three-component  system  can 
the  behavior  of  the  four-component  sys- 
tems containing  the  complex  ions,  poly- 
sulfide and  thiosulfate,  be  interpreted. 

Solutions  involved  in  ore  transport  are 
highly  corrosive;  therefore,  the  reaction 
vessel  (internally  plated  with  chromium), 
tubing,  and  valves  are  of  stainless  steel. 
Slight  corrosion  of  the  chamber  has  been 
observed  up  to  160°  C  and  1000  psi  by  runs 
in  the  system  ZnS-H2S-H20. 

Although  there  are  several  methods  of 
measuring  solubilities  at  moderate  pres- 
sures, the  use  of  double  valves  as  described 
below  is  believed  to  give  the  most  quanti- 
tative results.  The  entire  assembly  shown 



in  figure  42  is  agitated  to  hasten  equilib- 
rium by  tilting  through  an  arc  of  30° 
about  the  horizontal  rocker  axis.  Connec- 
tions to  valves  2  and  4  are  made  with  a 
flexible  spiral  coil  of  tubing  wound  about 
the  rocker  axis. 

Temperature  is  regulated  by  a  bridge 
circuit  in  which  the  variation  of  the  re- 
sistances of  the  nickel  furnace  windings 
controls  the  voltages  supplied  to  furnaces 
A  and  B.  The  temperature  is  measured 
and  recorded  by  thermocouples  inserted 
in  the  wells  shown  in  the  figure.  Pressure 

after  flushing  the  system  with  N2,  a 
vacuum  of  less  than  200  u  is  attained  by 
pumping  with  only  valve  4  closed.  Next 
valve  2  is  closed  and  the  valve  on  the  gas 
cylinder  opened  until  pressure  in  the  reac- 
tion chamber  indicates  a  calculated  gas 
content.  The  weight  of  gas  in  the  ap- 
paratus is  determined  by  reweighing  the 
gas  cylinder.  With  valves  2  and  3  closed, 
deaerated  water  is  pumped  from  a  weighed 
amount  in  a  calibrated  reservoir  (flushing 
the  gas  in  the  sampling  tube  ahead) 
through  an  ion-exchange  column  into  the 

Rocker  oxis 

Capillary   tubing 
to  pressure  gage 

Thermocouple  well 

Guide  pin 


To  ion  exchange 
column  and 
water  pump 

8   inches 

To  vacuum  pump  and 
nitrogen  tank 

Note:  The  closure  and  certain 
other  areas  have  been 
distorted   for  clarity 


aisiortea  tor  clarity 

Fig.  42.   Apparatus  for  sampling  chemical  systems  at  elevated  pressure  and  temperature  in  order 
determine  solubilities  of  minerals. 

is  measured  directly  with  a  stainless-steel 
bourdon  gage  through  a  U  tube  filled  with 
paraffin  oil  which  chemically  isolates  the 
gage  from  the  system.  Paraffin  oil  does 
not  react  with  runs  in  the  system  ZnS- 
H2S-H2O  if  the  U  tube  is  kept  near  room 

The  procedure  for  each  run  is  as  follows  : 
After  thorough  cleaning,  the  seal  of  the 
reaction  vessel  is  tested  with  argon  and 
water  to  avoid  leaks.  Without  breaking 
the  gasket  seal,  the  chamber  is  emptied 
through  the  tubing  ports  and  evacuated  to 
<1  mm  Hg  to  assure  a  dry  interior.  A 
weighed  amount  of  solid  is  inserted 
through  a  tubing  port,  and  the  entire  ap- 
paratus is  assembled.  A  weighed  cylinder 
of  gas  is  then  connected  to  valve  3  and, 

reaction  chamber  to  the  desired  volume. 
Valves  1  and  4  are  then  closed,  and  the 
water  in  the  sampling  tube  is  drained 
through  valve  3  (and  flushed  with  N2 
from  valve  2)  and  combined  with  that  in 
the  reservoir  to  determine  the  weight  loss 
and  the  weight  of  water  in  the  reaction 
chamber.  Thus,  weighed  amounts  of 
solids,  liquids,  and  a  gas  can  be  inserted 
into  the  reaction  chamber  without  con- 

After  the  apparatus  has  been  brought  to 
temperature  and  agitated  until  equilibrium 
is  assured,  a  sample  may  be  withdrawn 
simply  by  evacuating  the  sample  tube 
through  valve  2  with  valves  1,  3,  and  4 
closed,  opening  valve  1  momentarily  with 
valves  2,  3,  and  4  closed,  and  then  flushing 


the  sample  through  valve  3  with  N2  from 
valve  2.  Several  samples  at  the  same  or 
different  temperatures  can  be  taken  from 
each  charge  in  this  manner  because  vol- 
ume of  sample  is  small  compared  with  that 
of  the  main  vessel. 

The  sample  is  believed  to  be  representa- 
tive because  the  sample  tube  is  maintained 
near  the  same  temperature  as  the  reaction 
chamber  and  any  precipitate  formed  in 
the  reaction  vessel  near  the  sample  tube 
from    the    instantaneous    pressure    drop 

would  be  flushed  into  the  sample  as  the 
tube  fills.  N2  from  valve  2  cleans  any 
precipitate  from  the  sample  tube.  The 
sample  tube  is  0.45  per  cent  of  the  volume 
of  the  reaction  chamber;  therefore,  in  the 
system  Z11S-H2S-H2O,  the  observed  pres- 
sure drop  on  sampling  is  ~0.5  per  cent. 

The  apparatus  is  suitable  for  quantita- 
tive solubility  measurements  in  a  variety 
of  chemical  systems  limited  only  by  corro- 
sion and  the  mechanical  strength  of  the 
reaction  vessel. 


F.  Chayes 

Subtle  diffraction  effects — peak  shifts, 
changes  of  intensity,  appearance  or  disap- 
pearance of  weak  reflections — are  proving 
of  great  value  in  silicate  phase-equilibria 
studies,  particularly  in  the  delineation  of 
subsolidus  reactions.  It  is  becoming  in- 
creasingly popular  to  regard  such  effects 
as  consequences  of  ordering  (or  disorder- 
ing). Usually  no  geometrical  model  of  the 
ordering  process  is  presented,  and  in  the 
rare  exceptions  the  ordering  involved 
seems  to  be  of  the  "long-range"  variety, 
which  generates  additional  repeats  or  su- 
perperiodicity  along  one  or  more  of  the 
principal  directions  of  the  crystal.  The 
possibility  that  "short-range"  ordering,  i.e., 
ordering  concerned  with  nearest-neighbor 
pairings  rather  than  the  development  of  a 
superlattice,  might  generate  characteristic 
diffraction  effects  of  the  types  in  question 
seems  to  have  escaped  notice. 

Experimental  study  of  diffraction  effects 
generated  by  varying  levels  of  short-range 
ordering  in  layered  sequences,  begun  late 
in  1956,  continues  as  a  major  activity.  Most 
of  the  year  has  been  spent  in  refining  and 
improving  experimental  techniques  de- 
veloped toward  the  close  of  the  last  report 
year  and  described  briefly  in  Year  Book  56. 
As  often  happens  in  such  matters,  the 
chronological  order  of  events  is  determined 
by  psychological  rather  than  logical  con- 
siderations, and  makes  little  sense  to  any- 

one but  the  investigator  himself.  The 
year's  progress  is  accordingly  reviewed 
here  in  rational  rather  than  chronological 
sequence,  treating,  in  order,  developments 
affecting  the  design  of  random-layered  se- 
quences characterized  by  some  arbitrary 
level  of  short-range  order,  the  production 
of  experimental  diffraction  masks  based  on 
such  sequences,  the  generation  of  diffrac- 
tion transforms  from  these  masks,  and  the 
direct  calculation  of  the  diffraction  effects 
by  high-speed  computation. 

The  Distribution  of  Run  Lengths 

The  number  of  unbroken  runs  of  iden- 
tical elements  in  a  two-element  sequence 
such  as  ABBBAAB'  •  'determines  the  level 
of  short-range  order.  For  any  possible 
number,  d,  of  runs,  it  may  be  shown  that 
the  probability  that  any  run  chosen  at 
random  is  of  length  i  is: 


where  the  array  contains  Na  A's  and  Nb 
B's.  Since  there  are  to  be  d  runs  of  each 
element,  and  the  probabilities  q(a)i  and 
q{b)i  are  the  same  for  each  run  of  the 
appropriate  element,  the  expected  number 


of  runs  of  length  i  for  each  element  is 
simply  the  product  of  the  total  number 
of  runs  of  the  element  by  the  probability 
that  any  one  of  these  will  be  of  length  i, 

E(dAi)  =  dq  (a)  i  (2a) 

E(dm)  —dq(b) 


The  importance  of  these  relations  is  that 
they  permit  the  construction  of  experi- 
mental— and  therefore  necessarily  rather 
short — sequences  having  any  desired  level 
of  ordering  and  at  the  same  time  exhibit- 
ing relative  frequencies  of  run  lengths  as 
close  as  possible  to  those  required  by 
theory.  The  distribution  of  run  lengths  is 
thus  no  longer  an  experimental  variable. 
The  sequence  in  which  the  runs  of  dif- 
ferent lengths  occur  in  an  actual  mask  is 
randomized  by  a  standard  Monte  Carlo 
sampling  technique,  but,  for  a  mask  of 
any  given  length,  composition,  and  level 
of  ordering,  this  is  the  only  random  vari- 
able. This  is  essentially  the  situation  that 
confronts  us  when  we  examine  the  diffrac- 
tion pattern  of  a  real  crystal  in  which  the 
level  of  ordering  is  not  subject  to  change 
during  the  time  necessary  to  photograph 
the  pattern. 

Generation   of  Diffraction   Mas\s  of 
Layered  Sequences 

In  the  two-dimensional  mask,  the  place 
of  a  layer  is  taken  by  a  row  of  equally 
spaced  holes  in  an  opaque  matrix.  In 
the  work  reported  last  year,  these  were 
punched  individually  on  a  small  instru- 
ment, the  Peek-a-boo  Punch,  designed  by 
the  Basic  Instrumentation  Laboratory  of 
the  National  Bureau  of  Standards.  The 
masks  were  of  excellent  quality,  but  se- 
quences longer  than  150  layers  could  not 
be  managed,  and  even  a  100-layer  mask 
usually  required  more  than  3  hours  of 
machine  time. 

The  instrument  now  used  for  this  pur- 
pose, shown  in  figure  43,  is  an  adaptation 
of  one  described  by  B.  G.  M.  Willis  in 
1957.  It  is  essentially  a  contact  printer,  in 
which  the  photographic  element,  an  8  by 

10  inch  plate,  may  be  displaced  laterally 
between  exposures. 

On  the  stage  of  a  large  dividing  engine 
{A),  a  smaller  one  (B)  is  mounted,  the 
threads  of  the  two  being  normal  to  each 
other.  The  stage  of  the  smaller  engine 
carries  a  trough  (C)  in  which  the  plate- 
holder  (D)  rides.  The  light  box  (E)  is 
mounted  rigidly  to  the  frame  of  the  larger 
dividing  engine,  and  from  it  a  large  light 
shield  (F)  is  spring-loaded  against  the  top 
of  the  track  that  carries  the  plateholder. 
From  inside  the  light  box  a  rectangular 
shoe  containing  the  template  is  spring- 
loaded,  through  a  recess  in  the  light  shield, 
against  the  plate  cover.  Two  100-watt  pro- 
jection bulbs  mounted  in  the  removable 
top  of  the  light  box  (G)  are  activated  by 
the  switch  of  an  exposure  timer  (H). 

The  loaded  plateholder  is  slid  into  its 
track;  the  shoe  containing  the  template 
is  next  lowered  through  the  light  shield 
and  brought  to  rest  squarely  on  the  plate 
cover.  The  cover  of  the  light  box  is  put 
in  place,  making  the  entire  assembly  light- 
tight.  Finally  the  plate  cover  is  withdrawn, 
and  the  template,  which  is  resting  on  it, 
drops  gently  onto  the  plate. 

