Skip to main content

Full text of "Year book - Carnegie Institution of Washington"

See other formats

/?. &. i/Wfl^- 

J)jA.e trfi+S't 




Year Book 66 


Library of Congress Catalog Card Number 3-16716 
Garamond/Pridemark Press, Baltimore, Maryland 



Officers and Staff iv 

Report of the President 1 

Reports of Departments and Special Studies 1 

Department of Terrestrial Magnetism 3 

Committee on Image Tubes for Telescopes 133 

Department of Plant Biology 141 

Mount Wilson and Palomar Observatories 249 

Geophysical Laboratory 317 

Department of Embryology 573 

Genetics Research Unit 645 

Bibliography 673 

Report of the Executive Committee 675 

Report of Auditors 677 

Abstract of Minutes of Sixty-Ninth Meeting of the Board of Trustees 693 

Articles of Incorporation 695 

By-Laws of the Institution 699 

Index 705 


President and Trustees 


Caryl P. Haskins 


James N. White 

Henry S. Morgan 

Garrison Norton 

Eric Ashby 
Amory H. Bradford 
Omar N. Bradley 
Vannevar Bush 
Carl J. Gilbert 
Crawford H. Greenewalt 
Caryl P. Haskins 
Alfred L. Loomis 
Robert A. Lovett 
William McC. Martin, Jr. 
Keith S. McHugh 
Margaret Carnegie Miller 1 
Henry S. Morgan 
Seeley G. Mudd 
William I. Myers 
Garrison Norton 
Richard S. Perkins 
Elihu Root, Jr. 2 
William W. Rubey 
Frank Stanton 
Charles P. Taft 
Charles H. Townes 
Juan T. Trippe 
James N. White 

1 Resigned May 5, 1967. 

2 Resigned May 5, 19G7; died August 27, 1967. 


Trustees (continued) 


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


Henry S. Morgan, Chairman 
Carl J. Gilbert 
Crawford H. Greenewalt 
Caryl P. Haskins 
Robert A. Lovett 
William I. Myers 
Garrison Norton 
Richard S. Perkins 
James N. White 


Omar N. Bradley, Chairman 
Garrison Norton 
Richard S. Perkins 
Frank Stanton 



Seeley G. Mudd, Chairman 
Amory H. Bradford 
Crawford H. Greenewalt 

Richard S. Perkins, Chairman 
Crawford H. Greenewalt 
Alfred L. Loomis 
Keith S. McHugh 
Henry S. Morgan 


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


Carl J. Gilbert, Chairman 
Crawford H. Greenewalt 
Keith S. McHugh 
James N. White 


Juan T. Trippe, Chairman 
Richard S. Perkins 




SIS Santa Barbara Street 
Pasadena, California 91106 

Horace W. Babcock, Director 

Halton C. Arp 

Ira S. Bowen, Distinguished 

Service Staff Member 
Edwin W. Dennison 
Armin J. Deutsch 
Jesse L. Greenstein 
Robert F. Howard 
Robert P. Kraft 
Robert B. Leighton 
Guido Munch 
J. Beverley Oke 
Bruce H. Rule 
Allan R. Sandage 
Wallace L. W. Sargent 
Maarten Schmidt 
Arthur H. Vaughan, Jr. 
Olin C. Wilson 
Harold Zirin 
Fritz Zwicky 


2801 Upton Street, N.W. 
Washington, D.C. 20008 

Philip H. Abelson, Director 
Peter M. Bell 
Francis R. Boyd, Jr. 
Charles W. Burnham 1 
Felix Chayes 
Gordon L. Davis 
Gabrielle Donnay 
Joseph L. England 
P. Edgar Hare 
Thomas C. Hoering 
Thomas E. Krogh 2 
Gunnar Kullerud 
Donald H. Lindsley 
J. Frank Schairer 
Hatten S. Yoder, Jr. 


5241 Broad Branch Road, N.W. 
Washington, D.C. 20015 

Ellis T. Bolton, Director 
L. Thomas Aldrich, Associate 

Roy J. Britten 
Louis Brown 
Dean B. Cowie 
Scott E. Forbush 
W. Kent Ford, Jr. 
Stanley R. Hart 3 
David E. Kohne 
Richard B. Roberts 
Vera C. Rubin 
I. Selwyn Sacks 
T. Jefferson Smith 
John S. Steinhart 
Kenneth C. Turner 

1 Resigned August 31, 1966. 
- Appointed August 1, 1965. 
1 On leave of absence from September 1, 1966. 


Staff (continued) 


Stanford, California 94305 

C. Stacy French, Director 

Olle Bjorkman 

Jeanette S. Brown 

Jens C. Clausen, Emeritus 

David C. Fork 

William M. Hiesey 

Malcolm A. Nobs 

James H. C. Smith, Emeritus 


115 West University Parkway 
Baltimore, Maryland 21210 

James D. Ebert, Director 
David W. Bishop 
Bent G. Boving 
Donald D. Brown 
Igor B. Dawid 
Robert L. DeHaan 
Elizabeth M. Ramsey 


Cold Spring Harbor 
New York 1172k 

Alfred D. Hershey, Director 
Elizabeth Burgi 
Barbara McClintock 

Cytogenetics Laboratory 
Ann Arbor, Michigan 

Helen Gay 


Staff (continued) 


15S0 P Street, AMI"., Washington, B.C. 20005 

Caryl P. Haskins President 

Edward A. Ackerman Executive Officer 

James W. Boise Bursar; Secretary -Treasurer, Retirement Trust; 

Executive Secretary to the Finance Committee 

Marjorie H. Walburn Assistant to the President 

Donald J. Patton Director of Publications 

Eleanor F. Peck Editor 

Kenneth R. Henard Assistant Bursar; Assistant Treasurer, 

Retirement Trust 

Sheila A. McGough Associate Editor 

Joseph M. S. Haraburda Assistant to the Bursar 

Elizabeth R. Epstein Assistant to the Director of Publications 

Marshall Hornblower Counsel 


Tatiana Proskouriakoff 
Anna 0. Shepard 


Staff (continued) 


Carnegie Research Associate 

J. D. McGee 

Imperial College of Science and Technology, University of London 

Research Associates of the Carnegie Institution 

Richard A. Chase 

Johns Hopkins University 

Louis B. Flexner 

University of Pennsylvania 

Harry E. D. Pollock 

Carnegie Institution 

C. E. Tilley 

Cambridge University 


Former Presidents and Trustees 


Daniel Coit Gilman, 1902-1904 John Campbell Merriam, 1921-1938 ; 

Robert Simpson Woodward, 1904-1920 President Emeritus 19 39-19 A5 

Vannevar Bush, 1939-1955 

Alexander Agassiz 
George J. Baldwin 
Thomas Barbour 
James F. Bell 
John S. Billings 
Robert Woods Bliss 
Lindsay Bradford 
Robert S. Brookings 
John L. Cadwalader 
William W. Campbell 
John J. Carty 
Whitefoord R. Cole 
Frederic A. Delano 
Cleveland H. Dodge 
William E. Dodge 
Charles P. Fenner 
Homer L. Ferguson 
Simon Flexner 
W. Cameron Forbes 
James Forrestal 
William N. Frew 
Lyman J. Gage 
Walter S. Gifford 
Cass Gilbert 
Frederick H. Gillett 
Daniel C. Gilman 
John Hay 

Barklie McKee Henry 
Myron T. Herrick 
Abram S. Hewitt 
Henry L. Higginson 
Ethan A. Hitchcock 
Henry Hitchcock 
Herbert Hoover 
William Wirt Howe 
Charles L. Hutchinson 
Walter A. Jessup 
Frank B. Jewett 
Samuel P. Langley 
Ernest 0. Lawrence 
Charles A. Lindbergh 
William Lindsay 
Henry Cabot Lodge 



Seth Low 



Wayne MacVeagh 



Andrew W. Mellon 



Margaret Carnegie Miller 



Roswell Miller 



Darius O. Mills 



S. Weir Mitchell 



Andrew J. Montague 



William W. Morrow 



William Church Osborn 



James Parmelee 



Wm. Barclay Parsons 



Stewart Paton 



George W. Pepper 



John J. Pershing 



Henning W. Prentis, Jr. 



Henry S. Pritchett 



Gordon S. Rentschler 



David Rockefeller 



Elihu Root 



Elihu Root, Jr. 



Julius Rosenwald 



Martin A. Ryerson 



Henry R. Shepley 



Theobald Smith 



John C. Spooner 



William Benson Storey 



Richard P. Strong 



William H. Taft 



William S. Thayer 



James W. Wadsworth 



Charles D. Walcott 



Frederic C. Walcott 



Henry P. Walcott 



Lewis H. Weed 



William H. Welch 



Andrew D. White 



Edward D. White 



Henry White 



George W. Wickersham 



Robert E. Wilson 



Robert S. Woodward 



Carroll D. Wright 


Under the original charter, from the date of organization until April 28, 1904, the follow- 
ing 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. 


There is in our community an elderly painter of seascapes who knows 
the sea so well that he no longer goes to look at it while he paints. He 
dislikes intensely the work of a young painter ivho sets his easel on 
the beach and paints things his elder does not remember having seen. 

John Steinbeck — Between Pacific Tides 

We have all some responsibility for action, some area, however small, 
in which each of us and he alone can play the part of agent. There is 
a second field, wider and not congruent ivith the first, in which each 
of us can contribute to the making of policy. There is a third, wider 
still, in which each of us has power to give or ivithhold assent to the 
policy decisions of others. There is a fourth, yet wider, in which the 
only responsibility of each of us is the neglected but important re- 
sponsibility of giving or withholding the trust which supports or 
inhibits our fellows in the exercise of their inalienable responsibilities, 
as their trust or distrust supports or inhibits us. There is ... a fifth 
field . . . the creative function which shapes the work thus and not 
otherwise, ivhether the work be a building or an institution, a nation's 
history or a human life. Here lies the possibility for the vision that is 
manifest, for good or ill, ivhenever a 'state of art' is imposed on a 
l state of nature' . . . the authentic signature of the human mind. 

Geoffrey Vickers — The Art of Judgment 

. . . the educational and scientific estate . . . gains power in a socially 
complex society from its capacity for social invention. And while its 
poiver must rest on its ability to attract the support of attached and un- 
attached individuals, in the future its numbers will command respect. 

John Kenneth Galbraith— The New Industrial State 

present which, as yet, we are not well prepared to understand, 
much less to meet. They involve such fundamental issues as the real 
nature of our innovative society and the stage of evolution that it may 
presently have reached — and the optimum ways in which science and 
technology may contribute to its health and growth. What do we truly 
most wish to achieve in our current level of governmental support for 
science? Is it primarily to accelerate yet further the rate of discovery 
and the winning of new knowledge? Is it primarily to encourage and in- 
crease the development of additional centers of excellence in research 
and in teaching, academic and other, in the nation? Is it primarily to 
build yet more rapidly and more massively our resources of scientifically 
highly trained and talented young men and women, on whose shoulders, 
tomorrow, so much of our national prosperity, and indeed, of national 
survival itself, must rest? Is it to architect new and more effective 




means for providing and managing those massive services of a technical 
kind — in health, in commerce, in welfare — the need for which presses 
upon us so insistently today and may well prove still more insatiable 
tomorrow? Do our primary interests, in fact, include all of these ele- 
ments — evidently extremely divergent in their requirements and in- 
deed at times conflicting in their demands — and yet more? Are we 
primarily interested in that equality of opportunity upon which our 
whole philosophy and our very identity as a people have so long rested? 
Or are we most concerned with that luminous excellence whose existence 
continues to provide today, as it has throughout our past, our best 
guarantee of national health and development; upon which our control 
of our own destiny has so predominantly rested? Or are we still, as we 
have given repeated evidence in the past, equally interested in both? 
And if so, how are we going to manage the inconsistencies of action 
inherent in those judgments of value : inconsistencies that have always 
troubled us, but which now, in our current massive and complex world, 
are both far more formidable and demand workable solutions more 
importunately than ever before? In such a context, we have rarely been 
vouchsafed more illuminating views of the magnitude and the depth 
and the complexity of these issues, and of the elements of paradox 
within them, than during the year just past. 