The  template  now  in  use  is  simply  a 
brass  strip  containing  a  row  of  60  holes  of 
0.5-mm  diameter  spaced  3  mm  apart,  paral- 
lel to  the  screw  of  the  upper  dividing  en- 
gine. Each  exposure  thus  generates  a  row 
of  "scatterers"  which  serves  as  a  two- 
dimensional  projection,  or  model,  of  a 
layer  in  a  layered  structure.  The  distance 
between  layers  is  obtained  by  translation 
of  the  plate  with  the  large  dividing  engine; 
this  interval,  being  constant  in  the  problem 
now  under  consideration,  is  established  by 
means  of  a  detent,  so  that  the  large  drum 
need  be  read  only  as  a  check.  Each  exposure 
is  followed  by  a  rotation  of  the  large  drum; 
the  cross  engine,  however,  is  reset  only 
when  two  successive  layers  are  separated 
by  an  offset  or  "mistake."  After  a  little 
practice,  a  mask  of  200  layers  can  be  com- 
pleted in  a  little  less  than  an  hour.  If  of 
suitable  quality,  the  original  pattern  is 
reduced  by  a  linear  factor  of  not  less  than 
8.    If  an  eightfold  reduction  is  used,  the 


distance  between  layers  (rows)  in  the 
finished  mask  is  0.125  mm  and  the  viewing 
assembly  of  the  diffractometer  described 
below  will  contain  two  full  orders  of  the 
diffraction  pattern.  The  photographic 
work  has  so  far  been  done  entirely  on 
glass  Kodalith  plates  for  the  original  and 
high-contrast  Eastman  lantern  slides  for 
the  reductions.  A  typical  mask  is  shown 
in  figure  45#. 

The  Optical  Diffractometer 

The  very  small  diffractometer  used  to 
obtain  the  transforms  illustrating  last  year's 
report  has  now  been  replaced  by  the  instru- 
ment shown  in  figure  44#;  figure  44£  is  a 
schematic  diagram.  The  image  contained 
in  the  viewing  assembly  is  remarkably 
sharp  and  clear.  Photographs  of  equal 
clarity  can  be  obtained  with  early  versions 
of  the  2-watt  concentrated  (Zr)  arc  lamp 
as  light  source.  The  currently  available 
bulbs  of  this  type  are  subject  to  consider- 
able flicker,  and,  since  exposure  time  is  of 
the  order  of  1  or  2  minutes,  the  photo- 
graphs are  commonly  rather  inferior.  The 
difficulty  is  much  reduced  by  use  of  a 
pinhole,  but  this  greatly  increases  exposure 

A  good  summary  of  the  impact  of  these 
developments  on  the  experimental  work 
may  be  obtained  by  comparing  figure  45£ 
with  figure  4£  (plate  2,  p.  192)  of  Year 
Book  56.  The  composition  and  level  of 
ordering  of  the  masks  yielding  these 
transforms  are  the  same. 

The  Calculation  of  Intensity  Profiles 

The  number  of  (atomic)  layers  in  a 
crystal  of  a  layered  mineral  is  ordinarily 
very  large,  and  a  considerable  increase  in 
mask  length  would  therefore  be  desirable. 
With  the  procedure  reviewed  above,  it  is 
possible  to  design  sequences  of  any  length, 
composition,  and  level  of  ordering.  Al- 
though in  principle  the  instrument  de- 
scribed above  could  be  made  to  generate 
very  long  masks,  the  practical  maximum 
seems  to  be  of  the  order  of  400.  And,  if 
a  mask  of  even  this  length  is  reduced  suf- 

ficiently to  permit  complete  illumination 
in  the  central  region  of  the  collimated 
beam  of  the  diffractometer,  the  viewing 
assembly  will  not  contain  a  full  order  of 
the  diffraction  pattern.  The  desired  exten- 
sion of  mask  length  thus  requires  either 
rather  fundamental  improvements  in  op- 
tical equipment  or  an  outright  transfer 
of  the  work  from  optical  to  numerical  ex- 
perimentation. Ultimately,  it  may  be  use- 
ful to  explore  both  possibilities,  but  at  the 
moment  full  attention  is  being  given  the 

In  a  mask  like  figure  45#  all  the  rows 
are  of  identical  scattering  power,  and  each 
row  is  either  in  register  with  the  first  row 
or  offset  from  it  by  a  fixed  proportion  (t) 
of  the  distance  between  points.  In  the  ex- 
perimental work  described  here  t=l/2 
and  the  x  coordinate  of  the  first  hole  in 
each  row  is  thus  0  or  l/2.  The  y  coordinate 
is  (n  —  l)/N,  where  n  is  the  number  of 
the  row  and  N  the  number  of  rows  in 
the  mask.  The  intensity  of  the  diffracted 
beam  of  order  {h\)  is  then  given  by  the 
Fourier  transform 


I  N  \ 

\l       > 


sin  2n(hxn  +  ^yn) 





Z  cos  2n(hxn  +  \yn) 



It  is  implicit  in  the  definitions  of  sine 
and  cosine  that,  for  any  h,  hk  =  IuN+k); 
i.e.,  the  transform  is  periodic  along  \  with 
a  repeat  of  N.  It  may  also  be  shown  that, 
ior0<{  (integral)  <N, 

hk  —  h(N-k)  =  4  sin  2nht' 


N  N 

I  cos  Bt  (n)  Z  Bo  (n) 

1  1 


Z  cos  Bo  ( 


Z  sin  Bt  (n)  ' 


0        (4) 

9  Derivation  of  equations  4  and  5  is  straight- 
forward but  tedious.  Details  will  be  presented  in 
journal  publication. 

Plate  1 

Geophysical  Laboratory 

Fig.  43.    Mask  generator.    For  explanation  of  letters  see  text. 



//  ,«■ 










Fig.  44.    Optical  diffractometer.    (a)    Photograph  of  the  instrument.    (Focal  length   of  lenses   is 
45  cm.)    (b)   Schematic  diagram  identifying  components. 

Plate  2 

Geophysical  Laboratory 


iTl'I'i'i'lTl  1 1 1 1  1 1 1  III  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1.I.I.IJ.II.I.IJ.IJ.IJ.U.I.I.U.U  . 

i* **********  jVjVjVj'jVj'jV  ***'»'  f  ***** *':'*'****:'  ****«*t*:*s"**:*x"*"t"t*x"x*s***»^t":":^*^x"«**"«"t*j*»^«^»* 


t"t***»*x*i"i*t*t*x*x"x*i"x*x"t"i***x*x*x***x*x"i*x*i"x"**t*»*»*j"**i*s**'s"i"i"x"x  t"x  x.x.x  *  :  :  :  J.J.i.J.i.J.J.i.i. 


i*»***»*x*x"x*****x"x***x***x"x"x*x*****x  :'::  :s  i  »  ::::::::::  s:  s::::::::::::  i  i  ::  t  J  : 

rrrrri"i"i"ri"i"i"i"ri"i"i"rri"i"i"i*i"i"i"i*i*i"i"i*i!i*i"i"i"i"i*i*i"i"i!i"iii  j  j  .i  .•  j  .i.i.  j  .*  j  .i  j  .1.1.  *. 

I.  J. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.  I.I.I.  1. 1. 1. 1. 1. 1. 1.1. 1. 1. 1. 1. 1 .  I.J.I.  *.  '.'.". '.*.'.  '.'.'.".'.*.*.*.".*.".*.*.*.  ,.,.,.».*. 




rriYiYi*ri"i*i*rri"i*ri  ij'jirrijjjjjjjjjjjjj'.ij.i.i.i.i.r.r.r.r.i.rrri.rijj.o  o*i 

i*j*j)j*j*j*j*ri*rrrrri*ri*i^i[j|j|j  j  j  j*i|i[j*i|j*i*i*i*t*i'i"t*j*i"j[i*i)jj|*.i[j[j|*|j[i]i^j"j  j*j*j*j*j*j* 
1*1*1*1  i*i*i*i*t  i*i*i*i*i*i*$*t*i  t  i  i  i  t  «"r«"i*t  i*i  i  i  i  i  i  i  i  i  i  i  »  i  i"i"«  i"i"i"i"i*i*i"i*t*i*i*i*i  i  i* 

1 1 1 1  i  I  I'i  1 1  i  i  i'i  i  i  i  I  i  i  i  M  i'i ! !  i  i'i  i'i  i !  H  I  i  i  i  i'i  I  i  1 1 1 1  l"l"l"f  lllliili 

I  1*1  lYlYlYlYiYi  i  I  !  i  I  i  !.!,!.!  !  i  !  !  I  !  i.l.i  1  I  i  !  !  !  i  i  i  !  !  i'i'l "l  lYl'l  iYl  I  lYlY 


j.jVjVjV* ** ********** "* ********** *t"i*}.i*iit*i*J j'l "*"*"* "i"t*t*x***x^x*x"x***x***x"x*x*x**"x^x*x"x"*"******x*x^ 
t\ t] J* J j* 1*1" i'i" ** s° :" :° t*:":*s*:*s*:I:":"i*:":* :Vi* :*s*»I OX'! !!:!!!:!!*!!*!:!*°!!*"s°:" s's's**!**!*** i* x*x*x*x* 

I  j'l'i't"  *'***':'*':] :':':':':':':':':':':":]:":":':':':':':":' :':":' :':':' :':":' :*:':" :':":"  :***:*:***:*********  ****** 

".•.U.'.'.'.'.U.U.'.'.U  U.U.UJ.'.'.'.U.'.U.'.U.U  :':  :  :  5.5  5  i'i  lYiYlYlYlYlYlYl" 
■.•.,.*.*.*.'.:.:.!.'.:.:.:.:.:.:.:.:.:.!.s.:.!.s.!.!.!.!.'.!.'.!.s.!.!.!  ::::::  i  J  J  :  :"i*J* ************ j*j*j*j" 
,.,.,.,.".,.*.".".*.*.*.!.!.!  '•  '•  :.:.!.!.!.:.:.:.:.!.!.!.!.,.:.s  :::::::::  :":":V*V:V:Vj*j*j*j*j*j*jV 


l.l.l.l.i.i.i.i.:.-.i.I.£.I.-.i.t.i.:.I.i.s.-.£.-.i.-.s.-.t.i  i.i.i.J.i  5.5.5  5.5  5  :  s  5  5  5  5  5  i  i  5  iYiYlYlY 



Fig.  45.     (a)  A  short-range  ordered  2  :  1  sequence,  N=  192.    (b)  Transform  of  (a).    (Compare 
with  fig.  46,  plate  2,  Year  Book  56.) 


where  B(n)  =  (2n{/N)(n-l),  the  sum- 
mations to  be  carried  through  in  each  case 
only  for  the  layers  having  the  x  value  indi- 
cated by  subscript. 

Finally,   it  may   be  shown  that,   for   /( 
integral,  equation  3  may  be  written 

Ink  =  2(l  —  cos  Inht) 


(     I  cos  Bo  0) 
LsmBo(n)        J      ( 

in  which  form  it  is  apparent  that  if  t=l/2, 
as  in  the  experimental  work  here  described, 
Ink  will  be  the  same  for  all  h  odd  and  0 
for  all  h  even. 
The  calculations  which  are  to  evaluate 

the  influence  of  short-range  ordering  on 
"diffuse"  scatter  may  thus  consist  merely 
of  the  computation  of  hk  in  the  region 
0<{<N/2  (instead  of  0<{<N,  as  in 
equation  3)  and  each  hk  value  will  require 
the  generation  of  only  N() — or  Nt — sines 
and  cosines  (instead  of  N,  as  in  equation 
3).9  At  the  present  writing  the  trial  and 
testing  stage  of  the  programing  is  nearing 
completion,  but  systematic  computation 
has  not  begun. 

Perhaps  the  best  summary  of  this  report 
is  to  say  that  the  year  has  been  spent  pre- 
paring to  examine,  and  perhaps  to  solve, 
some  of  the  problems  posed  by  the  work 
reported  last  year.  The  preparation  is 
about  completed. 



Magnetic  Structure  of  Chalcopyrite 

G.  Donnay   et  al. 

On  December  16,  1957,  the  California 
Institute  of  Technology  celebrated  the 
fortieth  anniversary  of  the  publication  of 
the  first  crystal  structure  to  have  been  de- 
termined in  the  United  States,  that  of 
chalcopyrite,  CuFeS2.  (Contribution  No.  3 
from  the  Chemical  Laboratories  of  the 
Throop  College  of  Technology.)  It  seems 
appropriate,  at  this  time,  to  recall  that  this 
pioneering  work  was  made  possible  by  a 
grant  of  the  Carnegie  Institution  of  Wash- 
ington to  A.  A.  Noyes,  at  whose  suggestion 
the  research  was  carried  out.  The  support 
which  the  Institution  has  given  to  X-ray 
crystallography  and  the  interest  it  has 
manifested  in  the  results  thus  go  back  to 
the  very  beginning  of  this  science  in  our 

Although  the  original  structure  of  chal- 
copyrite was  later  modified  and  carefully 
refined  by  Pauling  and  Brockway  (1932) 
on  a  crystal  from  Joplin,  Missouri,  it  was 
still  claimed  to  be  the  correct  one  by  Kozu 
and  Takane  (1934),  who  used  a  crystal 
from  the  Arakawa  Mine,  Ugo  province, 
Japan.  The  uncertainty  that  was  thus  cre- 
ated was  of  much  concern  to  Drs.  Corliss, 

Hastings,  and  Elliott  at  the  Brookhaven 
National  Laboratory.  They  found,  by  mag- 
netic susceptibility  measurements,  that 
chalcopyrite  is  antiferromagnetic  up  to 
200°  C,  the  highest  temperature  at  which 
measurements  were  made.  They  had  al- 
ready obtained  powder  neutron  diffraction 
data  on  material  from  the  locality  in  Japan. 
To  interpret  these  data  they  had  to  be  sure 
of  the  crystal  structure.  G.  Donnay  (to- 
gether with  J.  D.  H.  Donnay)  joined  this 
group  of  workers  at  Brookhaven  for  the 
summer  of  1957.  To  rule  out  the  possi- 
bility that  the  material  from  Japan  was 
a  quenched  high-temperature  form  of  chal- 
copyrite, single-crystal  X-ray  studies  were 
made  on  specimens  from  both  localities. 
The  outstanding  difference  between  the 
two  proposed  crystal  structures  is  that 
the  Kozu-Takane  structure  is  tetragonal 
but  metrically  pseudocubic,  cja  —  0.983, 
whereas  the  Pauling-Brockway  structure 
is  tetragonal  with  a  c/a  ratio  equal  to  1.97. 
Rotation  patterns  about  the  c  axis,  taken 
with  cobalt  radiation  so  as  to  enhance  the 
difference  in  scattering  powers  of  copper 
and  iron,  confirm  the  Pauling-Brockway 
structure  for  both  localities;  thus,  up  to 
now,  only  one  form  of  chalcopyrite  has 
been  found  in  nature. 


Single-crystal  neutron  data  were  then 
collected  on  specimens  from  both  locali- 
ties. These  data  can  be  accounted  for  as 
follows:  The  electron  spins  are  taken  to 
be  parallel  and  antiparallel  to  the  c  axis, 
the  moments  being  equal  in  magnitude  for 
atoms  of  the  same  species;  the  copper 
moments  must  be  set  equal  to  zero  (or  at 
any  rate  less  than  0.02),  and  the  contribu- 
tion of  the  aligned  iron  moments  must 
equal  3.85  Bohr  magnetons  per  atom.  With 
this  spin  arrangement  the  space  group 
M2d  of  the  Pauling-Brockway  structure  is 
preserved.  Equally  good  agreement  could 
be  obtained  with  a  somewhat  more  compli- 
cated model  in  which  the  moments  of  the 
iron  atoms  are  of  two  kinds,  slightly  dif- 
ferent in  magnitude,  and  the  difference  be- 
tween them  is  fortuitously  compensated  by 
small  unequal  moments  on  the  copper 
atoms;  this  model,  however,  requires  the 
"chemical"  space  group  to  be  abandoned. 
In  either  case  the  agreement  between  ob- 
served and  calculated  intensities  is  excel- 
lent; the  residual 

R=L||Fo|-|Fo||/Z|Fo|  is  0.04 

for  the  combined  data  from  the  two  locali- 

The  first  hypothesis  (copper  moments 
equal  to  zero)  is  preferred  because  it  re- 
quires the  material  to  be  antiferromagnetic 
without  having  to  depend  on  fortuitous 
numerical  values  for  the  moments.  This 
hypothesis  implies  that  the  chalcopyrite 
formula  should  be  written  Cu+Fe+++S2, 
which  in  turn  would  lead  to  the  predic- 
tion that  5  Bohr  magnetons  should  be  as- 
sociated with  each  iron  atom.  If  iron  is 
assumed  to  form  an  additional  bond  with 
sulfur,  an  assumption  supported  by  the 
fact  that  the  Fe-S  distance  is  shorter  than 
the  Cu-S  distance  by  0.12  A,  the  reduction 
of  the  iron  moment  from  5  to  3.85  Bohr 
magnetons  can  be  accounted  for. 

Symmetry  of  Magnetic  Structures 
G.  Donnay  et  al. 

In  crystal-structure  work,  atoms  are  im- 
plicitly assumed  to  have  complete  spherical 

symmetry,  so  that  they  can  occupy  any 
site  in  any  space  group.  The  point-group 
symmetry  of  the  site,  which  results  from 
the  neighboring  arrangement  of  matter, 
can  always  be  conferred  upon  the  atom 
that  occupies  it.  In  magnetic-structure 
studies,  however,  the  atom  that  carries  a 
magnetic  moment  has  its  own  symmetry 
lowered  to  that  of  an  axial  vector,  that  is 
oo  /m.  It  can  be  visualized  as  the  symmetry 
of  a  circular  loop  in  which  a  current  cir- 
culates in  a  given  sense.  Some  space-group 
sites  are,  therefore,  forbidden  to  magnetic 
atoms;  only  if  the  point-group  symmetry 
of  the  site  is  a  subgroup  of  °o  jm  can  a 
magnetic  atom  be  placed  on  it.  Such  a 
site  possesses  at  most  one  symmetry  direc- 
tion; the  magnetic  moment  must  be 
directed  along  it. 

In  the  early  days  of  crystal-structure  de- 
terminations, no  use  was  made  of  space- 
group  symmetry;  each  crystal  structure, 
considered  to  be  a  completely  new  prob- 
lem, was  solved  by  a  trial-and-error  pro- 
cedure requiring  considerable  intuition. 
Nishikawa's  suggestion  to  use  space-group 
criteria  systematically  in  order  to  limit  the 
number  of  trial  structures  met  considerable 
skepticism  at  first.  This  skepticism  is  in- 
deed difficult  to  appreciate  nowadays  when 
the  International  Tables  are  universally 
employed  by  all  workers  in  X-ray  structure 

We  should  now  like  to  repeat  Nishi- 
kawa's appeal,  addressing  it  this  time  to 
neutron-diffraction  workers  who  are  de- 
termining magnetic  structures.  They  are 
at  present  using  symmetry  considerations 
only  to  obtain  the  chemical  structure.  In 
the  next  and  most  interesting  step  of  their 
structure  determination,  namely  finding 
the  directions  of  the  axial  vectors  that  rep- 
resent the  magnetic  moments  on  the  vari- 
ous magnetic  atoms,  they  pay  no  attention 
to  the  restrictions  imposed  on  the  vector 
directions  by  the  presence  of  symmetry 
elements.  The  argument  usually  given 
to  defend  this  approach  is  the  following: 
the  symmetry  is  not  lowered  when  the 
substance  becomes  magnetic;  the  axial  vec- 


tors  alone  may  eliminate  some  or  all  sym- 
metry elements  of  the  chemical  space 
group  without  causing  sufficient  changes 
in  atomic  parameters  to  make  the  lowering 
of  symmetry  detectable  by  X  rays.  But  this 
reasoning  is  basically  no  more  valid  for 
magnetic  structures  than  for  chemical 
structures,  which  may  all  turn  out  to  be 
triclinic  too  if  we  wish  to  take  a  pessimistic 
point  of  view.  There  is  only  one  differ- 
ence: in  dealing  with  magnetic  moments 
we  must  keep  in  mind  that  antisymmetry 
elements  are  as  likely  to  occur  as  symmetry 

If,  then,  an  optimistic  approach,  analo- 
gous to  that  used  in  chemical-structure 
determinations,  is  taken,  we  shall  first 
make  use  of  the  magnetic  extinction  cri- 
teria to  determine  the  magnetic  diffraction 
aspect  which  will  lead  to  one  or  more  mag- 
netic space  groups.  If  one  of  these  mag- 
netic space  groups  is  the  same  as  the 
chemical  space  group,  it  will  be  tested  first. 
This  will,  in  general,  lead  to  a  very  small 
number  of  trial  structures.  If  none  of  them 
yields  satisfactory  agreement  between  cal- 
culated and  observed  structure  factors,  we 
next  try  the  isomorphous  space  groups  in 
which  some  or  all  of  the  symmetry  ele- 
ments are  replaced  by  antisymmetry  ele- 
ments. (These  space  groups  are  among 
the  1191  Shubnikov  groups  listed  by  Belov 
et  al.)  None  of  these  space  groups  intro- 
duces additional  atomic  parameters. 

Indeed,  the  justification  for  the  proposed 
procedure  is  the  desirability  of  minimizing 
the  number  of  parameters  to  be  deter- 
mined: it  does  not  appear  justifiable  to 
assume  that  the  atomic  parameters  remain, 
say  1/4,  1/4,  1/4,  in  a  triclinic  structure, 
where  they  should  be  given  as  0.25  ±  Ax, 
0.25 ±  Ay,  0.25  ±Az.  Finally,  if  none  of 
the  antigroups  leads  to  agreement,  a  space 
group  with  lower  symmetry  must  be  tried, 
and  the  number  of  atomic  parameters 
must  be  increased.  Eventually  a  triclinic 
space  group  may  thus  be  reached  although 
to  our  knowledge  no  such  case  is  on  record 
as  yet. 

Of  the  two  kinds  of  reflection,  nuclear 

and  magnetic,  that  may  contribute  to  a 
neutron-diffraction  peak,  let  us  consider 
the  magnetic  one  only.  Its  sign  depends  on 
the  sense  of  the  moment,  which  may  be 
reversed  by  some  of  the  symmetry  op- 
erations in  the  space  group,  leading  to  a 
reversal  of  some  of  the  extinction  rules. 
For  this  reason  the  appearance  of  addi- 
tional peaks  on  a  neutron-diffraction  pat- 
tern (as  compared  with  the  X-ray  pattern) 
indicates  antiferromagnetism  and  can  be 
used  to  determine  the  magnetic  diffraction 
aspect  and  associated  space  group  (s). 

The  example  of  chalcopyrite  (see  above) 
is  a  case  in  point.  Its  space  group  M2d, 
determined  by  X  rays  by  Pauling  and 
Brockway,  forbids  all  reflections  HHL  for 
which  the  sum  of  the  indices  is  not  a  mul- 
tiple of  4.  This  rule  holds  for  nuclear 
contributions.  Magnetic  reflections,  how- 
ever, obey  the  reversed  rule;  the  fact  that 
reflections  110,  114,  222  (with  sum  even 
but  not  divisible  by  4)  are  observed  is 
thus  a  confirmation  of  the  existence  of  the 
d  glide  planes  in  the  magnetic  structure. 