On December 16, 1966 — the year when the Congress of the United 
States finally decided to reject the completion of Project Mohole, the 
attempt to drill through the ocean bottom to the region of the Moho- 
rovicic discontinuity — the Atomic Energy Commission announced that 
the location of its 200-billion-electron-volt accelerator, estimated to 
cost $375 million to build and $60 million a year to operate, would be 
the village of Weston, Illinois, situated near Chicago. One hundred and 
twenty-six proposals for location had been received, submitted from 
forty-six of the fifty states. The award followed the longest and most 
intense public discussion and political struggle in our history involving 
the approval and the locating of a Federal scientific facility, and feelings 
and opinions about it were as intense and as mixed as its magnitude and 
nature would have predicted. As a climax, it became apparent in mid- 
1967 that local authorities in Illinois had failed to give adequate assur- 
ance that important social considerations, notably the provision of ap- 
propriately integrated housing and living facilities in the vicinity of 
the giant machine, would be sufficiently guaranteed. Faced with this 
serious omission, and perhaps in some doubt that national expenditures 
of this order for this purpose were wholly prudent at a moment in 
history when acute perils confronted the nation both abroad and at 


home and the demands for men and dollars to meet them were crucial, 
Congress debated the matter intensely and at length. For a time, the 
Congress seemed inclined to conclude that it was not justifiable to pro- 
ceed with the project. Finally — but once again only after intense dis- 
cussion — an affirmative judgment was rendered. 

In the long view, the correctness of this specific decision seems less 
significant than the syndrome of difficulties, trauma, and widespread 
uncertainty that attended its taking; in particular it seems less im- 
portant than the issues that proved of critical significance in judging 
a wholly scientific project which was, however, so costly that it could 
be accomplished only with extensive Federal support. Primary among 
them, necessarily, were issues related to the order of social imperatives 
in our nation and of the economic priorities in our society. Never in our 
recent history, perhaps, have the range and character of the problems 
surrounding the architecting of Federal support of science on the scale 
characteristic of our time been so dramatically underlined in a single 
decision — nor, perhaps, so prophetically limned for the future — as in 
this year just past. And when these developments are viewed against 
the background of certain other factors affecting the Federal support 
of science, the outlines of a contemporary dilemma are yet more sharply 
focused. Overall expenditures in the United States for research and 
development in 1966 have been estimated at $23 billion, with the Fed- 
eral Government accounting for slightly over seventy per cent of the 
whole. This is a formidable total. But within the past several years, 
and notably in the present one, the slope of the curve of growth of 
Federal support for research and development has flattened markedly, 
approximating a budgeted increase of five per cent. At the same time, 
the research costs themselves have risen at least as rapidly, and perhaps 
considerably more so. Put in a specific context, these increases seem 
even more dramatic. Thus Dael Wolfle has recently estimated that a 
dollar spent in research in 1965 purchased only eighty-two per cent 
of the technical work of a dollar spent in 1960, only fifty-seven per cent 
of that accomplished per dollar in 1955, only forty-one per cent of that 
in 1950. Though the nation's expenditures for research and development 
were, in dollar terms, eight or nine times as great in 1965 as fifteen 
years earlier, yet the actual volume of research effort may have in- 
creased by a factor of only three or four. The budgeted increase from 
1966 to 1967 in Federal expenditures for research, indeed, barely keeps 
ahead of the rise in cost of doing research, and, for all the costs of 
Federally supported research and development, it may actually fall 

It is natural to suppose that this slower rate of growth in the dollar 
volume of Federally supported research and development in our nation 


is directly — and only — a consequence of the immediate and importunate 
demands of other massive involvements both domestic and foreign, 
and that, were these difficulties somehow resolved, older growth rates 
would, more or less automatically, be restored. Yet these trends are by 
no means confined to our own country, with our own peculiar problems. 
Almost identical circumstances, for example, confront Great Britain 
today, and it is probable that they are currently besieging all or nearly 
all of the industrialized nations of the contemporary world — except 
possibly Japan. They suggest that some deeper social circumstances 
may underlie the trend. Are there in fact deeper factors that are rele- 
vant here? In considering such a difficult, complex, and overween- 
ingly important general question, a dimension of critical importance 
is provided by some of the events of the year in quite a different dimen- 
sion — that of the substantive achievements of science. 

Against the background of these difficult and challenging questions 
and problems surrounding the Federal support and deployment of huge 
sums and programs of science and technology in the nation that have 
been emphasized so forcibly during the current year, and by way of 
sharpest contrast, it is worth recalling some scientific events in a differ- 
ent context which have also marked this year. It was in 1960 that an 
attempt was made at the great British radio observatory at Jodrell 
Bank to resolve the smallest angular diameters of the then quite novel 
radio sources of the sky by interferometric methods with a base line ex- 
tended to 32,000 wavelengths. This attempt was, to the surprise of 
many, unsuccessful, indicating that some radio sources apparently could 
intercept less than three arc seconds in the sky. It was already known 
that at least three of these anomalous sources of radio energy appeared 
to coincide in their positions in the heavens with three optically iden- 
tified starlike objects showing pronounced emission in the blue. But 
only when that coincidence had been precisely established was it con- 
vincingly demonstrated that the radio emitters and the visible heavenly 
objects could be the same. Then came the critical observation of the 
spectrum of one of them, 3C 48, by Allan Sandage, which disclosed most 
unusual and hitherto unidentified broad emission lines. Three years 
later at the Australian radio observatory at Parkes the astronomical 
object 3C 273 was identified with a radio source determined in pre- 
cision to one arc second in the sky. There followed the striking identi- 
fication of its emission lines as of the Balmer series from hydrogen, 
but shifted by a factor of 1.158 toward the red end of the spectrum. 

Since that germinal discovery, about 150 quasi-stellar radio sources 
have been identified, and it is thought that 100,000 radio-quiet quasi- 


stellar sources, the "QSO's," ranging down to a magnitude limit of 
19.7 in blue light, may exist over the whole sky. To the limit of the 
3C catalog, the "quiet" sources outnumber those which are radio 
emitters by about five hundred to one. This year the discovery, identi- 
fication, and analysis of a new quasar which indicates in its spectrum 
the greatest redshif t so far recorded were accomplished by a combination 
of studies, on the radio side at the Parkes Observatory in Australia, and 
on the optical at the Mount Wilson and Palomar Observatories and the 
Lick Observatory in California. Its emission spectral lines are shifted 
toward the red by a factor of 3.223. If this redshift by Quasar Parkes 
0237-23 is interpreted entirely as a Doppler effect, the quasar stands 
as the brightest object known in the universe, and the most distant from 
us, receding at 82.4 per cent the velocity of light, compared with a 
calculated velocity of 81.4 per cent for the most distant object previously 
known, Quasar 1116 + 12. 

One of the most arresting features of the quasars has been their strik- 
ingly small apparent volumes, considering their immense apparent 
distances and their brightness in the sky. Parkes 0237-23, for instance 
appears to be of optical magnitude 16.6. A year ago a spectacular dis- 
covery, made at the Mount Wilson and Palomar Observatories, revealed 
that the light from Quasar 3C 446 had increased in brilliance twenty- 
fold over a period measured merely in months. This year the phenome- 
non was confirmed by new evidence, again from the Mount Wilson and 
Palomar Observatories. Quasar 3C 279 was found capable of increasing 
or decreasing in brightness by twenty-five per cent within a period of 
twenty-four hours. 

What is the nature of these extraordinary objects that are appar- 
ently so incredibly distant from us, which appear, comparatively, to 
be so small, and which yet radiate such extraordinary amounts of 
energy? Calculations involving the apparent distance and the observed 
intensity over the radio spectrum of Quasar 3C 273, assigning it an 
arbitrary life-span of 10 5 years, suggest that the energy emitted is 
equivalent to that which would be obtained by converting to helium, by 
normal solar processes, a mass of hydrogen equivalent to that con- 
tained in 10 9 suns. But in a whole galaxy there are estimated to be only 
10 11 stars, and even assuming that energy equivalent to that in a whole 
galaxy is involved, it is difficult to find a physical process of energy 
conversion efficient enough to account for this result. 

Do quasars then represent the whole mass of a normal galaxy, 
gravitationally compressed within a volume of diameter equal to only 
about the distance between our sun and the star nearest us? Or is it 
possible, as has recently been inferred from the rapid fluctuation in 
their apparent brightness, that quasars are the nuclei of galaxies actu- 


ally being formed, including a number of large exploding stars sur- 
rounded by a swarm of cool condensations of stellar material that pass 
randomly across the bright source, thus occluding it for us? Is all of 
the fantastic redshift of quasars attributable to the normal Doppler 
effect, and are the calculated almost incredible distances from us 
therefore substantially correct? Or is it possible that some proportion 
of the observed redshift is to be ascribed to gravitational effects — and if 
so, how much? What is the significance of the recent finding that, in 
some of the quasars apparently most distant, the speeds of recession 
indicated by the redshifts of their emission and absorption spectra 
seem not to agree? Is the assigned lifetime of 10 5 years too great? 

Perhaps no mystery of our universe today is deeper or more strik- 
ingly intriguing than that presented by the quasars. Surely no frontier 
of physical knowledge in our time offers higher potential for conceptual 
advances so great as to be, in effect, revolutions of kind. And in the 
opening context, it is especially noteworthy that two of the most spec- 
tacular recent advances on that frontier, and perhaps among the most 
significant — the demonstration of the greatest redshift value so far 
detected, and the demonstration of the extraordinary apparent short- 
term variations in energy output of which the nuclei of some quasars 
are capable — have been accomplished or consolidated during the past 

If the current year has brought critical advances at the frontiers of 
our knowledge of physical science and in concepts of our universe on 
both the conceptual and the experimental sides, this has been no less 
true of research in the areas of life and of mind. Ever since Oparin's 
classic work, for instance, first published in book form in the Soviet 
Union in 1923 and brought to wider notice in the West with the publi- 
cation of the English version of The Origin of Life in 1937, conviction 
has been widespread that life as we know it — with its extraordinary 
panoply and precision of protein and nucleic acid, its intricate and 
wonderfully coordinated systems of information storage and informa- 
tion exploitation, of growth and regulation and direction and replica- 
tion implicit in the functioning of systems of DNA and RNA and of 
the enzymes and their products, had a terrestrial origin in much sim- 
pler protobiological systems. Various experimental demonstrations of 
the possibility of the synthesis of amino acids from inorganic constitu- 
ents in the laboratory, and even of the artificial synthesis of proteins, 
have in recent years provided strong reinforcement for such conclusions. 

Yet the formulation of thoroughly satisfactory notions of precise 
chemical paths through which the earliest precursors of living things 


might have arisen lagged well behind the experimental findings. An 
early picture of the primordial oceans as a "thick soup," a puree of many 
amino acids and related compounds of which the materials of life were 
ultimately synthesized, went for some time critically unchallenged. 
Similarly, the hypothesis that the primordial terrestrial atmosphere, 
before it acquired its present high proportion of oxygen primarily de- 
rived from and maintained by the photosynthetic activity of green 
plants, consisted largely of methane and ammonia — a concept dating 
from an early stage in the development of modern theories about the 
origin of life — was held a basic premise. Only in the year just past has 
it been convincingly expounded that this assumption seriously con- 
tradicts well-established geological evidence that volatile substances, 
produced by outgassing from rocks early in the history of the earth, 
reacted with an alkaline crust to form alkaline seas and an atmosphere 
of which the primary constituents were nitrogen, hydrogen, carbon 
monoxide, and carbon dioxide — very different components from those 
earlier envisaged. When such a mixture is exposed to radiant energy 
in the laboratory, hydrogen cyanide can be produced. And hydrogen 
cyanide itself, when irradiated with ultraviolet light, evolves significant 
yields of amino acids, the building blocks of the proteins of living things. 
Thus, during the year just past, one of the most active and stimulating 
frontiers of biological research has been illumined by a concept of 
process much more in harmony with views of the chemical nature of 
the early earth and its oceans and atmosphere that are widely accepted 
by geologists today. 

For two hundred years, a curious nervous disease of sheep has been 
widely recognized. It has also been greatly feared, for it appears to be 
markedly contagious, and once established in an animal its progress 
seems inexorable, leading in all but the rarest cases to collapse and 
death. In France, the disease has long been known as "la tremblante," 
reflecting the hypersensitivity, the shuddering, and the increasing loss 
of muscular coordination that mark its progress, strongly suggestive of 
an extensive involvement of the nervous system. In England the same 
disease is known as "scrapie" from the intense itching that character- 
izes its early stages, resulting in the patchy loss of wool from constant 
rubbing against objects of all kinds — a further suggestion of the early 
involvement of the peripheral nervous system. 

Despite the two centuries that this disease has preyed on men's flocks 
and has been a constant dread to the shepherd, despite fifty years of 
research into its cause, despite especially intensive investigations of 
the past decade, the nature of scrapie remained elusive and unresolved. 


In 193S two investigators, J. Cuille and P. L. Chelle, reported 
that they had been able to infect experimental animals with material 
which passed through a filter with pores fine enough to screen the 
smallest bacteria. It was then widely believed that the agent of scrapie 
was a virus: a virus with a period of incubation extraordinarily long, 
varying from four months to four years for sheep, and so influenced 
by immunity factors apparently genetically mediated that in any 
randomly selected group of exposed animals only five to thirty per cent 
ever developed the disease. Yet over a period of ten years studies of 
scrapie in tissue culture pursued in several laboratories consistently 
failed to demonstrate the existence of any viral agent. Serological 
studies failed to demonstrate the production of antibodies in affected 
animals, or scrapie antigen in scrapie infective material, regardless of 
its nature. Consistently, studies with the electron microscope of tissues 
and extracts proved to carry scrapie infection failed to demonstrate the 
presence of any virus particles. It was found, moreover, that the infec- 
tive agent would withstand exposure to concentrations of formalin fatal 
to conventional viruses. In one experiment, indeed, brain tissue taken 
from an affected animal and preserved in ten per cent formalin for 
periods ranging from six to twenty-eight months proved capable of 
transmitting the disease. As added tokens of a refractoriness so great 
that it is hard to associate w T ith a living organism at all, heating to one 
hundred degrees centigrade for half an hour did not destroy the infec- 
tious power of affected tissues, nor did exposure to a five-per-cent con- 
centration of chloroform at thirty-seven degrees centigrade for thirteen 
days, nor to a concentration of phenol of two per cent for the same 
length of time under the same conditions. Rapid freezing and thawing, 
exposure to ribonuclease or deoxyribonuclease, ether, and chloroform- 
methanol were equally without effect. 

During this year, experiments have been reported which strongly 
suggest that the infective "particle" of scrapie, if such indeed it is, may 
be smaller than the smallest virus. A study of size by means of ionizing 
radiation, undertaken collaboratively in the Radiopathology Research 
Unit of the British Medical Research Council at Hammersmith Hos- 
pital in London and the Department of Pathology at the Institute for 
Research on Animal Diseases of the British Agricultural Research 
Council at Compton in Berkshire resulted in an estimate of a particle of 
about seven millionths of a millimeter in diameter — substantially 
smaller than the smallest virus particle recorded. Irradiation with 
ultraviolet light of a wavelength specifically absorbed by nucleic acids 
in exposures that normally destroy their function made no difference 
in the infective power of the scrapie particles, suggesting that they 
may lack nucleic acid. What, then, is the nature of this seemingly new 


kind of disease agent that can be passed from organism to organism by 
simple injection, like a normal infection, which obviously can increase 
in concentration with time in the infected animal as though it multi- 
plied like a virus, and that, like a normal organism of disease, produces 
a characteristic and reproducible — and in this case devastating — syn- 
drome? There is no clear answer as yet. Only the future can show how 
far the repercussions of this new frontier of medicine and of theoretical 
biology will develop. But the possibilities range from a further impact 
on theories of the origin of terrestrial life to notions of the causes of 
abnormal cell development to an enhanced understanding of a group 
of dread human diseases the nature of which, today, is little compre- 

One of the major obstacles to understanding how life on earth could 
have originated through the chemical paths now generally visualized 
is that, though the synthesis of proteins under conditions approxi- 
mating those believed to have prevailed on the primitive earth can be 
readily conceived, it is less easy to imagine such processes for either 
ribonucleic or deoxyribonucleic acid. Yet the very earliest life systems, 
however simple and crude they may have been by modern standards, 
must nonetheless have been highly adjusted and regulated relative to 
nonliving matter, else they could not have persisted at all. What then 
could have been the regulators? Could they, in the first instance, have 
been relatively simple proteins, performing vital tasks, to be sure, but 
much less complex than those that nucleic acids were later required to 
handle? An observation made by I. H. Pettison and D. A. Haig in the 
course of a study of the behavior of nerve cells from the brains of 
animals killed in advanced stages of scrapie and maintained in organ 
culture gives pause in this context as well as in that of abnormal 
cellular development in general. In several cases dividing nerve cells 
were thought to have been detected in these cultures, although nerve 
cells in cultures of normal brain tissue of adult animals are ordinarily 
of a strictly nondividing habit. It seems possible that the scrapie- 
inf ected nerve cells may have acquired an abnormal growth mechanism 
at least superficially reminiscent of that of a cancerous cell. 

This surprising interpretation has been reinforced by a parallel and 
remarkable finding announced this year by Gurdon. When the nuclei 
of adult nerve cells from the brain of the clawed toad Xenopus laevis 
were injected into developing toad oocytes, it was found that in mature, 
unfertilized eggs, DNA synthesis might be resumed by the adult nerve- 
cell nucleus, under the influence, apparently, of a cytoplasmic factor 
which appears in the egg a few hours after the administration of pitui- 
tary hormone to the animal. It has long been a classic concept among 
cell physiologists that the power of growth and division is permanently 


lost at maturity. The idea that it is only inhibited is new and radical. 
But it has been foreshadowed for several years by work in the De- 
partment of Embryology, described elsewhere in this volume, in which 
it has proved possible to "reactivate" the synthesis of DNA in mature 
muscle cells infected with Rous sarcoma virus. This effect, moreover, 
has been confirmed by Yaffe at Rehovoth in cultures of mature muscle 
cells infected with polyoma virus. It seems probable that what we con- 
sidered so long to be a permanent loss of the power of specialized cells 
to grow and divide in higher animals at their maturity may in fact 
represent only the repression involved in an exquisitely regulated cellu- 
lar system. 

There are some resemblances between the behavior of the presumed 
infective particle of scrapie and an agent involved in a condition which 
has been termed "allergic encephalomyelitis." A small basic protein, 
thought by some to be a polypeptide, has been isolated from the central 
nervous system of an animal exhibiting nervous symptoms resembling 
those of true encephalomyelitis. This protein, when injected into the 
footpad of a guinea pig, results two or three weeks later in an acute 
reaction in the nervous system of the animal, terminating in paralysis 
or even death. This "allergic encephalomyelitis" does not seem to be 
infectious, and it may be only a single episode of transfer. In this it is 
unlike scrapie. But it is like the scrapie agent in its remarkable resist- 
ance to high temperatures, formalin, solvents, and ultraviolet light. 
Does it belong in the same class of substances as the scrapie agent? 

It seems possible that future findings on the nature of scrapie could 
shed light on the fundamental nature of that group of important human 
diseases of the central nervous system such as multiple sclerosis where, 
as in scrapie, there is slow destruction of that system. Particularly 
striking is the possibility that perhaps the most enigmatic human 
nervous disease known today, the notorious kuru of the New Guinean 
Fore people, which has variously been held to be transmitted by he- 
redity, or congenitally, or infectiously, may be illumined by the findings 
on scrapie. 

At yet another research frontier, where novel theoretical findings 
in biology combine with significant potential gains in a different theater 
of preventive medicine, this year has once again witnessed striking 
events. Although it has been eliminated from a part of the world, the 
human body louse still persists in many areas. There it can serve as 
the vector transmitting the dreaded typhus or plague from man to 
man, as it has done in the scourges of that disease since the dawn of 
human history. The louse, and the agent of typhus that it carries, are 


as characteristically associated today as they have always been with 
hunger, filth, and extreme poverty, and so thrive today in parts of the 
world least prepared to meet their onslaught. They continue to be, as 
they always have been, the particular accompaniment of the sufferings 
and the social disorganization of wars. 

Since World War II an intensive battle has been waged against the 
human body louse with modern chemical insecticides, particularly 
DDT. There have been two consequences. The louse has been eradicated 
from many areas. But, like many other insects, the louse is a biochem- 
ically versatile organism. In areas where it has not been eliminated, 
mutants that successfully resist such agents have largely replaced the 
susceptible forms which made up the greatest proportion of the original 
populations — as has also happened extensively among the disease-trans- 
mitting houseflies and mosquitoes of the world. This development has 
created epidemiological emergencies comparable to those produced by 
parallel situations among some of the most serious bacterial agents of 
disease, where the causative organisms have acquired similar immuni- 
ties to the newer antibiotics. Once this has happened, whether among 
disease organisms or among the insect vectors that carry them, a way is 
reopened to the most dangerous consequences. For our defenses are re- 
duced to essentially what they were before the antibiotics and the in- 
secticides, on which we have relied so implicitly through the past 
decades, were discovered. It is not simple to devise new chemical weap- 
ons of comparable effectiveness. More important, it is very likely that, 
even when such new weapons are designed and produced, they will once 
again gradually become ineffective as further biochemical mutants of 
the targets appear and are rapidly selected. 

A dramatic new approach announced during the current year could 
well foreshadow a major change in the whole method of control of the 
vectors of insect-borne disease. More than a decade ago, Carroll M. 
Williams of Harvard University discovered in the male of a species 
of silkworm moth a rich natural source of a remarkable insect hor- 
mone. It serves as one of the elements in a delicately balanced endocrine 
system modulating and controlling the complex and precisely adjusted 
patterns of development in insects. Its particular function is to main- 
tain insect tissue in a youthful, or larval, condition, inhibiting matura- 
tion. In the normal insect its action is balanced by another hormone 
system, which promotes the maturity of tissues. The opposing systems 
ensure the precise structuring of the complex and highly regulated 
metamorphosis of the normal insect. 

Almost at the time of this discovery, its author predicted that if ever 
the naturally occurring juvenile hormone could be made available in 
quantity, or if a biologically active analogue could be synthesized arti- 


ficially at reasonable cost, a new class of insecticides might become 
available — a class distinguished from its predecessors by at least three 
important characteristics. Unlike substances of the type of DDT, they 
would be harmless to organisms outside the class of insects, such as fish, 
birds, mammals, and man himself. They might even be made specific 
to certain targets, so that beneficial insects need not be indiscriminately 
slaughtered with noxious ones. And since these hormones are evolved 
by the targets themselves, it is extremely unlikely that the insects could 
acquire the immunity to the hormones that they have developed so 
effectively for the chemical classes of poisons. 

Recently, the first part of this prediction came true. It proved possible 
to synthesize an analogue of the insect juvenile hormone which was 
biologically active, and also to find cheap and abundant sources of a 
natural product. During the year just past, the synthetic hormone was 
tested on the human body louse with decisive effects. Eggs continuously 
exposed to it failed to hatch. Exposure of the immature insect to mod- 
erate concentrations inhibited further development, and heavier dosages 
proved fatal to the adults. Thus a potentially powerful weapon against 
the further spread of typhus in the world — a possibly invulnerable 
weapon which may in time be effectively used also against other men- 
acing insect carriers of disease, a weapon which fifteen years ago would 
have seemed only generally predictable — has become concrete. That 
achievement is surely another major milestone both in biological re- 
search and in the control of disease of the year just past. 

A keen interest in the nature and workings of the human brain — and 
in what has perhaps too often been assumed to be an exact equivalent 
of it, the human mind — must be older than civilization, as the numerous 
trepanned skulls found in the ancient graves of many lands bear silent, 
pointed witness. But it is only within this half century that more 
sophisticated biological understanding, the newer and exquisite tech- 
niques of neurophysiology, the discoveries of the actions of brain-influ- 
encing chemicals applied generally or placed at precise sites on brain- 
stem or cortex, and the greatly increased scope and refinement of many 
of the modern diagnostic techniques have given us a feeling of increased 
promise in that field. 