An  example  of  the  change  of  symmetry 
from  the  "chemical"  to  the  magnetic  space 
group  is  provided  by  Cr2Os.  The  structure 
is  known;  the  space  group  determined  by 
X  rays  is  R3c,  and  the  four  chromium 
atoms  in  the  rhombohedral  cell  lie  on  the 
threefold  axis,  in  position  4r  of  multiplicity 
4.  Brockhouse  (1953)  found  Cr203  to  be 
antiferromagnetic  and  determined  the 
spin  arrangement  from  neutron-diffraction 
data.  Pairs  of  chromium  atoms  related  by 
a  center  of  symmetry  in  the  chemical  space 
group  carry  compensating  magnetic  mo- 

Symmetry  considerations  immediately 
give  us  information  about  the  spin  direc- 
tion: the  spins  must  be  directed  along  the 
threefold  axis  if  the  magnetic  structure  is 
based  on  a  rhombohedral  lattice.  But  we 
know  that  the  operation  of  inversion  leaves 
the  direction  of  an  axial  vector  unchanged. 
It  follows  that  the  chemical  space  group 
Rlc  is  incompatible  with  the  spin  arrange- 
ment determined  by  Brockhouse.  The 
antigroup  in  which  the  centers  of  sym- 



metry  are  replaced  by  anticenters  is  R3'c 
(No.  106  in  the  listing  of  Belov  et  al., 
1957).  This  group  correctly  describes  the 
magnetic  structure  of  Cr203.  To  our 
knowledge  this  is  the  first  magnetic  struc- 
ture shown  to  belong  to  a  space  group  of 


High-Temperature  Chalcopyrite 

G.  Donnay  and  G.  Kullerud 

Only  one  form  of  chalcopyrite  has  been 
found  in  nature;  it  is  tetragonal,  a  =  5.24 
to  5.32  A,  c- 10.34  to  10.45  A  (see  above, 
under  Magnetic  Structure  of  Chalcopy- 
rite). The  existence  of  a  cubic,  high-tem- 
perature modification  has  now  been  es- 
tablished. It  has  been  obtained  in  the 
laboratory  in  the  form  of  single  crystals 
(0.05  mm  in  largest  dimension),  when  a 
sample  of  pure  CuFeS2  was  held  at  600°  C 
over  a  period  of  21  months  and  then 
quenched  in  water.  To  confirm  the  cubic 
symmetry,  crystals  were  examined  with 
Co  Ka  radiation  on  precession,  Weissen- 
berg,  and  rotation  cameras.  The  diffraction 
aspect  is  F###.  The  value  of  the  cell  edge 
was  refined  by  a  least-squares  analysis  of 
powder  data  obtained  on  the  Norelco  dif- 
fractometer:  a  =  5.264  ±0.003  A. 

Although  the  crystal  structure  of  the 
high-temperature  form  has  not  yet  been 
analyzed  in  detail,  a  preliminary  study 
indicates  that  it  is  isostructural  with  sphal- 
erite, ZnS,  in  which  case  Cu  and  Fe  atoms 
must  randomly  replace  Zn  atoms.  The  cell 
edge  of  ZnS,  a  =  5.406  ±  0.005  A,  differs 
from  that  of  cubic  CuFeS2  by  only  2.8  per 
cent,  so  that  a  complete  series  of  substitu- 
tion solid  solutions  would  be  expected  at 
high  temperature.  This  is  not  observed; 
10  per  cent  (by  weight)  of  CuFeS2  in  ZnS 
at  600°  C  exceeds  the  solubility  limit.  The 
saturated  (Zn,Cu,Fe)S  composition  has 
a  —  5.410  ±0.005  A,  equal  within  experi- 
mental error  to  the  cell  edge  of  the  pure 
end  member.  Similarly,  10  per  cent  (by 
weight)  of  ZnS  in  CuFeS2  at  600°  C  ex- 
ceeds the  solubility  limit.    Again  the  cell 

edge  does  not  change  measurably  on  maxi- 
mum substitution  of  Zn  for  Cu  and  Fe. 

N.  Morimoto 

Arsenopyrite  is  one  of  the  common  sul- 
fide minerals.  Buerger  has  studied  the 
structure  of  the  arsenopyrite  group  in  de- 
tail (1936,  1937).  According  to  him,  ar- 
senopyrite usually  shows  twinning,  and  it 
is  very  difficult  to  find  single  crystals  in 
nature.  Using  the  data  of  twinned  crystals 
of  arsenopyrite  combined  with  his  knowl- 
edge of  the  gudmundite  (FeSbS)  structure, 
he  obtained  a  marcasite-type  structure  for 

In  an  attempt  to  clarify  the  mechanism 
of  twinning  in  arsenopyrite,  to  determine 
its  structure  more  precisely,  and  to  study 
possible  structural  changes  with  changing 
composition  and  temperature  of  formation, 
the  examination  of  natural  and  synthetic 
arsenopyrite  was  undertaken  in  coopera- 
tion with  L.  A.  Clark.  The  following 
summary  shows  some  of  the  results  ob- 
tained so  far  by  X-ray  diffraction  methods. 

Natural  arsenopyrite.  Arsenopyrite  from 
Freiberg,  Germany,  was  studied.  Spectro- 
graphic  analysis  revealed  less  than  0.2 
per  cent  total  impurities  in  this  material 
with  Fe,  As,  and  S  as  the  only  major 

The  cell  dimensions  obtained  from  pow- 
der patterns  taken  on  the  Philips  diflfrac- 
tometer  are  ^  =  5.744  A,  £  =  5.676,  <r=5.784, 
and  0  =  112°  10'  (by  L.  A.  Clark).  Met- 
rically the  lattice  is  monoclinic.  The  pre- 

10  Buerger  (1939)  uses  an  unconventional 
monoclinic  setting,  in  which  the  positive  senses 
of  the  c  and  a  axes  enclose  an  acute  angle  [3.  To 
obtain  the  conventional  monoclinic  setting  ((3  ob- 
tuse) from  the  Buerger  setting  (1939),  we  must 
apply  the  transformation:  700/OTO/OO1.  To  ob- 
tain the  orthorhombic  B  setting  from  the  mono- 
clinic P  setting,  Buerger's  transformation  is  101/ 
010/T01;  the  transformation  lOl/OW/101  gives 
Buerger's  orthorhombic  B  cell  from  the  conven- 
tional monoclinic  P  cell.  _(To  keep  the  sense  of 
b,  the  transformation  101/010/101  might  have 
been  preferable.)  The  Buerger  B  cell  will  be  re- 
tained to  facilitate  comparisons. 


cession  and  Weissenberg  photographs  of 
five  specimens  of  this  material  always  gave 
a  triclinic  symmetry  of  intensity  distribu- 
tion and  a  splitting  of  reflections.  This 
suggests  that  the  real  symmetry  of  this 
arsenopyrite  is  triclinic  and  that  twinning 
always  exists. 

Some  of  the  specimens  gave  photographs 
that  clearly  show  the  splitting  of  reflec- 
tions, with  twin  plane  (101)  and  twin  axis 
[101].  This  splitting  is  easily  observed, 
owing  to  the  difference  in  length  of  the 
a*  and  c*  axes.  An  arsenopyrite  crystal 
heated  at  600°  C  for  4  days  in  a  sealed 
evacuated  silica  glass  tube  shows  different 
twinning,  with  twin  plane  (101)  and  twin 
axis  [101].  This  specimen,  however,  had 
not  been  studied  before  heating. 

Of  five  unheated  specimens  examined, 
two  did  not  show  any  of  the  above-men- 
tioned twinning;  however,  they  all  gave 
photographs  on  which  the  (^00),  (0^0), 
and  (00/)  reflections  were  split  into  two 
parts  and  the  {h\l)  reflections  into  three 
or  four  parts.  Usually  the  splitting  was 
very  slight  and  the  individual  reflections 
were  joined  by  diffuse  intensity  streaks. 
The  multiple  splitting  is  best  explained 
by  the  assumption  that  albite  and  pericline 
twinning  are  present  jointly  as  in  micro- 
cline  (Laves,  1950). 

Therefore,  this  natural  arsenopyrite  con- 
tains two  types  of  twinning.  The  combina- 
tion of  albite  and  pericline  twinning  which 
takes  place  as  the  result  of  the  pseudo- 
monoclinic  symmetry  of  the  mineral  is  so 
fine  that  it  is  impossible  to  obtain  a  truly 
single  crystal.  The  second  type  of  twin- 
ning is  due  to  the  pseudo-orthorhombic 
symmetry  and  has  twin  planes  (101)  or 
(101).  As  the  scale  of  this  twinning  is 
rather  large  compared  with  the  former 
type,  we  can  occasionally  obtain  an  edifice 
which  shows  only  the  small-scale  type  of 

The  existence  of  a  symmetry  center  was 
postulated  on  the  basis  of  a  Wilson  test  on 
the  intensity  distribution  of  the  (h{0) 
reflections.  The  space  group  of  the_natural 
arsenopyrite    is   accordingly    d-Pl9   with 

some  structural  extinctions.  An  accurate 
structure  analysis  is  in  progress  based  on 
(hOl),  (0/</),  and  (h{0)  reflections.  The 
structure  determined  by  Buerger  is  funda- 
mentally correct,  except  that  it  was  based 
on  space  group  C<n?-Pl\/c. 

Synthetic  arsenopyrite.  The  only  "single 
crystal"  examined  so  far  was  synthesized 
at  350°  C  by  L.  A.  Clark  from  a  bulk 
composition  lying  in  the  arsenopyrite-pyr- 
rhotite-liquid-vapor  four-phase  region  of 
the  Fe-As-S  system.  The  composition  of 
the  arsenopyrite  is  near  FeAsi.iSo.9. 

The  cell  dimensions  determined  from 
the  powder  patterns  (by  L.  A.  Clark)  are 
am  =cm  =  5.796  A,  bm  =  5.720  A,  and  p  =  113° 
21'.  As  the  length  of  the  a  axis  is  the  same 
as  that  of  the  c  axis,  it  is  possible  to  calcu- 
late the  dimensions  of  the  orthorhombic 
cell  in  which  a0  =  9.661  A,  b0  —  5.710  A,  and 
ca  =  6.403  A. 

The  "single-crystal"  precession  and 
Weissenberg  photographs  also  give  ortho- 
rhombic  symmetry  of  intensity  distribu- 
tion, with  values  of  a0  —  9.669  A,  £0  =  5.710 
A,  and  (T0  =  6.400  A  (all  ±0.3  per  cent). 
Examination  of  all  reflections  observed  re- 
veals that  (h^l)  appears  only  when  h  +  l= 
2n  and  (0^0)  only  when  \  —  2n.  If  we 
accept  the  existence  of  a  symmetry  center, 
the  only  possible  space  group  is  D27119- 
Bmmm  and  structural  extinctions  occur 
for  (0^0),  ^  odd.  Further  examination 
also  gives  the  additional  rules:  (1)  (0^/) 
appears  only  when  \Jrll2  —  2n>  and  (2) 
(kOl)    only  when  h+l  =  4n   or  h—\  —  \n. 

To  account  for  space  group  Bmmm,  we 
must  (1)  assume  disorder  of  As  and  S  in 
the  structure,  or  (2)  postulate  minute 
twinning  of  a  monoclinic  crystal,  or  (3) 
discard  the  marcasite-type  structure.  This 
structure,  however,  gives  fairly  good  agree- 
ment between  observed  and  calculated  F 
values  in  natural  arsenopyrite.  Taking 
into  account  the  additional  extinction  rules, 
the  second  assumption  seems  to  be  the 
most  reasonable. 

If  we  assume  that  the  apparent  ortho- 
rhombic  symmetry  in  the  synthetic  arseno- 
pyrite is  due  to  twinning  of  a  monoclinic 


crystal  (cell  dimensions  of  L.  A.  Clark, 
above)  by  reflection  on  the  (101)  plane, 
the  transformed  extinction  rules  lead  to 
space  group  Czh-P2\jc.  The  twinning  of 
this  space  group  explains  the  second  addi- 
tional extinction  rule  noted  in  the  ortho- 
rhombic  cell. 

The  accurate  structure  analysis  of  syn- 
thetic arsenopyrite  is  hampered  by  the  fact 
that  it  has  been  impossible  to  get  a  crystal 
free  from  twinning.  Nevertheless,  the 
analysis  is  in  progress,  using  reflections 
that  are  not  affected  by  twinning.  The 
cell  dimensions  determined  from  powder 
patterns  of  various  samples  of  synthetic 
arsenopyrite  are  constant,  and  for  all  sam- 
ples the  composition  appears  to  be  close  to 
FeAsi.iSo.9  (see  report  of  L.  A.  Clark). 
For  natural  arsenopyrites,  however,  the  cell 
dimensions  differ  measurably  from  locality 
to  locality,  suggesting  that  their  composi- 
tions deviate  from  that  of  the  synthetic 

G.  Donnay,  J.  D.  H.  Donnay,  and  G.  Kullerud 

Bornite,  CusFeS*,  although  a  common 
and  widespread  sulfide  mineral,  has  rarely 
been  reported  in  single  crystals.  Well 
formed,  steel-blue,  euhedral  crystals  were 
found  by  Kullerud  at  Copper  Corp.  Mine, 
Ontario.  The  hand  specimen  shows  them 
occurring  with  barite  crystals,  an  associa- 
tion previously  reported  only  from  the 
Cheshire  locality.  The  dominant  form  of 
the  bornite  crystals  is  the  cube,  truncated 
by  small  octahedral  faces.  Their  cube  edge 
ranges  in  length  from  0.2  to  about  1.5  mm. 
All  cube  edges  are  replaced  by  staircase- 
like indentations,  consisting  of  an  alterna- 
tion of  two  faces  of  the  octahedron;  for 
instance,  the  three  cube  edges  that  meet  at 
the  corner  truncated  by  (111)  are  replaced 
by_an  alternation  of  (111)  faces  with  (Til), 
(111),  or  (HI),  respectively,  according  as 
the  edge  is  parallel  to  the  a,  b,  or  c  axis. 
The  height  of  these  steps  ranges  from  0.01 
to  0.5  mm. 

Several  chips  were  examined  on  the  rota- 
tion, Weissenberg,  and  precession  cameras 

using  Mo  Ka,  Cu  Ka,  and  Co  Ka  radia- 
tions. Two  types  of  bornite  were  observed : 
one  is  cubic,  diffraction  aspect  FJ##,  space 
group  Fd3m  if  holohedral,  cell  edge  21.94 
±0.06  A;  the  other  has  a  primitive  ortho- 
rhombic  lattice,  with  pseudotetragonal  cell 
dimensions,  a  =  £=21.90 ±0.006,  <r=10.95 
±0.03  A.  The  latter  type  has  previously 
been  reported  by  Frueh  (1950)  on  low- 
temperature  crystals  from  Bristol,  Connec- 
ticut. In  addition  Tunell  and  Adams 
(1949)  have  reported  a  cubic  type  with 
cell  edge  32.8  A.  The  original  structure 
determination  by  Lundquist  and  Westgren 
(1936)  assumed  a  cubic  cell,  10.93  A  along 
the  edge.  (It  is  not  certain,  however, 
whether  weak  reflections  were  taken  into 
account  or  ignored  in  this  study.)  Thus 
we  now  know  of  at  least  three  types  of  low- 
temperature  bornite,  two  of  which  have 
been  found  on  the  same  hand  specimen 
and  appear  to  belong  to  a  single  genera- 

Like  many  other  sulfides,  bornite  has 
a  crystal  structure  composed  of  sulfur 
atoms  in  a  face-centered  cubic  arrange- 
ment, cube  edge  5.475  A,  and  of  metal 
atoms  that  must  be  distributed  in  the  avail- 
able holes:  at  1/4  1/4  1/4,  etc.,  in  tetrahe- 
dral  coordination,  at  1/3  1/3  1/3,  etc.,  in 
trigonal  coordination,  and  at  1/2  1/2  1/2  in 
octahedral  coordination.  The  various  low- 
temperature  forms  must  differ  in  the  ar- 
rangement of  the  metal  atoms.  There  is 
good  evidence  (Frueh,  1950)  that,  in  the 
high-temperature  form  (cubic,  with  cell 
edge  10.93  A),  copper,  iron,  and  vacant 
sites  form  a  substitution  solid  solution;  in 
other  words,  disorder  exists  among  cations 
both  as  to  chemical  nature  and  as  to  atomic 

The  large  cell  contents  of  the  various 
bornite  types — 1080  copper  atoms,  216  iron 
atoms,  and  864  sulfur  atoms  in  the  large 
cube  described  by  Tunell  and  Adams — 
make  us  suspect  that  we  are  dealing  with 
a  complex  edifice,  either  a  twin  as  in  di- 
genite  (Donnay,  Donnay,  and  Kullerud, 
1958)  or  a  syntaxic  intergrowth  of  two 
distinct  phases  as  in  andorites  IV  and  VI 


(Donnay  and  Donnay,  1954).  The  latter 
hypothesis  accounts  for  the  specimens  that 
show  a  cube  edge  of  32.8  =  6x5.47  A  and 
have  their  first,  fifth,  seventh,  and  eleventh 
layer  lines  completely  absent  on  the  tf-axis 
rotation  photographs  presented  by  Tunell 
and  Adams.  The  absences  are  accounted 
for  if  we  postulate  a  syntaxic  intergrowth 
of  two  cubic  phases,  a  known  one  with  cell 
edge  10.94  =  2x5.47  A  and  a  hypothetical 
one  with  cell  edge  16.41  =  3x5.47  A 
(fig.    46).     Unfortunately,    however,    we 










-■  3 






■  ■  0 
°lw  s'0.94  =  2x5.47 


=  3x 

5.47  A 

=  32.8 

=6x5. 47A 

Fig.  46.  Schematic  explanation  of  bornite  rota- 
tion pattern  (observed  by  Tunell  and  Adams). 
The  proposed  intergrowth  of  two  cubic  forms, 
one  with  a  =  2.547  A  (found  by  Lundquist  and 
Westgren)  and  a  second  one  with  a  —  3.547  A 
(not  yet  reported  in  the  literature)  accounts  for 
the  observed  absence  of  first,  fifth,  seventh,  and 
eleventh  layer  lines. 

cannot  account  for  the  other  two  forms 
in  this  way.  Attempts  to  find  systematic 
structural  absences  among  the  observed 
reflections  of  the  other  two  forms,  which 
would  give  clues  about  twinning,  have  so 
far  been  unsuccessful.  Work  on  the  bornite 
puzzle  is  being  continued. 



N.  Morimoto  et  al. 

The  determination  of  accurate  crystal 
structures  of  rock-forming  minerals  of 
known  chemical  composition  formed  un- 

der various  conditions  has  become  an  im- 
portant field  in  mineralogy.  Crystallog- 
raphers  have  carried  out  excellent  studies 
on  feldspars  and  amphiboles.  The  corre- 
sponding study  of  pyroxene  structure  in 
relation  to  chemical  composition  and  gene- 
sis, however,  has  not  yet  been  made  in 
spite  of  recent  developments  in  the  peno- 
logical study  of  minerals  of  this  group. 
Some  accurate  measurements  of  the  cell 
dimensions  have  been  made  by  Hess  and 
Kuno.  To  cover  the  gap  between  the 
petrology  and  crystallography  of  pyrox- 
enes the  X-ray  study  of  pyroxenes  of  vari- 
ous compositions  has  been  undertaken. 

The  pyroxenes  present  exceptional  com- 
plexity both  of  crystalline  modifications 
and  of  chemical  composition.  The  most 
important  natural  pyroxenes,  however,  oc- 
cur in  the  ternary  system  MgSi03-FeSi03- 
CaSiOs,  with  less  than  50  mole  per  cent 
CaSiOs,  although  a  minor  amount  of  re- 
placement of  other  cations  (Al,  Ti,  etc.) 
usually  takes  place.  Over  most  of  this  field, 
the  pyroxenes  are  monoclinic  and  an 
orthorhombic  modification  is  more  stable 
only  when  there  is  less  than  10  per  cent 
CaSi03  at  low  temperature.  In  the 
MgSiOs-CaSiOs  join,  however,  Atlas 
(1952)  and  Schairer  and  Boyd  (1957) 
showed  that  protoenstatite,  an  orthorhom- 
bic modification  other  than  the  usual  en- 
statite,  is  stable  at  high  temperature,  and 
clinoenstatite  is  a  metastable  form. 

Clinopyroxenes.  Clinopyroxene  in  the 
ternary  system  MgSi03-FeSi03-CaSi03 
can  be  divided  into  two  different  types  on 
the  basis  of  space  group,  (a)  diopside- 
hedenbergite  join  (augite  type),  (#)  clino- 
enstatite-clinoferrosilite  join  (pigeonite 

The  augite-type  clinopyroxenes  belong  to 
space  group  C2h6-C2i/c  (Warren  and  Bis- 
coe,  1931).  Their  3  angle  and  cell  volume 
change  continuously  as  the  replacement  of 
metallic  atoms  takes  place,  suggesting  the 
existence  of  complete  solid  solution  in  this 
type.  In  the  pigeonite-type  clinopyroxenes, 
however,  the  3  angle  is  nearly  constant 
throughout  the  type,  though  the  cell  vol- 
ume changes  gradually.  The  space  group 


of  this  type  was  found  to  be  Cth-Flija 
(Morimoto,  1956;  Gay  and  Bown,  1957). 

The  existence  of  a  miscibility  gap  be- 
tween diopside  and  clinoenstatite  (Atlas, 
1952;  Schairer  and  Boyd,  1957)  was  con- 
firmed. It  is  not  certain,  however,  at  what 
compositions  the  solvus  intersects  the 
solidus  in  clinopyroxenes,  though  it  is  be- 
lieved that  a  complete  solid  solution  be- 
tween the  augite-  and  pigeonite-type  clino- 
pyroxene  begins  to  exist  when  the  pyrox- 
enes become  rich  in  Fe+2.  The  following 
information  (obtained  jointly  with  T.  Ito, 
Tokyo  University)  throws  some  light  on 
this  problem. 