In the last decade, the Age of Computers has affected the nature of 
questions asked and the kinds of research undertaken to elucidate them 
quite as much in the field of neurophysiology and brain research as in 
functional and biochemical genetics, though frequently in very different 
ways. We speak often now of computer "communication," for example 
— among computers and between computer and operator. Programmed 


"languages" for computers, in turn, have become more and more re- 
fined and versatile, conveying new degrees of range and precision. To- 
morrow it may well be that information can, quite literally, be trans- 
mitted with the speed of light, traveling along the coherent beam of a 
laser. And no characteristic of the computer that suggests the action of 
the brain has been developed more dramatically in recent years than its 
"memory." Computer capacities for the storage of information used 
to be measured in thousands of bits. Now they are measured in millions, 
or even in thousands of millions. Perhaps it is this spectacular advance 
in the memory capabilities of computers that has particularly fostered 
the idea that machines thus endowed, immensely inferior to brain mem- 
ory though we know the endowment to be in a quantitative sense, could 
nevertheless by their operations shed some light on the specific nature 
of human recollection. 

Such an analogy, however, must be regarded with reserve. It would 
not be argued that computer coding is a precise analogue of genetic 
coding. So it may prove in the case of computer versus human memory, 
and to an even greater degree. For the brain and the computer, in last 
analysis, are very different things. As Donald MacKay has pointed out, 
a primary challenge to research on the nature and the function of the 
brain is posed by the very fact of its structural and functional com- 
plexity: a structural and functional complexity unexampled elsewhere 
in the living world, except perhaps among the great conurbations of 
our time. And since even in the greatest urban centers we do not yet 
deal with individual aggregates of citizens numbering ten billion or 
more, while the human brain does include approximately that number 
of individual neurons in its compass, the brain may well be the most 
staggeringly complex biological organization in the world. To approach 
such a subject with some hope of ultimate real understanding requires 
even more than newer, more powerful, more searching techniques. 

Over three centuries ago, Descartes was tempted to look at the 
mechanisms of body and mind primarily in terms of the clockworks 
that were then so novel and so exciting. A decade or more ago, when we 
were just beginning generally to appreciate the new and spectacular 
developments in telephonic communication, we were constantly tempted 
to compare the brain and the nervous system with a telephone network 
and a telephone exchange. The comparison clearly stimulated valuable 
investigative work on the nature of nerve and brain. But in the end, of 
course, it proved much too narrow. So may the comparison ultimately 
prove with ideas of "computer thought" and "computer memory," if 
conceived in terms of specific models rather than of purely functional 


parallels. This lesson may be particularly pointed with respect to mem- 
ory. For, as MacEay has pointed out, it is the act of remembering, the 
process taking place, that we seek to understand, and this may be a 
somewhat different goal for the investigator from that of seeking evi- 
dence for the physical existence of a record, in the sense that one would 
seek it in the storage bank of a computer. This subtle distinction may 
in fact be extraordinarily profound. It has been illustrated with unusual 
vividness by developments of the year, involving a mode of investigation 
that has produced some of the most suggestive findings relating to the 
nature of biological memory that have been reported. 

Not long ago it was discovered that when mice, trained to a maze 
until they had learned it thoroughly and could readily recognize and 
again successfully thread it a considerable period after the training, 
were injected intracerebrally with the antibiotic puromycin, they suf- 
fered apparently complete and permanent oblivion of that training 
within ten to twenty hours after the injection. The loss of memory for 
a particular routine persisted for at least three months after the treat- 
ment. The same effect was later fully confirmed with goldfish similarly 
trained to a maze. 

Puromycin, a naturally produced antibiotic, powerfully inhibits pro- 
tein synthesis. From the beginning this was thought to be a decisive 
factor in its action — indeed, this was the property which, on theoretical 
grounds, led to its initial selection for the experiment. But puzzling 
features appeared. Acetoxycycloheximide and cycloheximide, for ex- 
ample, also markedly inhibit protein synthesis. Yet they appeared to be 
without effect on memory when similarly injected. It was also later 
found that cycloheximide, and also chloramphenicol, could actually pro- 
tect against the destruction of memory by puromycin. 

The modes of inhibition of protein synthesis of puromycin on the one 
hand and of acetoxycycloheximide and cycloheximide on the other are 
quite different. Acetoxycycloheximide is known to impede the transfer 
of an amino acid from RNA to polypeptide, thus hindering protein 
building. Puromycin, however, inhibits protein formation by interfer- 
ing with the growth of polypeptide chains in quite a different way. The 
absence of an inhibiting effect on memory by cycloheximide, together 
with its property of protecting against the inhibiting effects of puro- 
mycin, invoked a more complex picture. The new hypothesis suggested 
that initial changes within the brain accompanying the fixation of mem- 
ory might involve corresponding changes in the amounts of one or more 
types of molecules of messenger RNA, altering the rates of synthesis of 
one or more proteins essential for memory expression, while these pro- 
teins, or possibly their products, might in turn act as inducers of their 
related messenger RNA. Thus expression of memory would depend on 


changes in protein concentration brought about by corresponding mes- 
senger RNA, changes in turn modified by the experience of learning. 
Loss of either protein or the corresponding messenger RNA could thus 
upset the cycle, with presumed loss of memory. Since puromycin in con- 
centrations comparable to those injected is known to inhibit the syn- 
thesis of RNA in the cells of vertebrates, this seemed a reasonable 
explanation of its memory-destroying power. By the same token, the 
protection which cycloheximide is known to give messenger RNA could 
account both for its ineffectiveness in destroying memory and for its 
memory-protecting action against puromycin. 

This reasonable and consistent hypothesis was challenged by a new 
finding during the past year. A series of experiments was undertaken 
to determine how long puromycin must remain in contact with the 
brain to produce complete and long-lasting amnesia. Saline solution was 
injected after puromycin treatment, diluting and finally washing out 
the drug. Unexpectedly, when the puromycin was removed, the memory 
record which had apparently been destroyed proved to have been only 
blocked, and was restored. It became evident that mechanisms quite 
different from those envisaged earlier must be implicated. 

Shortly thereafter results of another approach to the same problem 
were described, and a new mechanism was suggested. Recordings of the 
electrical activity of the mouse brain taken from the hippocampal region 
a few hours after the injection of puromycin gave evidence of the occur- 
rence of nervous convulsions which were frequent, but not so severe as 
to affect the overt behavior of the animal. When similar recordings were 
made from the brains of mice that had received injections of cyclo- 
heximide or saline, however, or when cycloheximide injections were 
combined with those of puromycin, such abnormal electrical activity 
was much attenuated or absent. A conclusion that could be drawn from 
these new findings is that the effects of puromycin in producing mem- 
ory loss are related primarily to induced abnormalities in the electrical 
activity of the brain, involving in fact subclinical seizures; and that 
failure in functional recall of the memory store, rather than loss of a 
specific memory trace, was the prime factor involved. 

Time may show that this picture is not as different from nor so 
incompatible with the earlier concepts as might at first appear. Yet, 
set in opposition, the two interpretations provide a vivid example of 
the kind of contrast between viewpoints emphasizing structure and 
viewpoints emphasizing process that may well form a major theme in 
our approaches to the brain, and perhaps to biological entities of com- 
parable complexity over a wider range, in the years to come. 


Superficially, the past year seems to have been one of sharp contrasts 
on the scientific scene. There has been the satisfaction and the sense of 
genuine cultural gain and cultural certainty conveyed by the striking 
substantive advances on a multitude of frontiers of scientific discovery 
and application typified by these few but representative examples. But 
there have also been the prominent difficulties and doubts and uncertain- 
ties attending the social scientific concerns of the same year, as con- 
comitants of their weight and significance. We have lived with these 
latter problems for some years — though they have rarely been so mas- 
sive — and they will surely be with us in even more acute form far into 
the future. They are concerns of great national import. For not only 
do they involve relatively vast sums of public money but, more im- 
portant, they deploy great numbers of men and women who, potentially 
or actually, are of special significance to our society. In contrast to the 
uncommitted ends of purely scientific discovery, the major orientation 
of these concerns usually is, as in its very nature it must be, toward 
goals more nearly preconceived, and often biased primarily to other 
ends than those of scientific discovery per se, or of cultural advance- 
ment in the strict sense of cultural content. 

It is difficult a priori to imagine a sharper contrast in purpose or 
design, or approach, or requirements than that demarking the pursuit 
of scientific discovery from the use of the total scientific enterprise as a 
new and extremely powerful tool of social change. Yet the two enter- 
prises, vividly contrasting as they are in many aspects, not only exist 
together but are in fact indissolubly linked. On the one hand, the shape 
and the magnitude of exploratory science in our time, for better or for 
worse, are deeply affected by the massive deployment of science to social 
ends that is so commanding a feature of our day. On the other, science 
is, by its very nature, one of the extremely powerful forces for social 
change and social molding with which we live. To fail to ponder the 
quality of its impact is to be irresponsible. The most important task of 
science in our day, in fact, may be to meld these two aspects of the 
scientific way without damage to the precious core on which all effec- 
tiveness must ultimately rest — the philosophy and the continuing oppor- 
tunity of independent, uncommitted scientific exploration. Perhaps the 
greatest challenge for our time in this field is to understand truly new 
dimensions without discarding or even failing to maintain in pristine 
brightness the precious values that we inherit and that are basic to the 
significance of science in society. One aid on the long and difficult road 
of understanding such relationships and of protecting and conserving 
these precious values while bringing them to bear on the overriding 
needs of a changing society may be to look on these linked aspects as 
both contemporary and sequential stages in a vast social evolutionary 


process which exhibits some of the same characteristics as do evolu- 
tionary processes elsewhere in nature, and to deal with them in that 

So it may be pertinent to recall that it was in February, just one 
hundred and ten years ago, that across the mind of Alfred Russell 
Wallace, as he lay racked with a bout of severe intermittent malaria at 
Ternate in the Moluccas, "there suddenly flashed . . . the idea of natural 
selection." During the rest of that bout of ague he thought the notion 
through, and made a preliminary draft of it the same evening. During 
the following two evenings he composed the famous letter to Darwin 
which prompted the recipient, when he read it on June 18, to write on 
the same day to Sir Charles Lyell : "I never saw a more striking coinci- 
dence. If Wallace had my ms. sketch written out in 1842, he could not 
have made a better abstract ! Even his terms now stand as the heads of 
my chapters." This was the essay that, with a summary of Darwin's 
views, was read as the joint paper before that memorable meeting of the 
Linnean Society on July 1, 1858, when the theory of organic evolution 
was first officially laid before the world. 

Through the century since, we have gradually come to realize how 
fundamental were the principles enunciated at that meeting of the 
Linnean Society, and how generally they apply in living nature — well 
beyond the horizons that could have been even dimly visioned by a 
Darwin or a Wallace, or by a generation of their followers. Among 
living organisms we are now finding that the processes that Darwin and 
Wallace elucidated are displayed at far more elementary levels of 
biological organization than those at which they were first observed. At 
microscopic and submicroscopic levels within a single cell, indeed, the 
evolutionary aspects of living systems are displayed in exquisite detail, 
in realms of nature where the very scope of knowledge of Darwin's time 
made more than a vestige of appreciation impossible. Even so, the 
astonishing circumstance is clear today that the genius of Darwin did 
lead him to suspect something of this in the realm of genetics. In that 
critical year of 1858 he wrote to Huxley: "Approaching the subject 
from the side which attracts me most, viz. inheritance, I have lately been 
inclined to speculate, very crudely and indistinctly, that propagation by 
true fertilization will turn out to be a sort of mixture, and not true 
fusion, of two distinct individuals, as each parent has its parents and 
ancestors. I can understand on no other view the way in which crossed 
forms go back to so large an extent to ancestral forms." In that 
pre-Mendelian age, dominated by notions of blending heredity to which 
Darwin himself subscribed in practice, it was a moment when a pro- 


phetic vision captured one fleeting glimpse of the particulate quality of 
inheritance and of the nature of the germ plasm. But not even that 
genius, nor yet the insight and the investigations of more than two 
subsequent generations of geneticists, were able to bring fully to light 
what is so clear today. 