Phenocrysts  in  an  andesite  from  Ha- 
kone,  Japan,  studied  by  the  Weissenberg 
method,  show  very  minute  intergrowths  of 
pigeonite  (Woi6En45Fs39)  and  augite 
(Wo32En37Fs3i)  having  the  (001)  plane  in 
common  (Morimoto,  1956).  Some  speci- 
mens from  the  same  locality  show  an  un- 
usual twinlike  structure  of  two  such  inter- 
growths of  pigeonite  and  augite  as  in 
figure  47.  The  b  axes  of  the  four  crystals 
in  this  edifice  are  the  same  not  only  in 
orientation  but  also  in  length.  The  c  axes 
of  the  two  pigeonite  crystals  enclose  an 
angle  of  1°,  and  of  the  two  augite  crystals, 
an  angle  of  3.6°.  This  combination  of  two 
pigeonite-augite  intergrowths  can  be  in- 
terpreted to  be  in  "twin"  relation  on  a 
plane  whose  indices  are  not  integers  but 
which  is  close  to  (100)  for  all  four  crystals. 
Although,  of  course,  we  cannot  refer  to 
this  edifice  as  a  twin,  we  call  it  a  "pseudo- 
twin."  On  the  microscope,  the  sample 
used  in  X-ray  studies  does  simulate  poly- 
synthetic  twinning  of  pigeonite  parallel 
to  the  (100)  plane,  with  rather  broad 
lamellae.  Between  pigeonite  bands  there 
are  narrow  strips,  so  narrow  in  fact  that  it 
is  impossible  optically  to  decide  whether 
they  are  pigeonite  or  augite.  To  judge 
from  X-ray  experiments,  they  are  probably 

The  existence  of  this  polysynthetic  twin- 
like structure  of  pigeonite  and  augite  can 
reasonably  be  understood  as  an  exsolution 
product  of  a  mixed  crystal.  Clinopyroxene, 

originally  crystallized  as  a  homogeneous 
phase,  later  unmixed  to  pigeonite  and 
augite.  The  composition  plane  of  the  pi- 
geonites  and  augites  is  supposed  to  be  the 
(100)  plane  of  the  homogeneous  phase. 
The  coincidence  of  the  b  axes  of  pigeonite 
and  augite  also  suggests  that  they  have 
unmixed  from  one  high-temperature  py- 

It  is  possible  to  say  from  the  above  re- 
sults that  there  is  a  mixed  crystal  between 
augite  and  pigeonite  at  a  high  tempera- 
ture but  that  unmixing  to  augite  and  pi- 


Qp    and  Q^ 

CI     and  0 
Ptt  A 

Fig.  47.  Unusual  pseudo-twin  of  pigeonite  and 
augite,  which  suggests  exsolution  from  homo- 
geneous clinopyroxene.  «PI,rPI:  the  a  and  c 
axes  of  pigeonite  I.  «PII,fpn:  the  a  and  c  axes  of 
pigeonite  II.  aAl,cA1:  the  a  and  c  axes  of  augite  I. 
tfAII,<rAII:  the  a  and  c  axes  of  augite  II.  Pigeonite 
I  and  augite  I  have  the  (001)  plane  in  common. 
The  relation  between  pigeonite  II  and  augite  II 
is  the  same  as  that  of  pigeonite  I  and  augite  I. 
The  composite  plane  is  supposed  to  be  the  (100) 
plane  of  the  original  clinopyroxene. 

geonite  takes  place  on  cooling  at  the  com- 
position of  Woi8En44Fs3s.  (This  composi- 
tion is  estimated  from  the  ratio  of  the 
intensities  of  the  reflections  of  pigeonite 
and  augite  on  X-ray  photographs.)  We 
are  planning  heating  and  quenching  ex- 
periments on  phenocrysts  that  consist  of 
pseudo-twins.  The  case  of  pseudo-twin- 
ning  reported  here  is  the  second  one  on 
record.  The  first  one  was  found  in  a 
feldspar  specimen  (Spencer  N)  by  G.  Don- 
nay  and  J.  D.  H.  Donnay  (Year  Book  53). 
Clinopyroxene  of  the  pigeonite  type  has 
a  primitive  lattice  and  shows  the  reflections 
h\l  with  (h  +  f()  odd,  which  do  not  occur 
in  the  augite-type  clinopyroxenes.   In  fig- 



ure  48  parts  of  the  powder  patterns  of  four 
materials  belonging  to  this  type  are  shown, 
together  with  their  chemical  compositions. 
The  intensity  of  the  reflection  (231)  de- 
creases as  the  Ca  and  Fe  contents  in- 
crease, and  almost  disappears  in  the  pat- 

440      450      46°     47°      48°     47°      50° 
i     ■      i      .      i      ■      i      ■ I i I — i — I 





(Heated  hypersthene) 








44°      45e 
I i l_ 

46°     47° 







Fig.  48.  The  diffraction  patterns  of  (A)  syn- 
thetic clinoenstatite;  (B)  pigeonite  transformed 
from  hypersthene,  Kamakura,  Japan  (heated  at 
1400°  C  for  16  hours  in  a  vacuum  silica  tube) ; 
(C)  pigeonite,  Usugoyazawa,  Japan;  (D)  ferro- 
pigeonite, Ashio,  Japan,  taken  by  Norelco  X-ray 
diffractometer  using  Fe-radiation. 

tern  of  ferropigeonite  of  the  composition 
Woi2En26Fs62,  suggesting  a  C-centered 
lattice,  which  is  that  of  the  augite-type 

The  single-crystal  photographs  of  this 
ferropigeonite,  however,  give  a  fairly 
strong  reflection  (231)  and  other,  addi- 
tional reflections  with  (h  +  J()  odd,  which 

are  proof  of  the  primitive  lattice.  These 
additional  reflections  are  usually  some- 
what diffuse  (Gay,  1957)  and  seem  to 
have  streaks  along  the  a  axis.  On  the 
other  hand,  pigeonite  of  composition 
Woi6En45Fs39  shows  very  sharp  additional 
reflections.  The  intensity  of  the  additional 
reflections  is  generally  weaker  than  that 
of  the  usual  reflections  (h  +  ^  =  2n)  in  the 
ferropigeonite,  and  even  weaker  in  the 
pigeonite.  This  observation  suggests  that 
the  intensity  of  the  additional  reflections 
does  not  depend  on  the  Fe  content. 

Thus  we  can  conclude  that  (1)  the  in- 
tensity of  the  additional  reflections  de- 
pends on  the  amount  of  Ca  in  the  struc- 
ture, and  (2)  the  difTuseness  of  the  addi- 
tional reflections  depends  on  the  amount 
of  Fe.11 

The  accurate  structure  analysis  of  clino- 
enstatite and  ferropigeonite  is  now  in  prog- 
ress (with  H.  T.  Evans  and  D.  E.  Apple- 
man,  U.  S.  Geological  Survey)  and  should 
clarify  the  following  questions:  (1)  To 
what  extent  are  the  atomic  coordinates  of 
clinoenstatite  and  ferropigeonite  different 
from  those  of  diopside?  (2)  To  what  ex- 
tent are  the  atomic  coordinates  of  clino- 
enstatite different  from  those  of  ferropi- 
geonite? (3)  What  is  the  configuration  of 
cations  such  as  Ca,  Mg,  and  Fe  atoms  in 
ferropigeonite  structure?  (4)  Why  does 
the  gradual  change  of  intensity  and  diffuse- 
ness  of  the  additional  reflections  in  the 
pigeonite-type  clinopyroxenes  take  place  ? 

Specimens  of  clinoenstatite  were  ob- 
tained by  heating  Webster  26  bronzite 
(En9iFs9)  and  Bishopville  enstatite  at 
1400°  C  for  24  hours.  They  usually  show 
fine  poly  synthetic  twinning  on  the  (100) 
plane,  suggesting  the  transition  from  pro- 
toenstatite,  which  is  supposed  to  be  a 
stable  high-temperature  form.  The  single 

11  According  to  Drs.  Gay  and  Bown,  Cam- 
bridge, England  (private  communication),  the 
difTuseness  of  these  reflections  does  not  change 
after  2  days  of  heating  at  1000°  C  and  quenching 
in  air.  Therefore,  they  conclude,  the  difTuseness 
does  not  depend  on  the  rate  of  cooling  of  the 


crystals  of  ferropigeonite  (W012E1126FS62) 
were  obtained  from  a  dike  at  Ashio, 
Japan,  by  Kuno. 

The  structures  of  clinoenstatite  and  fer- 
ropigeonite were  obtained  by  deforming 
the  structure  of  diopside,  giving  the  addi- 
tional reflections.  Their  structures  are  now 
in  the  stage  of  refinement  using  (h^O), 
(hOl),  and  (0/(/)  reflections. 

Orthopyroxenes.  The  structure  analysis 
of  ferropigeonite  suggests  that  Mg,  Fe,  and 
Ca  atoms  are  randomly  distributed  be- 
tween two  crystallographically  different 
positions.  In  orthopyroxenes,  which  are 
a  stable  form  at  low  temperature,  we  can 
expect  some  ordering  of  cations  into  dif- 
ferent positions.  Furthermore,  T.  Ito 
(1950),  on  the  basis  of  his  twinning 
theory,  proposed  a  structure  for  bronzite 
(En84Fsi6)  slightly  different  from  that  pro- 
posed by  Warren  and  Modell  for  hyper- 
sthene  (En7oFs3o).  Refinement  of  enstatite 
(Bishopville,  Enioo)  and  ferrohypersthene 
(Hirogawara,  En55Fs45)  structures  is  now 
under  way. 

In  the  course  of  the  study  of  orthopyrox- 
enes, it  was  found  that  orthopyroxenes  of 
plutonic  origin  (Great  Dike,  Webster,  etc.) 
always  show  some  additional  reflections, 
such  as  (012),  (032),  etc.,  though  they  are 
very  weak.  These  reflections  are  prohibited 
by  the  extinction  rule  of  the  space  group 
of  orthopyroxenes  D2h15-Pbca.  In  ortho- 
pyroxenes of  volcanic  or  meteoritic  origin 
(Bishopville,  Bonin  Island,  Hirogawara, 
etc.),  however,  it  was  impossible  to  observe 
these  reflections  even  in  photographs  of 
long  exposure.  These  reflections,  therefore, 
can  be  interpreted  as  (112),  (132),  etc.,  of 
the  diopsidic  pyroxene  on  the  basis  of  the 
exsolution  of  diopsidic  pyroxene  from  hy- 
persthene  by  cooling  and  the  intergrowth 
of  the  diopsidic  pyroxene  and  hypersthene 
on  the  (100)  plane. 

Structural  relation  between  enstatite, 
clinoenstatite,  and  protoenstatite.  It  is  im- 
portant to  study  the  mechanism  of  transi- 
tion from  orthopyroxene  to  clinopyroxene. 
Diffuse  reflections  of  monoclinic  symmetry 
along  the  a  *  axis  can  be  observed  even  in 

natural  enstatite;  the  symmetry  of  the  dif- 
fuse reflections  changes  to  orthorhombic 
when  the  mineral  inverts  to  clinoenstatite 
by  heating.  The  study  of  the  transition 
mechanism  between  clinoenstatite,  ensta- 
tite, and  protoenstatite  and  the  interpreta- 
tion of  the  diffuse  reflections  can  be  un- 
dertaken by  means  of  heating  experiments 
on  single  crystals,  once  the  precise  struc- 
ture of  these  minerals  is  known.  Some 
heating  experiments  have  been  conducted 
this  year,  and  the  work  will  be  continued 
during  the  coming  year. 

Synthetic  Mica  of  Type  3M 
G.  Donnay  and  P.  Kingman 

Although  numerous  studies  have  been 
made  on  the  mica  group  of  minerals  be- 
cause of  their  abundance  and  geological 
importance,  precise  crystal-structure  data 
are  still  unavailable.  Information  about 
bond  angles  and  interatomic  distances  is 
still  of  a  comparatively  crude  nature. 

A  structure  determination  of  a  synthetic 
mica  would  be  of  considerable  value,  since 
one  of  the  chief  features  of  the  group  is 
the  multiplicity  of  isomorphous  substitu- 
tions, and  information  about  bond  angles, 
lengths,  etc.,  obtained  from  a  single-crystal 
study  would  be  of  great  assistance  in  un- 
derstanding the  crystal  chemistry  of  these 

In  the  extensive  study  of  the  mica  sys- 
tem, KFe3(FeSi3Oio)  (OH)2  has  been  syn- 
thesized (Wones  and  Eugster)  in  sheets 
thick  enough  to  be  used  for  X-ray  studies 
by  single-crystal  methods.  This  synthetic 
mica  has  distinct  advantages  for  a  struc- 
ture determination.  First,  the  composition 
of  the  synthetic  mineral  is  much  more  ac- 
curately known  than  that  of  naturally  oc- 
curring micas.  Second,  the  substitution  of 
nothing  but  Fe+++  atoms  for  some  of  the  Si 
atoms  in  tetrahedral  coordination — a  sub- 
stitution apparently  not  occurring  in  na- 
ture— offers  the  possibility  of  determining 
whether  such  substitutions  are  ordered  or 
disordered.  In  natural  micas  the  element 
that  substitutes  for  silicon  is  aluminum, 


whose  atomic  weight  is  so  similar  to  that 
of  silicon  that  distinguishing  between  the 
two  by  X-ray  diffraction  is  difficult.  It  is 
also  possible  that  the  substance  may  have 
interesting  magnetic  properties  due  to  the 
presence  of  Fe++  and  Fe+++,  and  examination 
by  neutron  diffraction  might  prove  fruitful. 
Crystals  have  been  examined  on  preces- 
sion and  Weissenberg  cameras  with  Fe  Ka 
(A  =  1.9373  A)  and  Mo  Ka  (A  =  0.7108  A) 
radiations.    The  lattice  is  metrically  hex- 

agonal, but  the  symmetry  is  monoclinic, 
with  diffraction  aspect  C#.  The  cell  di- 
mensions are:  a  =  5.434,  £  =  9.404,  <r  =  30.49 
A,  all  ±0.3  per  cent,  (3  =  90°  0'.  For  an 
orthohexagonal  lattice  b  =  2a  cos  30°  = 
9.41 2  A,  which  is  equal  within  experi- 
mental error  to  the  measured  b  value. 
The  hexagonal  plates,  with  excellent  cleav- 
age (001),  as  well  as  the  cell  dimensions, 
indicate  a  mica-type  structure  with  a  cell 
three  layers  high. 


Petrologists'  Club 

The  Petrologists'  Club  held  an  increased 
number  of  meetings  this  year  and  revived 
its  annual  field  trip.  Ores  and  ore  solu- 
tions, together  with  isotope  studies,  were 
subjects  discussed  at  the  majority  of  the 
meetings.  A  field  trip  to  examine  the 
metamorphic  rocks  in  the  Fairfax  quad- 
rangle, Virginia,  was  led  by  Charles  Mil- 
ton and  Edward  C.  T.  Chao.  D.  B.  Stew- 
art and  E.  H.  Roseboom,  Jr.,  were  elected 
to  serve  as  chairmen  for  1959.  Speakers 
whom  the  club  was  privileged  to  hear  this 
past  season  are  listed  below. 

"016/018  ratios  in  coexisting  minerals  from 
various  geologic  deposits,"  by  Sam  Epstein 
(California  Institute  of  Technology). 

"Some  observations  on  natural  and  syn- 
thetic carbonates,"  by  Julian  R.  Goldsmith 
(University  of  Chicago). 

"Composition  of  magmatic  gases  at  high 
temperatures,"  by  Konrad  B.  Krauskopf 
(Stanford  University). 

"Thermochemical  data  and  its  application 
to  limestone  replacement,"  by  H.  Holland 
(Princeton  University). 

"Fractionation  of  deuterium  in  earth  proc- 
esses," by  Irving  Friedman  (U.  S.  Geological 

"Some  aspects  of  current  sulfide  research," 
by  G.  Kullerud,  E.  H.  Roseboom,  Jr.,  and  R. 
G.  Arnold  (Geophysical  Laboratory). 

"Fluid  pressure  hypothesis  and  overthrust- 
ing  in  western  Wyoming,"  by  W.  W.  Rubey 
(U.  S.  Geological  Survey). 

"Sediments  of  California  marine  basins," 
by  Kenneth  O.  Emery  (University  of  South- 
ern California). 


The  Laboratory  continued  its  weekly 
series  of  seminars,  with  papers  presented 
largely  by  staff  members  and  concerned 
mainly  with  discussions  of  work  in  prog- 
ress. The  following  talks  were  given  by 
guest  speakers  from  outside  the  Labora- 
tory : 

"The  olivine-spinel  transition  and  its  bear- 
ing on  the  composition  of  the  mantle,"  by  A. 
E.  Ringwood  (University  of  Melbourne,  Aus- 

"Standard  free  energies  of  formation  de- 
duced from  mineral  occurrences,"  by  R.  M. 
Garrels  (Harvard  University). 

"Infra-red  spectroscopy  of  inorganic  sys- 
tems," by  E.  R.  Lippencott  (University  of 

"Activation  analysis  applied  to  geochemical 
problems,"  by  George  W.  Reed  (Argonne 
National  Laboratory). 

"Some  variations  in  018/016  ratios  in  nat- 
ural minerals,"  by  Sam  Epstein  (California 
Institute  of  Technology). 

"Attenuation  of  small-amplitude  stress 
waves  in  solids,"  by  Gordon  J.  F.  MacDonald 
(Massachusetts  Institute  of  Technology). 

"Problems  in  mineral  fades,"  by  James  B. 
Thompson  (Harvard  University). 

"Carbon-14  dating  and  the  Pleistocene  geol- 
ogy of  Europe,"  by  Hessel  de  Vries  (Univer- 
sity of  Groningen,  Netherlands). 

"Geochemical  prospecting,"  by  H.  E. 
Hawkes  (University  of  California,  Berkeley). 

"Geochemistry  of  the  isotopes  of  nitrogen," 
by  T.  Hoering  (University  of  Arkansas). 

"Geochemical  indications  of  the  environ- 
ment of  deposition  of  sedimentary  rocks,"  by 
M.  L.  Keith  (Pennsylvania  State  University). 


"Magnetic  structure  of  chalcopyrite  as  de- 
termined by  neutron  diffraction,"  by  L.  M. 
Corliss  (Brookhaven  National  Laboratory) 
and  G.  Donnay  (Geophysical  Laboratory). 

"Some  aspects  of  the  geochemistry  of  sul- 
fur isotopes,"  by  Wayne  U.  Ault  (U.  S.  Geo- 
logical Survey). 

"Research  in  ceramics,"  by  A.  T.  Green 
(British  Ceramic  Research  Association). 

"The  dating  of  marine  sediments:  Prob- 
lems and  progress,"  by  Gustaf  O.  Arrhenius 
(Scripps  Institution  of  Oceanography). 

"Some  aspects  of  the  geochemistry  of  pe- 
troleum," by  W.  E.  Hanson  (Mellon  Insti- 

"The  dynamic  consequences  of  phase  tran- 
sitions in  the  mantle,"  by  Gordon  }.  F.  Mac- 
Donald  (Massachusetts  Institute  of  Technol- 

"Operation  Rainier,"  by  G.  W.  Morey 
(U.  S.  Geological  Survey). 

"The  application  of  optical  methods  to  the 
study  of  imperfect  structures,"  by  H.  Lipson 
(College  of  Science  and  Technology,  Man- 

]ohns  H opsins  University  and  Geophysi- 
cal Laboratory  Graduate  Seminars  on 
"Researches  in    Geochemistry" 

During  the  academic  year  1957-1958  the 
Geophysical  Laboratory  arranged  a  weekly 
series  of  graduate  seminars  on  research  in 
geochemistry  at  the  Johns  Hopkins  Uni- 
versity. Many  of  the  speakers  also  visited 
and  gave  talks  at  the  Geophysical  Labora- 
tory. The  content  of  these  lectures  has 
been  assembled  for  publication  by  John 
Wiley  &  Sons  in  a  book  entitled  Researches 
in  Geochemistry. 

Speakers  and  topics  included:  Gunnar 
Kullerud  (Geophysical  Laboratory),  "Sul- 
fide systems  as  geological  thermometers"; 
Hans  P.  Eugster  (Geophysical  Labora- 
tory), "Reduction  and  oxidation  in  meta- 
morphism";  Robert  M.  Garrels  (Harvard 
University),  "Rates  of  geochemical  reac- 
tions at  low  temperatures  and  pressures"; 
George  R.  Tilton  (Geophysical  Labora- 
tory), "Geochronology";  George  W.  Reed 
(Argonne  National  Laboratory),  "Activa- 
tion analysis  applied  to  geochemical  prob- 

lems"; Sam  Epstein  (California  Institute 
of  Technology),  "Variation  of  the  018/016 
ratio  in  nature";  Julian  R.  Goldsmith 
(University  of  Chicago),  "Some  aspects  of 
the  geochemistry  of  carbonates";  Willard 
F.  Libby  (Atomic  Energy  Commission 
and  Geophysical  Laboratory),  "Tritium  in 
hydrology  and  meteorology";  Konrad  B. 
Krauskopf  (Stanford  University),  "The 
use  of  equilibrium  calculations  in  finding 
the  composition  of  a  magmatic  gas  phase"; 
Gordon  J.  F.  MacDonald  (Massachusetts 
Institute  of  Technology),  "Chondrites  and 
the  chemical  composition  of  the  earth"; 
James  B.  Thompson  (Harvard  Univer- 
sity), "The  phase  rule  and  metasomatism"; 
Francis  R.  Boyd  (Geophysical  Labora- 
tory), "Hydrothermal  investigations  of 
amphiboles";  Paul  B.  Barton,  Jr.  (U.  S. 
Geological  Survey),  "The  chemical  en- 
vironment of  ore  deposition  and  the  prob- 
lem of  low-temperature  ore  transport"; 
H.  E.  Hawkes  (University  of  California), 
"Geochemical  prospecting";  Wayne  U. 
Ault  (U.  S.  Geological  Survey),  "Sulfur 
isotopic  fractionation  in  geochemical  proc- 
esses"; M.  L.  Keith  (Pennsylvania  State 
University),  "Geochemical  indicators  of 
marine  and  fresh-water  sediments";  Philip 
H.  Abelson  (Geophysical  Laboratory), 
"Geochemistry  of  organic  substances"; 
Charles  Milton  (U.  S.  Geological  Survey), 
"Mineral  assemblages  of  the  Green  River 
formation";  Felix  Chayes  (Geophysical 
Laboratory),  "Diffraction  effects  of  short- 
range  ordering  in  layered  sequences"; 
William  E.  Hanson  (Mellon  Institute  of 
Industrial  Research),  "Some  aspects  of  the 
geochemistry  of  petroleum";  Gustaf  O. 
Arrhenius  (Scripps  Institution  of  Ocea- 
nography), "Sedimentary  processes  and 
records  in  the  ocean";  Sydney  P.  Clark,  Jr. 
(Geophysical  Laboratory),  "Equations  of 
state  and  polymorphism  at  high  pressure"; 
Kenneth  O.  Emery  (University  of  South- 
ern California),  "Sediments  of  California 
marine  basins";  Hessel  de  Vries  (Univer- 
sity of  Groningen),  "Measurement  and 
use  of  natural  radiocarbon." 


Symposium  on  C14  Dating,  Pleistocene 

Stratigraphy,  and  Archaeologic 


A  one-day  conference  devoted  mainly 
to  problems  in  Pleistocene  geology  was 
held  at  the  Geophysical  Laboratory  on 
May  22,  1958.  Professor  Hessel  de  Vries 
(University  of  Groningen,  Netherlands) 
discussed  the  results  of  recent  C14  work 
at  Groningen  with  particular  reference  to 
Pleistocene  geology  in  western  Europe. 
Professor  Friedrich  Brandtner  (currently 
a  Fellow  at  Yale  University)  gave  an  ac- 
count of  his  field  observations  in  the 
Pleistocene  of  western  Europe,  and  com- 
pared them  with  the  C14  results. 

After  a  buffet  luncheon,  Dr.  Ralph 
Solecki  (Smithsonian  Institution)  spoke 
on  C14  dates  from  the  Shanidar  cave  in 
Iraq  and  their  significance  to  Pleistocene 
geology.  Professor  R.  F.  Flint  (Yale  Uni- 
versity) discussed  Pleistocene  problems  in 
North  America,  especially  the  studies  that 
have  been  made  at  Searles  Lake.  The 
glacial  sequence  of  North  America  was 
discussed  by  Dr.  Meyer  Rubin  (U.  S.  Geo- 
logical Survey).  Dr.  W.  S.  Broecker  (La- 
mont  Geological  Observatory)  presented 
C14  dates  from  pluvial  lakes  in  the  Great 
Basin  of  western  North  America.  Profes- 
sor de  Vries  then  reviewed  data  on  the 
fluctuation  of  C14  concentration  in  samples 
of  modern  carbon.  Finally,  Dr.  Irving 
Friedman  (U.  S.  Geological  Survey)  spoke 
on  the  hydration  of  obsidian  artifacts  as  a 
new  dating  tool. 

Besides  personnel  from  the  Geophysical 
Laboratory,  the  conference  was  attended 
by  approximately  twenty-five  representa- 
tives from  the  U.  S.  Geological  Survey,  the 
Smithsonian  Institution,  the  Lamont  Geo- 
logical Observatory,  Yale  University,  and 
the  Peabody  Foundation. 