It would surely have delighted Darwin immeasurably that evolu- 
tionary processes can be seen at work just as surely and just as typically 
in the subtle realms of genetic form and structure and genetic mecha- 
nism as at the grosser levels of morphology in the organisms whose 
heredity they mediate. The early evolution of sexuality in nonsexually 
reproducing primitive organisms and the various modes of its determi- 
nation, the later reverse evolutionary courses in which bisexuality has 
been discarded, culminating in populations of advanced partheno- 
genetically reproducing organisms that have appeared again and again 
in the history of life in adaptation to special conditions of the environ- 
ment ; the structural evolution of the chromosomes themselves from less 
precisely disposed genetic material, and the evolution of the complex 
and finely regulated processes by which they are proportioned at mitosis 
in the division of body cells and between new individuals in the mecha- 
nisms of meiosis; the biological "management" by which diversity is 
secured in parallel with continuity; the selection of mutations and the 
perpetuation of advantageous ones; the manner in which such "acci- 
dents" to the chromosome as deletion and translocation are repaired or 
put to adaptive ends; the evolution of the diploid from the haploid 
chromosome condition and the evolution and adaptive "use" of poly- 
ploidy; the development of the controls for the expression of new char- 
acters in the plant or animal body with its accompanying evolution of 
dominance or recessiveness, and the evolution of sex-limited and sex- 
linked inheritance — all these circumstances strikingly repeat and illus- 
trate at new levels the grand features of the process. They are being 
even more explicitly demonstrated today at the still finer level of 
molecular structure. Pioneering studies of the apparently excessive 
replication of particular genetic loci in the evolving chromosomes of 
some higher organisms, described elsewhere in this Report, emphasize 
one aspect of this development. Another is represented by striking 
findings concerning the mechanisms through which the effects of partic- 
ular genes are reinforced or inhibited or controlled in their appearance 
during development by other genes or other chemical entities of molec- 
ular dimensions — an approach first conceived some two decades ago by 
Dr. Barbara McClintock at the Carnegie Institution, dramatically con- 
firmed in bacteria by Jacob and Monod at the Pasteur Institute, and 
now so widely recognized in living things that the study of regulator 


genes and their modes of action is a heavily populated field of investiga- 
tion at the frontiers of biology today. 

In the classical case discussed by Jacob and Monod, the product of a 
regulator gene, called a repressor, acts by switching off a highly specific 
target gene. The control is therefore negative. It might be brought about 
directly by the repressor product. Or it might be achieved by an inter- 
action of the repressor substance with messenger RNA to block the 
translation of the genetic message into specific protein. 

During the year investigators have isolated such inhibiting sub- 
stances produced by specific repressor genes. It has been possible to 
study the action of one of these, the protein made by the Ci gene of 
lambda phage, which acts to keep the phage chromosome dormant within 
its colon bacterium host until the appropriate time for growth. From 
experiments conducted at molecular level it has been possible to demon- 
strate that the control is indeed achieved by the direct "occlusion" of a 
specific site in the DNA by the repressor substance. It is becoming more 
and more clear that actions of this highly specific kind at molecular level 
must be responsible for the deployment in time of all the precisely 
patterned and adjusted sequential developments in the growing em- 
bryos of many-celled plants and animals. The evolution of the molecular 
means by which such adjustments are secured displays an intricacy and 
a regulated precision to dwarf all that we have known before. Yet it 
bears, unmistakably, the marks of the same basic evolutionary processes 
that have so impressed us, on the stage of the organism as a whole, since 
the days of Darwin. 

Even at molecular levels where only the components of life, rather 
than the whole living organism are directly involved, evidence of the 
exquisite patterns of self -regulation imposed by the requirements of 
adaptive evolution are becoming evident. Already the three-dimensional 
structures of some biologically critical enzymes have been or are being 
completely worked out — the structures of lysozyme, ribonuclease, 
chymotrypsin, carboxypeptidase (a protein that digests other pro- 
teins) ; of the respiratory enzyme cytochrome c, and the plant digestive 
enzyme papain. Max Perutz has recently pointed out that as the modes 
of action of such biocatalysts become clear they may prove to be the first 
substances of any sort in which the mechanisms of catalysis are fully 
understood, although industry has depended heavily upon catalysts in a 
myriad of processes for many years. 

Even among these biologically vital, and yet nonliving, molecular 
entities, self -regulation is extraordinarily precise, suggestively parallel- 
ing the self -regulatory mechanisms that evolution has imposed on living 
things. Consider, for instance, the cycle through which cytidine tri- 


phosphate, one of the building blocks for the manufacture of nucleic 
acids, is formed. The starting material is aspartic acid or aspartate, 
and the catalyst promoting the first step in the five-stage process is the 
enzyme aspartate transcarbamylase. The first molecules of aspartic 
acid or aspartate that bind to the enzyme change its structure some- 
what, making other binding sites on the enzyme more receptive for 
aspartate molecules than before and thus accelerating the conversion of 
the raw material to the first intermediate product, carbamyl aspartate. 
As the reaction proceeds toward its terminal stages and cytidine tri- 
phosphate begins to accumulate, however, the final product too be- 
comes attached to the enzyme. This sharply diminishes the enzyme 
activity, perhaps by competing with the aspartate for binding sites. 
Thus, so long as molecules of cytidine triphosphate are accumulating in 
excess — that is, so long as the factory for this particular structural 
block of a nucleic acid is producing more than is immediately required — 
the aspartate transcarbamylase is a less effective catalyst and the 
supply of carbamyl aspartate is reduced. Thus the whole manufactur- 
ing process is occluded at its beginning. Conversely, as the surplus of 
cytidine triphosphate is used up in nucleic acid manufacture and its 
concentration in the medium falls, the supply of free carbamyl aspar- 
tate is replenished from the population of bound molecules. The reaction 
picks up again, until, perhaps, it gets out of step with the general 
process once more. In the truest biological sense, this is a mechanism for 
the maintenance of homeostasis by negative feedback from the reaction 
product, closely simulating living processes and indeed fulfilling the 
requirements of such processes. 

This is but a single example — and by no means one of the most com- 
plex — of a number of self -regulating enzyme systems that have been 
analyzed, governing with fine precision an array of biological processes 
in living systems. It would be hard to find more dramatic illustrations 
of evolution at the most elementary level in the structuring of life where 
it has been investigated, a level indeed but newly opened to us. To be 
sure, most of the actual steps in such evolutionary processes remain to 
be specifically demonstrated, though a few have already been detected 
experimentally, such as mutations in the details of allosteric reactions 
of enzymes governing metabolic syntheses in certain bacteria. But 
whatever the specific paths through which enzymes evolved to govern 
the process of life, it seems highly probable that the earliest lacked reg- 
ulatory control mechanisms. Once such an enzyme had become firmly 
incorporated in an organism, however, and had become indispensable to 
its metabolic economy, then the "design" of special active sites on its 
molecule, and the selection of the most efficient of these to regulate the 
catalyst, must have followed the typical courses of evolution, closely 


coupled with the evolution of its "host." In recent years, the discovery 
and elucidation of this new setting for biological regulation within the 
biological molecule itself has opened vistas that, in their demonstration 
of the ubiquity of the processes of evolution through every level of life, 
would surely have brought renewed wonder and delight to a Darwin 
and a Wallace were they living at this hour. 

From this panorama of evolving biological systems, certain common 
features of the evolutionary way, at whatever levels and in whatever 
contexts they are played out, strike the observer with special force. One 
of the most fundamental features, conspicuous at every level, is the 
universal circumstance that such evolutionary courses have evidently 
unfolded in relatively small successive steps, each new step profoundly 
influenced by preceding ones and by all the surrounding requirements 
for survival, and each step exquisitely meshed with the total situation. 
This is evidently a primary requisite for success and indeed for per- 
sistence: the great saltations that some earlier students of animal 
evolution expected to find have yet to be satisfactorily demonstrated. 
Each new limited change is firmly embedded in the matrix of a prior 
context, so that the total pattern of evolution in effect develops as a 
mosaic, of which the greater proportion is conservative and older in 
design, the smaller, radical and new. But characteristically the older 
elements too were sifted earlier in their turn, and selected for adaptive 
value with equal rigor. Those that have been tenaciously retained have 
been well proven. Great evolutionary changes, for example, have obvi- 
ously occurred among the vertebrates since the rise of early fishes. Yet 
in neither amphibians nor reptiles nor birds nor mammals nor man have 
the great structural principles of the vertebrate eye or the spinal col- 
umn or the central nervous system been discarded. Rarely if ever in 
successful organic evolutionary change has the baby been thrown out 
with — or after — the bath. 

Another striking and fairly universal quality of evolutionary change 
is that, in order to be successful, the evolving systems must sustain a 
tight-woven network of communication permitting the most intimate 
and sensitive continuing interactions among their components, whether 
at the chemical level, or at that of intracellular organelles, or of cells in 
a many-celled system, or of whole organisms in a biological society. 
These are qualities of biological evolution to be remembered. 

Ever since the days of Herbert Spencer, observers have been struck 
by the fact that human societies frequently seem to resemble biological 
systems in some aspects of both evolution and structure. The mecha- 
nisms involved, of course, have been quite different, as has been pointed 


out repeatedly over the years. The capacity to react sensitively to in- 
formation, to conserve, communicate, and use it, and constantly to 
accumulate new stores is central to successful evolution in either con- 
text. But the biological mechanisms of accumulation, selection, use, and 
transfer of information differ in obvious and extremely important 
practical ways from the mechanisms involved in the accumulation, use, 
and transfer of social information by the processes of learning and 
through the medium of culture. The potential for what is in effect the 
inheritance of acquired characteristics confers upon the evolution of the 
culturally mediated society a dynamism, a flexibility, and a versatility 
that the biological organism or system can probably never remotely 
match. The obvious penalty attached to this versatility and flexibility, 
of course, is a degree of vulnerability also seldom matched in the biolog- 
ical world, as the disappearance of past great civilizations bears silent, 
poignant witness. 

But these differences of mechanism, conspicuous and indeed basic as 
they are, should not obscure the parallel requirements for success that 
face the evolving society and the evolving organism. There are a number 
of such similar requirements. They are characteristically elementary, 
and by that very token may be especially noteworthy. Thus, precisely as 
in the biological world, there is a sharp limit to the rate at which evolu- 
tionary change can take place in social evolution, and to the magnitude 
of any given step, if the innovations are to be successful and if the 
society that they will affect is not to be severely disoriented or even 
crippled by the process. As in biological evolution, effective social evolu- 
tion must be at once radical and conservative, freely embracing the new 
yet scrupulously preserving basic and well-tested elements that have 
had a high survival value in the past and which remain relevant to the 
present. Like successful biological evolution, successful social evolution 
must constantly guard against discarding the essential with the trivial 
— against throwing the baby out with the bath. 

The more rapid the social evolution, the more imperative becomes this 
requirement and the more vigilance is demanded to satisfy it. Through 
history this has been one of the most difficult lessons for man to learn. 
Repeated failures to understand it or to act upon it adequately must 
have accounted for major historical debacles — and indeed for major 
distortions of social evolution. It is a danger if anything more acute 
today than ever before, because of the immensely increased dynamism 
of social movements and evolutionary forces with which we live, and 
because the stakes of failure, like those of success, are so much higher 
than they have ever been. It would be difficult to find a more apposite 
general caveat for our time than this of exercising due care that, in 
embracing new and experimental courses on myriad fronts of move- 


ment with the ardor that we must, we do not at the same time discard 
long-tested values and long-tried adaptive courses which, if they are 
lost, will only have, one day, to be re-won — and probably at enormous 
cost. This is a consideration that is before us in all our affairs, every 
minute of every day. 

Finally, it is very clear that, as in biological evolution but to a yet 
higher degree, successful adaptive evolution within any society can only 
be accomplished in the presence of as complete a nexus of communica- 
tion among all parts of the social structure as can possibly be secured. 
Only if superior communication prevails can the highly pluralistic social 
structure of today be constantly adapted to the selective changes con- 
fronting it in the many-sectored environment of society. In the realm of 
science, this requirement becomes more difficult, and demands more and 
more effort and ingenuity to satisfy, as the volume of scientific litera- 
ture burgeons and the ranks of those committed to the scientific way 
increase ever further. 

With the delineation of these general parameters, we return to issues 
raised at the beginning — the vital, challenging, intimately interlocked, 
and often also frustrating problems of science policy in American 
society, with all their potential opportunities, which events of the 
current year illustrate so well, coexisting with the rich harvest of more 
evident and satisfying individual conquests at substantive scientific 
frontiers that the same year has provided in such impressive measure. 
How are the two related? 

Throughout the history of science, as John Steinbeck has remarked, 
parts of the universe of experience have been rediscovered, redescribed, 
reclassified. Again and again an investigator of genius, gazing at the 
same world as his predecessor, has lifted a new corner of the veil of 
incomprehension and has glimpsed a new scene. In the excitement of the 
fresh vision there has typically been, at first, skepticism and controversy 
among those to whom the vision has not yet been granted. This, perhaps, 
is the most bracing and dynamic and stimulating period. Then, if 
acceptance follows, a band of believers slowly accumulates, extending 
the new ideas and confirming them in added ways. Gradually this group 
of disciples becomes more formalized, and the typical steps of post- 
discovery follow — evaluation, taxonomy, rearrangement in the new 
world of experience — until, a half generation or so later, a new struc- 
ture has supplanted the old. Then it is that the ancient danger, against 
which the whole vista of organic evolution warns, becomes particularly 
acute. As Thomas Huxley emphasized many years ago, it is the common 
fate of knowledge to start as heresy and end as superstition. Character- 


istically it is the band of later believers — not the first discoverers — who, 
as the new system becomes codified and orthodox, forget the invariants 
in the evolution, the older common factors long proved viable that are as 
essential in the new structure as they were in the old, and that should 
provide a bond between the two. In their enthusiasm for the newly 
orthodox, they may press to discard all the features of the older frame: 
to start wholly afresh. Ultimately, as patternmaking proceeds in the 
new conceptual world, these older invariants, so lightly discarded in the 
period of change, typically emerge once more, to reoccupy in the new 
system the central position that they held within the old. But time and 
expensive effort will have been lost. How much more economical it might 
have been had their invariance been earlier recognized! 

One such invariant, in the realm of the creation of new knowledge, 
stands out particularly when the substantive scientific gains of the year 
are viewed against the wider background of the whole technological 
activity of the nation and its relation to public policy. Every one of the 
substantive advances of the year that has been cited — and many more 
could be added — has typically been the work of a single investigator, or 
of a small group. None has been achieved by the specifically united 
efforts of a massive team — though it has many times been proven that a 
close-knit research group can effectively breach the frontiers of new 
knowledge when embedded within a large organization, provided that 
its own integrity be recognized and rigorously protected. 

The exploitation of the implications and the consequences of new 
knowledge once won and its application to the purposes of society, may 
be, and commonly is, truly the organized communal work of thousands. 
The creation of that knowledge is typically the province of a compara- 
tive few, dedicated to the life that Newton epitomized when asked how 
he made his discoveries : ". . . by always thinking unto them. I keep the 
subject constantly before me and wait till the first dawnings open little 
by little into the full light." This has been the history of advance in 
science since the beginning of science itself. Once, when scientists were a 
small group in the society, the work was characteristically that of a 
single individual — the Newton, or the Pasteur, or the Michael Faraday. 
Today, with the number of scientists so vastly greater, similar pioneer- 
ing is more apt to be accomplished by bands of close associates, often 
working in parallel with others, at the center of that wider network of 
communications that the successful evolutionary pattern demands here 
as elsewhere, but themselves small, defined units where flexibility and 
the opportunities for intimate, moment-by-moment exchange of in- 
formation are at a maximum. This structuring of the research group, 
clearly, is an essential core in the evolution of the creation of new 
knowledge, and of new viewpoints about the world, which we neglect or 
negate at our peril. 


Yet the temptation to forget the importance of this invariant is 
surely greater today than it has ever been. There are powerful reasons 
for this. The acts of creating new knowledge in science on the one hand, 
and of consolidating that knowledge and using it for the purposes of the 
society on the other, are typically and properly conducted in close con- 
junction and form a continuous spectrum. Thus, to the casual observer, 
they appear to merge, and to be essentially alike in character and re- 
quirements. In fact, at the extremes of that spectrum, genuinely but not 
often obviously different processes are invoked which require somewhat 
dissimilar approaches. They are commonly shaped by significantly 
divergent philosophies and are likely to be carried out by men and 
women of rather dissimilar orientation, although the capacities and 
training of those men and women may be very similar and despite the 
fact that some are able to work in both environments and to discharge 
both roles successfully. 

Now the instruments for the creation of knowledge, which have re- 
mained constant in their basic features over several generations, must 
operate effectively today in an environment where the means for the 
marshaling and exploitation of that knowledge, once gained, are evolv- 
ing radically and rapidly and where experimentation with new forms is 
proceeding with unprecedented speed. In this context, science, for indus- 
trial nations the world over, has come into a relationship with society 
almost precisely opposite from what it was when the mechanisms for 
the creation of new knowledge were themselves first evolving — when the 
pioneering work of a Galileo or a Leibnitz created little stir at the time 
of its accomplishment beyond very limited professional and intellectual 
circles. Today the products of this same science have become major 
factors in the policies of nations. As Sir Eric Ashby has vividly ex- 
pressed the problem, in the broader setting of the whole environment of 
universities: "Formerly they (the universities) were detached orga- 
nisms, assimilating and growing in accordance with their own internal 
laws. Now they have become absolutely essential to the economy and the 
very survival of nations. Under the patronage of princes or bishops they 
were cultivated as garden flowers, of no more significance to the state's 
economy than the court musician. Under the patronage of modern gov- 
ernments they are cultivated as intensive crops, heavily manured, and 
expected to give a high yield essential for the nourishment of the state." 
In such a climate, the protection of the essentials of the research envi- 
ronment is both more difficult and doubly demanded. 

Ours is a society, originally heavily dependent upon technological 
innovation for its very existence, which has now become equally depend- 
ent upon the continual creation of new knowledge for its survival, and 


which we may truly style innovative in its basic character; a society 
in which the costs of some kinds of research have already mounted to 
the point where they can be borne only by the Federal Government, and 
threaten soon to exceed even that resource; a society in which the twin 
streams of the creation and the use of knowledge are now thoroughly 
merged. It is an egalitarian society, moreover, where, as de Tocqueville 
long ago predicted, free and voluntary associations of citizens flourish 
and multiply in every sphere of activity — and ought, as de Tocqueville 
said, to fulfill the functions typically assumed by autocratic individuals 
or an autocratic bureaucracy in more rigidly structured situations. How, 
in such a society, are we to mesh the invariants that are so vital to the 
continued creation of new knowledge with the overarching considera- 
tions of common purpose that are now so deeply affected by the trends 
and developments of science itself? How are we to do this while main- 
taining the integrity and independence of the forms and of the very 
institutions which are clearly essential to the creation of new scientific 
knowledge? How are we to protect, and fully tap, those clear springs 
from which high excellence flows — to maintain the sources and to hold 
high the selection and training of that precious human reservoir of 
extraordinary quality, in science as elsewhere, which has always pro- 
vided the ultimate guarantee of our survival — within the massive set- 
ting of a technological, pragmatic society of immense power and 
resource, whose vast capacities shape these very sources and reservoirs? 
And conversely, how is science itself, pragmatically considered, best to 
be supplied the most effective channels through which to serve the 
national purpose? 

To this last question, of course, there is a prior one, itself not yet 
clearly answered. How are we to define that national purpose and for- 
mulate it coherently with respect to science? How far we are from any 
agreed answer to this most basic and difficult of questions the events of 
the past year in the arena of publicly supported science give vivid 
evidence. As far as any general consensus on the whole social position 
and significance of science goes, we are surely living through times as 
inchoate and contradictory as any that our society — or perhaps any 
society — has breasted, in a world itself grown far more complex and 

But if our time poses totally novel problems in this sphere, it is also 
surely a time of the richest opportunity. Given suitable leadership and 
understanding, given a steady and consistent will to surmount the pre- 
occupations and the special, short-range interests that distract us, there 
is no reason why we should not reap benefits from the interwoven con- 
cerns and objectives of science and society in the years to come to dwarf 
what w r e have known. But the effort will require even more than capable 


leadership, and even more than continuing vision to rise above imme- 
diate problems. It will require the will and the capacity constantly to 
innovate in the field of social mechanisms, constantly to devise and to 
try out new patterns, and to discard quickly and decisively those which 
prove unsuitable to our purposes. 

In this effort, of course, those concerned with the future of science 
and public policy — and indeed with science as public policy — are only 
called upon to do precisely what America, from the very beginning of 
its history, has always been peculiarly adept at doing in a more general 
frame : to give full rein to our national talent to be constantly inventive, 
and to foster highly pragmatic and selective use of our inventions. In- 
deed, the whole challenge to the scientific way in this context offers a 
paradigm for the similar challenge confronting the broader intellectual 
community of the nation, a challenge epitomized by the whole complex 
of the problems facing the universities and indeed of the actual future 
role of the universities in American society that so vex us today, but 
also extending well beyond this milieu. It is this aspect of paradigm for 
a still larger but closely related national problem that lends to these 
questions of science and the national policy, of the antitheses between 
excellence and equality, of the adequate protection of environments and 
modes which we know to be absolutely indispensable to creative thought 
in any field, a range and significance even greater than that assured by 
their intrinsic importance. 

In an economic message to Congress in 1963, President Kennedy esti- 
mated that two thirds of the trained people in the nation available for 
scientific and technical research were involved, in one way or another, 
with the defense, space, or atomic energy programs of the nation. Is this 
the optimum distribution of our talent? Or should we devote more 
specific attention than we have done to diversification and the means of 
bringing it about? Are we making the wisest choices in seeking to 
partition our human scientific resources between more highly focused 
governmental and more plural civilian enterprises, in the best possible 
interest of the ultimate strength of the country? It is a striking circum- 
stance that, though Japan spends only about nine tenths of one per 
cent of its gross national product on research and development, its 
annual rate of economic growth has approximated seven per cent, while 
we, who spend more than three times as high a proportion of our own 
far greater G.N.P. — nearly three per cent — on research and develop- 
ment, are unable to nearly match this growth figure. Does the contrast 
to any degree reflect the fact that in our case only about one tenth of this 
three per cent, or a total of about three tenths of one per cent, is devoted 


to our civilian sector, while a large share of the much smaller total 
outlay in Japan is so committed? And if there is a real connection, do we 
in fact have any other option, given our situation in the world? Is the 
case of Japan instructive, or only a special one? 

Many more questions might be put within this particular frame of 
reference. But the framework itself illustrates with remarkable vivid- 
ness a central circumstance confronting alike the specialized areas of 
scientific research and the broader intellectual structure of our nation 
as a whole, epitomized by the problems facing American universities in 
our time. As early as the days of the Manhattan Project it had become 
abundantly evident that the expenditures required to carry forward 
some kinds of research and development closest to the national interest 
were so enormous that they could be met only by the Federal Govern- 
ment, and eventually might well exceed even its capacity. The penetrat- 
ing estimates of Dael Wolfle earlier cited clearly demonstrate not only 
that the aggregate costs of conducting research today in many areas are 
well beyond the means of private institutions, but that limitations are 
evident even at the Federal level. It may well be that we shall conclude 
in the future that some scientific undertakings can be carried forward 
only with international participation. In any event, it is already very 
clear that such enterprises will permanently require the decisive sup- 
port of the Federal Government, and that on a huge scale. Yet, as events 
of the year have demonstrated so vividly, discoveries of major impor- 
tance on some of the most significant and exciting frontiers of science 
may still be achieved at very modest cost, in terms both of money and of 
numbers — but not quality — of investigators. This striking paradox is 
to be remembered. 