During  the  report  year  staff  members 
were  invited  to  present  lectures  as  follows : 

P.  H.  Abelson  lectured  at  a  symposium 
on  organic  geochemistry  sponsored  by  the 

American  Petroleum  Institute  at  Dallas, 
Texas;  the  D.  C.  Chapter  of  the  American 
Institute  of  Chemists;  the  Department  of 
Genetics,  Cold  Spring  Harbor,  N.  Y.;  the 
National  Bureau  of  Standards;  a  sym- 
posium on  biochemical  origins  at  the  San 
Francisco  meeting  of  the  American  Chemi- 
cal Society;  the  College  of  Mineral  Indus- 
tries at  Pennsylvania  State  University;  the 
Kansas  Chapter  of  Phi  Sigma  and  the  De- 
partment of  Bacteriology  at  the  University 
of  Kansas;  and  at  the  Army  Conference 
on  Basic  and  Applied  Research  and  Com- 
ponent Development  at  Operations  Re- 
search Office  of  Johns  Hopkins  University. 
F.  R.  Boyd  addressed  the  Washington 
Academy  of  Sciences  and  Howard  Univer- 
sity, D.  C. 

F.  Chayes  lectured  at  the  Department  of 
Geology  and  Geophysics,  Massachusetts 
Institute  of  Technology. 

H.  P.  Eugster  again  served  as  Lecturer 
in  the  Department  of  Geology,  Johns  Hop- 
kins University,  each  Friday  during  the 
academic  year  1957-1958.  He  also  ad- 
dressed the  Department  of  Geology  and 
the  Institute  of  Geophysics  at  the  Univer- 
sity of  California  at  Los  Angeles. 

G.  Kullerud  lectured  at  the  New  York 
meeting  of  the  American  Institute  of 
Mining  and  Engineering. 

N.  Morimoto  presented  a  symposium  on 
the  crystal  structure  of  borax  to  the  Wash- 
ington Crystal  Colloquium. 

J.  F.  Schairer  addressed  a  meeting  of  the 
Philosophical  Society  of  Washington. 

G.  R.  Tilton  gave  several  lectures  on  the 
geochemistry  of  stable  isotopes  at  Johns 
Hopkins  University. 

H.  S.  Yoder  lectured  at  the  Conference 
on  the  Structure  and  Properties  of  Natural 
and  Synthetic  Minerals,  Pennsylvania  State 
University;  the  Sixth  Clay  Conference  at 
the  University  of  California,  Berkeley; 
the  Department  of  Geology,  Michigan 
State  University;  and  the  Geology  Club, 
McGill  University,  Montreal.  He  also  gave 
a  series  of  six  lectures  at  the  Summer 
Institute  in  Geology  for  college  teachers 


at  the  University  of  Illinois.  Three  lectures 
were  given  at  the  Department  of  Geology, 
University  of  Minnesota,  Minneapolis. 

The  "Summary  of  Published  Work"  be- 
low briefly  describes  the  papers  published 
in  scientific  journals  during  the  report 
year.  In  addition,  the  following  papers 
are  now  prepared  for  publication :  G.  Don- 
nay,  "Crystal  data  on  chlorophyll  a"; 
J.  D.  H.  Donnay  and  G.  Donnay,  "Sine 
table  for  indexing  powder  patterns";  G. 
Donnay  and  J.  G.  Smith,  "Calibration 
sights  for  X-ray  powder  camera";  G.  Kul- 

lerud  and  G.  Donnay,  "Natural  and  syn- 
thetic ferroselite:  A  roentgenographic 
mimesis  of  rammelsbergite";  J.  F.  Schairer, 
"Phase  equilibria  in  silicate  systems";  J.  R. 
Smith,  "The  optical  properties  of  heated 
plagioclases";  G.  R.  Tilton,  "Isotopic  com- 
position of  lead  from  tektites";  G.  R.  Til- 
ton,  G.  W.  Wetherill,  G.  L.  Davis,  and 
C.  A.  Hopson,  "Ages  of  minerals  from 
the  Baltimore  gneiss";  O.  F.  Tuttle  and 
N.  L.  Bowen,  "The  origin  of  granites  in 
the  light  of  experimental  studies  in  the 
system  NaAlSi808-K AlSisOs-SiOa-HaO" ; 
H.  S.  Yoder,  "Experimental  studies  on 
micas:    A  synthesis." 


(1269)  The  system  water-nepheline-albite:  A 
theoretical  discussion.  G.  W.  Morey. 
Am.  J.  Sci.,  255,  461-480  (1957). 

At  an  invariant  or  quintuple  point  in  a 
three-component  system,  five  pressure-temper- 
ature curves  of  univariant  equilibria  intersect. 
The  sequence  of  these  P-T  curves,  that  is,  the 
order  in  which  they  are  met  in  either  their 
stable  portions  or  their  metastable  prolonga- 
tions, is  determined  by  the  compositions  of 
the  phases  at  that  point.  This  proposition  is 
proved,  and  its  application  is  illustrated  by 
a  discussion  of  possible  P-T  curves  and  in- 
variant points  in  the  system  water-nepheline- 

(1270)  Isotopic  ages  of  zircon  from  granites 
and  pegmatites.  G.  R.  Tilton,  G.  L. 
Davis,  G.  W.  Wetherill,  and  L.  T.  Al- 
drich.  Trans.  Am.  Geophys.  Union,  38, 
360-371  (1957). 

Isotopic  lead  age  determinations  have  been 
made  on  13  zircons  obtained  from  rocks  185 
to  1400  million  years  old.  Concordant  or 
nearly  concordant  ages  are  found  for  all  the 
samples  containing  no  detectable  common 
lead,  and  discordant  ages  are  found  for  most, 
if  not  all,  of  the  samples  containing  common 
lead.  A  comparison  is  made  between  the  con- 
cordant isotopic  age  patterns  given  by  3  zir- 
cons from  the  Grenville  subprovince  in  On- 
tario and  the  discordant  patterns  given  by  3 
zircons  from  the  Cordilleran  region  of  western 
United  States.  This  comparison  indicates 
that  the  discordant  ages  can  be  related  to  the 

recent  orogenies  that  occurred  in  the  Cordil- 
leran region.  The  Grenville  is  a  stable  shield 
area.  There  is  no  relation  between  the  agree- 
ment of  the  isotopic  ages  of  zircon  and  crystal 
size,  amount  of  radiation  damage,  or  optical 
appearance — that  is,  zoning,  cloudiness,  or  in- 

When  a  discordant  age  result  is  compared 
with  the  potassium-argon  and  rubidium-stron- 
tium ages  of  associated  mica  the  Pb207-Pb206 
age  is  found  to  be  the  closest  to  the  mica  age. 
Isotopic  ages  are  compared  with  the  simpler 
a-lead  and  chemical-lead  ages,  which  do  not 
require  isotopic  analysis  of  lead.  The  noniso- 
topic  ages  are  approximately  correct  for  zir- 
cons that  have  concordant  isotopic  ages  but 
are  in  error  when  discordant  isotopic  ages  are 
found.  No  explanation  is  offered  as  to  why 
the  mica  ages  are  apparently  unaffected  by 
the  process  or  processes  that  altered  the  zir- 
con ages.  An  understanding  of  this  phe- 
nomenon would  doubtless  provide  valuable 
information  about  the  post-crystallization  his- 
tory of  the  samples. 

(1271)  Melting  relations  of  the  common  rock- 
forming  oxides.  J.  F.  Schairer.  /.  Am. 
Ceram.  Soc,  40,  215-235  (1957). 

Geological  science  is  concerned  with  the 
nature  of  the  minerals  of  the  earth  and  par- 
ticularly with  the  processes  by  which  earth 
materials  have  been  changed  and  modified. 
Laboratory  studies  of  the  melting  behavior  of 
the  common  rock-forming  oxides  have  been 
an  important  adjunct  to  the  observations  of 


the  field  geologist.  For  the  past  fifty  years 
investigators  at  the  Geophysical  Laboratory 
have  been  obtaining  quantitative  information 
on  the  melting  relations  of  many  of  the  im- 
portant rock-forming  minerals.  These  studies 
of  the  fundamental  chemistry  of  igneous  and 
metamorphic  rocks  have  yielded  much  infor- 
mation of  value  to  ceramists,  metallurgists, 
and  mineral  technologists.  This  paper  sum- 
marizes the  most  important  phase-equilibrium 
studies  of  unary,  binary,  ternary,  quaternary, 
and  portions  of  quinary  systems  of  the  com- 
mon rock-forming  oxides — Si02,  A1203,  FeO, 
Fe2Os,  CaO,  MgO,  Na20,  and  K20. 

(1272)  The  crystalline  modifications  of  NaAl- 
Sio08.  W.  S.  MacKenzie.  Am.  J.  Sci., 
255,  481-516  (1957). 

X-ray  studies  of  synthetic  albites  reveal  a 
wide  variation  in  lattice  parameters  of  the 
crystals,  depending  on  the  conditions  of  crys- 

A  glass  of  composition  NaAlSi3Os  has  been 
crystallized  in  the  presence  of  water  vapor 
under  pressure,  at  temperatures  between  450° 
and  1000°  C  for  varying  periods  of  time. 
X-ray  study  of  the  crystals  reveals  that  at  each 
temperature  the  crystals  formed  after  a  few 
hours  are  similar  to  high-temperature  albite 
in  lattice  parameters.  Experiments  of  longer 
duration,  however,  show  that  the  lattice  pa- 
rameters of  the  initially  formed  crystals 
change  gradually,  finally  reaching  a  steady 
value  characteristic  of  the  temperature  of  crys- 
tallization. The  results  suggest  that  for  each 
temperature  there  is  a  stable  crystalline  form 
of  NaAlSi3Og  intermediate  between  high- 
temperature  albite  and  low-temperature  al- 
bite, high-temperature  albite  being  stable 
only  above  about  1000°  C  and  low-tempera- 
ture albite  only  below  about  450°  C. 

Previous  investigations  have  indicated  that 
some  natural  potassium  feldspars  may  have 
crystallized  metastably  as  the  high-tempera- 
ture monoclinic  modification  sanidine,  and 
subsequently  inverted  to  the  low-temperature 
triclinic  modification  microcline.  The  experi- 
ments reported  here  support  a  similar  conclu- 
sion in  respect  to  sodium  feldspar,  namely 
that  some  natural  albites  now  in  the  low- 
temperature  form  may  have  crystallized  meta- 
stably in  the  high-temperature  form  and  grad- 
ually inverted  to  the  low-temperature  form. 
The    variations   in   the   properties   of  albites 

from  low-temperature  veins  of  Alpine  type 
are  consistent  with  this  interpretation. 

(1273)  Olivine  X-ray  determinative  curve.  H.  S. 
Yoder,  Jr.,  and  Th.  G.  Sahama.  Am. 
Mineralogist,  42,  475-491  (1957). 

The  (130)  spacing  of  31  chemically  ana- 
lyzed natural  olivines  and  7  synthetic  olivines 
has  been  measured.  A  determinative  curve 
has  been  calculated  from  26  of  the  chemically 
analyzed  natural  olivines: 

Fo   (mole   per  cent)  =  4233.91  -  1494.59i130. 

The  fictive  end  points  are  d130  (Fo  =  100) 
=  2.7659  and  d130  (Fo  =  0)  =2.8328.  The 
error  attached  to  an  individual  measurement 
ranges  from  3  to  4  mole  per  cent,  depending 
on  the  composition. 

Portions  of  the  powder  X-ray  diffraction 
patterns  for  synthetic  forsterite  and  synthetic 
fayalite  have  been  indexed.  The  cell  con- 
stants, density,  and  molar  volumes  are  given. 

(1274)  Some  aspects  of  paleobiochemistry.  P.  H. 
Abelson.  Annals  N.  Y.  Acad.  Sci.,  69, 
276-285  (1957).