For many years now, science has faced the extremely challenging 
question of how to advance and guide major projects supported and 
operated largely by government — of how to inject scientific judgments 
effectively into what must be, in essence, political domains. It is not an 
easy task, nor a unidimensional one. It is amply evident that one ele- 
ment urgently required is advice and assistance from a multitude of 
hands and minds of a wide range of capacities and preparation and 
experience, operating at least as much in the private sector as in govern- 
ment but bound by a sense of common concern into a diversified and yet 
distinctly identifiable whole. In time of full-scale conflict, as in World 
War II, when an overriding common peril was universally felt and 
clearly identified, when the need for swift and effective action was para- 
mount, when the vastly greater part of Federally supported research 
and development was sharply oriented to a single unequivocally vital 
mission, the direction of such an effort, the mapping of its ends, the 
deployment of those engaged in it, constituted a relatively unified mis- 
sion, primarily determined by the Executive branch of government. 


Even under these circumstances, however, the injection of the scientific 
viewpoint at Federal level, the coupling of a scientific effort overrid- 
ingly in the national interest and conducted on a national scale with 
adequate protection for the modes of operation indispensable to the 
productivity of science and scientists, required nothing less than genius 
in those who accomplished it. Yet demanding as that task was, the prob- 
lem may inherently be yet more difficult in times like the present, when 
the effort is not only larger but far more pluralistic, when its goals are 
more diffuse, and when it is far more a daily concern of the whole 
American people. 

Under such conditions, it is in the Congress of the United States as 
much as in the Executive branch that the ultimate responsibility for 
determining the magnitude and direction of a large part of the Federal 
scientific effort properly rests. This circumstance imposes on the Con- 
gress a truly awesome responsibility. And it is not one for which Con- 
gress has traditionally been particularly well prepared, because hitherto 
so few of its members have dealt extensively with scientific issues before 
assuming national responsibilities. The year's record of debate on 
locating the accelerator, mentioned earlier, illustrates the difficulties 
well. Yet progress is surely being made. On the side of government, 
many members of Congress have taken great pains to understand the 
subtle requirements of the scientific way, to make the acquaintance of 
scientists and to listen to their views. On the side of science itself, 
various groups especially interested in policy have sought to aid the 
Congress, as similar groups earlier took the signal step of serving as 
advisors for the Executive. 

Science still has a long way to go in these matters. But imperfect as 
its leadership now is in this vital sector of our national life, it is indeed 
real leadership, and may well prove significant in an even wider theater. 
For the experience of science in this field has been longer than that of 
most other segments of the intellectual life of the nation. And there are 
close parallels between the problems that science has already encoun- 
tered and those that currently face the universities as such in basically 
similar contexts. There is the same acute financial need that has 
brought the Federal Government, vis-a-vis the universities as vis-a-vis 
science, into the position of a major arbiter of social change. There is 
the same urgency for the institutions that articulate the intellectual 
nexus of the nation as a whole, to present a coordinate front in force- 
fully representing both the requirements and obligations of their way 
of life, and in explaining their proper functions to the patron that they 
share with science — the body politic of the nation — who, as in the 
special case of science, may sorely need information of this kind. As 
with science itself, this is nowhere more vital than in shielding creative 
university environments from demands for the execution of too spe- 


cifically oriented missions originated from without, and for the per- 
formance of particular services which could be better or more appro- 
priately rendered elsewhere. There is the same critical national need for 
continuing advice from the best minds in the fields of the humanities 
and social sciences, whose primary commitments lie outside the sphere 
of government, that there is in the case of natural scientists similarly 
placed. How great that need may be has been vividly demonstrated 
recently in the debacle of Camelot, which surely could hardly have 
urred if the most cogent advice in relevant fields had been appro- 
priately marshaled. In all these areas the needs, the vulnerabilities, and 
the unique potentials of science parallel remarkably those of the rest of 
the intellectual nexus of the nation. 

In a wider assessment, too, the problems and duties of the community 
of science and of the community of universities in the arena of national 
policy show remarkable parallels. Each community shares at least three 
common relationships with government. Research must be supported 
without damaging the research climate; services must be provided to 
the government and limits on such services must be set; and, perhaps 
most difficult of all, advice must be given by qualified professionals who 
can think deeply and prophetically about the future to colleagues on the 
fronts of action — colleagues who are far too completely engaged with 
the heat and burden of the day to reflect in this fashion, but who can and 
must make real the consequences of such thought. How indeed can those 
who are able to think most effectively also effectively advise those who 
must act with high public responsibility? It is a question daily posed to 
science and to the rest of the intellectual world alike. 

Science and the humanistic disciplines are at one as primary pro- 
ducers of excellence in the society. They are at one, too, as bastions of a 
philosophy of equal opportunity. To them alike is posed the challenge of 
how best to serve both values without destroying or debilitating either — 
a vital, consuming, and extraordinarily difficult assignment that will 
stretch as far into the future as we can see. Again, science shares with 
the rest of the intellectual world a particularly critical — and most diffi- 
cult — task and duty: the task and the duty of preparing the society for 
what is to come; of reinforcing its values against the shock of sweeping 
change when such change suddenly becomes apparent. Science failed to 
do this with respect to the power of nuclear fission, even though the 
release of nuclear energy was regarded in some quarters as virtually a 
the^ 1 certainty as early as the fourth decade of the century. It 

failed for a variety of reasons. The scientists concerned were deeply 
absorbed in their work. Early there was some question whether the 


process had any practicability. Later there were overriding military 
considerations that irrevocably prevented public communication. There 
was the ethic then dominant which held that the responsibility of 
scientists was entirely to their work, that the social results of that work 
were not properly their concern. Surely these were reasons enough. Yet 
the ultimate cost of those circumstances was certainly high in terms 
both of the intellectual and the institutional unpreparedness of the post- 
war society on which the knowledge ultimately broke. It is abundantly 
clear that it is not fully paid even today. 

It is possible that we are now confronting a situation of equal power 
and potential in the biological sciences. The problems inherent in 
controlling genetic change in organisms are extraordinarily complex 
and difficult— and indeed such controlled change is scarcely feasible in 
even the simplest case today. The chromosomes of a single human germ 
cell may include in their coded messages on the order of a thousand 
million nucleotide base pairs. The task of deleting or mutating a specific 
one or even a small group, preselected and at will, seems technically 
insuperable. Yet it is currently well known that there are bacterio- 
phages capable of deleting from the chromosome of a bacterial host sec- 
tions that include genetic information mediating particular processes of 
chemical metabolism in that host, which sections, carried by the virus 
to a new mutant host of the same species lacking in these particular 
capacities, can be incorporated at the proper point in its chromosome 
and become functional there, repairing the deficient metabolic mechan- 
ism in that cell and in all of its descendants. Bacteria are far simpler 
genetically than higher organisms, and such processes of transduction 
have never been observed in the cells of mammals, though they have been 
sought assiduously there. On the other hand, the evidence is now quite 
clear that a large proportion if not all of the forms of life on our planet 
use and thus "understand" the same genetic code. And it has been esti- 
mated by highly qualified authorities that within a quarter of a century 
it may be possible to "program" cells, at least those of bacteria, with 
synthetic messages. If and when that is done, the possibility of doing 
the same thing with cells of higher organisms might not be too far 

Should such a capability ever be realized, the resulting shock to our 
society, if the society were not properly prepared, could be at least as 
great as that inflicted by the advent of the atom bomb. And the social 
problems that such a technical capacity might pose could be even greater 
and more subtle. For, given the awesome power to bring about controlled 
genetic change, the paramount question then would be : change for, or 
toward, what? It is a question that involves another which hitherto we 
have never really had to face — or to evade — because it has never been 


truly meaningful for us and has seemed, in practice, to lie so far beyond 
our ken. It is the question "What, in a specific sense, perhaps even in a 
biological sense, does constitute a 'good' society for us: what do we 
really want it to be?" If this sort of power were ours, that question 
would be posed in the starkest way. Too much emphasis cannot be laid 
upon the importance of preparing a society to meet such issues long in 
advance of their possible appearance on the horizon. To do this force- 
fully and yet soundly, with conviction and yet without sensationalism, 
is a task demanding the greatest skill and care both in judgment and in 
communication. It is a responsibility shared by scientists and sociolo- 
gists and humanists alike, and to be successfully discharged only by 
their joint efforts. 

There is a final and particularly significant way in which the natural 
sciences are at one with the humanities. In Germany it was long held 
that exposure to the processes of scientific research ipso facto consti- 
tuted a liberal education for the student who himself intended to enter 
quite different fields. As with Latin and Greek in the British tradition, 
it was held that the rigorous inner logic of the system, its sheer intel- 
lectual discipline, its power to identify and to screen minds of great 
general potential, its introduction to new realms of concept and new and 
difficult techniques of dealing with them, made it a powerful agent for 
excellence in education, regardless of the particular ends to which that 
education would ultimately be directed. 

Similarly, science must share with society as a whole another task of 
general humanism, emphasized primarily by science itself. As Edward 
Shils has penetratingly noted, the tremendous advances in the life 
sciences that characterize our day, with the promise that they hold of 
further and even more striking achievements following on a yet keener 
and deeper understanding of the human physical condition, also carry a 
special risk which we must understand and for which we must be pre- 
pared. The prospect of such developments as the substitution of arti- 
ficial for worn-out natural vital organs or the replacement of such 
organs with other living ones maintained in organ banks, of directed 
change in the genetic code or in the expression of genetic factors 
through such modifications as could be made possible by our ever- 
increasing understanding of protein molecules and particularly of 
enzymes, of effective mental therapy or reorientation in desperate cases 
that our increasing knowledge of the mind drugs may one day permit us 
to achieve, carry an implication beyond that of medical benefit. Inevita- 
bly, they also carry the shadowy but sinister possibility of a new kind 
of control of one human being over another. And the hint of such a 
prospect, whether justified or not, whether real or only imagined, comes 
at a time when the experience of widespread violence to human life, and 


the memories of that experience on a yet more terrible scale, are all too 
vivid. A cumulative effect of such a conjunction of developments could 
be that we might in future come to hold individual human life less 
sacred than in the past. Such a trend would surely spell one of the worst 
social, as well as moral, catastrophes that could befall us. The threat of 
such a prospect, however remote, lays the gravest responsibility upon 
science and the other disciplines of the intellect alike constantly to labor 
at placing new discoveries and newly acquired powers in their proper 

Science and the other intellectual disciplines, the research establish- 
ments of science and the universities where scientists are made, the 
community of science and the community formed by the educational 
fabric of the nation, the scientific nexus and the intellectual nexus as a 
whole, clearly share common duties, common opportunities, and common 
problems to an extraordinary degree. In some areas the body of science, 
by virtue of its own nature and of circumstances beyond its own fore- 
telling and control, was touched by the opportunities and immersed in 
these problems somewhat earlier than, for example, the corpus of edu- 
cation at large. Particularly in its specific relations with government, 
scientific research has had a longer history of intensive association than 
education, whether in science or in broader realms, and this disparity of 
timing has raised some peculiar and often seemingly intractable prob- 
lems in the relationships of both science and education to public policy. 
But there may be another and salutary aspect of this disparity. The 
longer experience of scientific research in this arena may serve as a 
paradigm for the parallel but later experience of the educational world 
and indeed of the intellectual world as a whole — a paradigm both of 
successes and of failures, at once an example and a warning, and above 
all, a token for the future. As John Gardner has said: "The nation's 
leaders must serve as symbols of the moral unity of society. . . . They 
can conceive and articulate goals that lift people out of their petty pre- 
occupations, carry them above the conflicts that tear a society apart, and 
unite them in the pursuit of objectives worthy of their best efforts." If 
the experience of science, both in its successes and in its failures, can 
significantly implement these goals, it may truly be accounted of solid 
and significant worth for our society and our nation. 

We in this tiventieth century have learnt, by bitter experience, that a 
climate of liberal, rational opinion is not something which can be 
taken for granted, as part of western Europe's intellectual heritage. 
It is an ideal to be constantly fought for, ivith unremitting vigilance: 
a precious acquisition, all too easily lost. 

"Freeing Science From Myth" — Times Literary Supplement 

The Year in Review 

"Our sun is at least a second generation star. It inherited from pre- 
vious stars a rich treasure-house of elements ivhich makes our world 
the complex and wonderful environment ive find it to be. 
"All of us arc part of that world. We, too, are star dust." 

William A. Fowler, California Institute of Technology 

The research in the Carnegie Institution for the year just past in- 
cluded the uncovering of further information on the still mysterious 
astronomical objects known as quasars, the discovery of a probable 
physical connection between our Galaxy and another galaxy, the gaining 
of new information on the depth and composition of the earth's crust 
in eastern North America, the exploration of new vistas in the bio- 
chemistry of evolution, a demonstration as to how common marine 
sediments form petroleum, the discovery of a way to make some com- 
mon plants double their growth by changing their environment, and 
the addition of new information on the action of the very important 
genes that control other genes. All will be reported upon in this Review. 

Such striking findings, which, in one or another form, typify each 
year of work in the Institution, by their very nature are deeply exciting. 
Yet the less spectacular, but constant and steady progress on difficult 
scientific frontiers, the growth of every Staff Member that each year 
also typically brings, is no less important. If we were to sum the always 
freely chosen individual interests of all our Staff Members, the Insti- 
tution's "field" would literally be the history, structure, and organiza- 
tion of the universe. This Review will sample the efforts of both Staff 
Members and Fellows as they have applied their special talents and 
experience to problems that they consider especially attractive and sig- 
nificant from this great field. 

The Scientist's Concern for History 

Although the history of human society is eternally fascinating to 
men's minds, it is but a minuscule part of all history. The scientist, 



too, concerns himself with history — history of a complexity, and scope, 
and grandeur that inspire both awe and redoubled curiosity in any 
person prepared to investigate it. Some of the grandest and most mem- 
orable of all scientific visions have had a strong element of history in 
them, like Darwin's and Wallace's on the history of life, or Lyell's and 
Agassiz's descriptions of earth history, or Eddington and Jeans on 
stellar structure, or Hubble, Baade and Hoyle on cosmology. 

The investigation of structure is an interest closely associated with 
history. Structure is the stage on which cosmic, planetary, biotic, and 
other forms of evolution run their dramatic courses. Thus studies of 
the structure of the earth's crust and mantle, or of Galactic structure, 
or of intergalactic connections, to mention a few among the examples 
in this year's work, all ultimately contribute to an understanding of 
the grand history that many scientists find fascinating. 

In reading the departmental reports of the Institution this year one 
is struck with the number of investigations that are in the context of 
this universal history. Such are many of the research programs of the 
Mount Wilson and Palomar Observatories in their search for the true 
pattern of cosmic history, the programs of the Geophysical Laboratory 
and the Department of Terrestrial Magnetism to decipher the history 
of formation of the earth's crust and mantle, the studies of paleogeo- 
chemistry, experimental taxonomy, the evolutionary aspects of genetics, 
and others. 

Study of the Earth 9 s Crust and Mantle 

These interests in history and its structural setting are nowhere 
better illustrated than in the geophysical and geological investigations 
undertaken within the Geophysical Laboratory and the Department 
of Terrestrial Magnetism. Without knowledge of the context of earth 
history into which they fit, they easily appear to be a long series of 
recondite accounts of specialists' activities. The error of such an im- 
pression is very quickly revealed when one examines the relation of 
each research program to earth history. Almost every investigation 
has a carefully reviewed rationale; it has been specifically designed as 
a "probe" into the structure of the earth's crust or mantle, or as an 
historical indicator. 

It is not so many years ago that historical geology was a subject 
somewhat apart from other principal branches of geological studies. 
It depended heavily for its progress on paleontology, with assistance 
from geomorphology, stratigraphy, and volcanology. These subjects, 
of course, continue to illuminate the gradually unfolding history of 
the earth. Indeed, the recent and continuing discoveries of micro- 
paleontology have extended our knowledge of earth history through f os- 


sils to periods once thought unbelievably remote for such a record. But 
newer techniques have taken their place alongside the older approaches. 
The techniques of experimental petrology are prominent, but isotopic 
dating of minerals, seismology, heat-flow studies, crystallography, and 
others are making continual contributions. Their importance in ob- 
taining the full history of the earth's crust and mantle is very great 
because only these techniques can get at the record lying in all of the 
igneous and most of the metamorphic rocks that comprise such an over- 
whelming proportion of the crust and mantle. 

Among the many contributions to our knowledge of earth history pre- 
sented in Year Book 66, five illustrate the diversity of the year's work 
in geophysics and its relation to the grand design for understanding 
the record in the earth's crust and mantle. They are experimental 
petrology studies, biogeochemistry, seismologic investigations, earth 
heat-flow studies, and isotopic dating studies. 

Experimental Petrology. The group of minerals known as pyroxenes 
are a major constituent of the earth's mantle 1 and also are commonly 
encountered in the earth's crust. They are a very diverse group. Al- 
though closely related crystallographically they have varied chemical 
compositions and textures resulting from different physical conditions 
of formation such as temperature, pressure, and oxygen fugacity. 
Pyroxenes are associated with basaltic rocks, metamorphic rocks, 
layered igneous intrusives, and with some less commonly encountered 
rocks thought to have been formed deep within the crust (ultramafic 
nodules of kimberlites). 2 P. H. Abelson, Director of the Geophysical 
Laboratory, observes in his report that petrologists are hopeful that 
study of the pyroxenes "will be an extraordinarily useful tool in deter- 
mining complex petrogenetic history." A number of studies of the 
pyroxenes were carried out at the Geophysical Laboratory during the 
year; from them two illustrations are given: study of the phase equi- 
libria of two pyroxene minerals known as hedenbergite and wollastonite, 
and the electron-probe study of exsolution in pyroxenes. 

Some of the older intrusive igneous rocks that outcrop at the earth's 
surface are found in the Skaergaard intrusion of Greenland, the Still- 
water intrusion in Montana, and the Bushveld complex in South Africa. 
Minerals from all of them have been studied at the Geophysical Labora- 
tory for their indications of the conditions of mineral formation within 
these rock bodies. 

D. H. Lindsley, G. M. Brown, and I. D. Muir investigated the phase 

1 That part of the earth's interior between the Mohorovicic discontinuity and the core. 
^ 2 Ultramafic refers to minerals having- a relatively high content of iron and magne- 
sium; kimberlite is a granular rock composed chiefly of pyroxene, olivine (Mg-Fe2Si04) 
and biotite mica, found in the diamond "pipes" in South Africa and elsewhere. 



relations of two pyroxenes characteristic of the upper layers of the 
Skaergaard intrusion, hedenbergite, Ca,Fe(Si 2 06), and a pyroxene 
derived from solid solution between wollastonite, CaCSiO.-J, and ferro- 
silite, FeSiO.3. By experiment the pressure-temperature conditions 
under which inversion (change from one mineral structure to another) 
takes place was determined ( Fig. 1 ) . Results from these experiments 
were combined with previously obtained experimental data on the in- 
version of two other minerals from the same rock body, quartz and 
tridymite, both Si0 2 . The combination of these data yields a well- 
defined model for the conditions of crystallization of these rocks. 
Lindsley, Brown, and Muir conclude that "crystallization of upper zone 
c of the Skaergaard intrusion must have taken place at 600 ± 100 bars 
and over a temperature interval that included the range 900° to 950 °C." 
With these data inferences can be made about the depth within the crust 
at which the crystallization took place. The Lindsley-Brown-Muir 








Z 3 

Pressure, kilobars 

Fig. 1. Diagram showing phase relations for green pyroxene from Skaergaard intrusion. Amount 
of shading approximately proportional to amount of hedenbergite in run products. Wo ss , ferrifer- 
ous wollastonite (Ca,FeSi 2 06); HD SS/ hedenbergite (Ca,Fe); Trid, tridymite (Si0 2 ); Qtz, quartz. 


experiments might be described as employing the classic methods of 
experimental petrology, using pressure-temperature apparatus. 

F. R. Boyd and G. M. Brown examined some other pyroxenes with 
the aid of an instrument new to the Geophysical Laboratory, the elec- 
tron microprobe. This instrument produces a very finely focused beam 
of electrons. 1.0 0.1 microns, which, upon impact on a specimen, 
generates X rays that can rive accurate information on element com- 
position. Very successfully employed for several years in metallurgy, 
its use in mineralogy has been limited because of problems attending 
the analysis of elements of low atomic number. As F. R. Boyd explains 
in a section of his report of this year, the probe permits nondestructive 
quantitative analysis of mineral grains at least as small as 8 to 10 
microns. He adds, "The degree of homogeneity as well as the compo- 
sition of a material can be estimated in probe analysis, and this capacity 
will probably revolutionize our concept of equilibrium in mineral 

Boyd and Brown applied the probe to analysis of exsolved 3 pyroxene 
lamellae from the Skaergaard and Bushveld intrusions. These la- 
mellae are considered of particular interest because of their contrast 
in calcium content to the crystals within which they occur. The con- 
trasting compositions appear to be related principally to the tempera- 
tures of crystallization. They are also of interest in determining the 
extent of cation migration and ordering in the fine texture of rocks. 
Semiquantitative analyses were made of augite lamellae (calcium rich) 
in hypersthene (magnesium rich) and hypersthene lamellae in augite. 
These 8-to-30-micron-thick lamellae were discovered to have the same 
chemical composition as "host" crystals of the same mineral (Fig. 2). 
However, the crystal symmetry was found to differ, resembling the 
host augite in the case of the hypersthene lamellae (Plate 1). From 
these data and those of other experiments Boyd and Brown state that, 
for the two minerals considered, extensive cation migration over dis- 
tances of 20 to 30 microns took place within the host crystals at tem- 
peratures between 1050° and 1000°C, and that migrations over shorter 
distances (2 microns) took place below 1000 °C. Although these studies 
must be considered as only exploratory, they strongly suggest that the 
electron probe may be expected to yield valuable information where its 
scanning is applied to the fine structure of diagnostic minerals in rock 

For a number of years G. Kullerud and his co-workers have pursued 
another aspect of experimental petrology that concerns much more 
limited sections of the earth's crust, but sections that have been of enor- 

3 Solid solutions stable at high temperature can separate into phases with different 
chemical compositions on slow cooling. This process is known as exsolution. 

Plate 1 . Microphotographs in polarized light of pyroxene crystals from the Stillwater, 
Skaergaard, and Bushveld intrusions showing exsolved lamellae. 

A. Orthorhombic bronzite crystal with thin lamellae of calcium-rich pyroxene exsolved. 
From Stillwater intrusion, Wyoming, United States. 

8. Inverted pigeonite" crystal, now orthorhombic hypersthene with thick augite lamellae 
in herringbone pattern. Texture suggests exsolution of augite of twinned monoclinic pigeonite, 
prior to inversion of the latter to orthorhombic hypersthene. Skaergaard intrusion, Greenland. 

C. Inverted pigeonite crystal showing a second, thin set of lamellae exsolved in hypersthene, 
presumably after inversion from pigeonite. Bushveld intrusion, South Africa. 

D. Complex pattern of exsolution in the hypersthene host of an inverted pigeonite crystal 
base). This is additional to the coarse, preinversion lamellae and the parallel thin lamellae 
top right). The angle between these two sets of lamellae does not suggest exsolution of 

coarse lamellae along the (001) plane of pigeonite; cf. (C). Bushveld intrusion. 

Plate 1 

Report of the President 



Hw OW l »l^ » " ' W * « i 


Fig. 2. Electron-probe scans across pyroxene crystals containing exsolved lamellae of another 
pyroxene phase. G—J represent scans of augites from the Bushveld formation. Differences in cal- 
cium content between "host" and lamellae are depicted. E, scan across a silver lamination in 
steel, monitoring the specimen current, and testing sensitivity of the microprobe. Slopes of the 
sides of the peak indicate a beam size of 1.0—1.1 microns. F, scan across the same lamination 
shown in E, monitoring the FeKa X ray. X-ray spot size is 1.6—1.9 microns. 


mous importance to human society. These are ore bodies, many of 
which are associated with sulfide mineral systems. Experiments have 
been concentrated especially on the copper-iron-nickel sulfides asso- 
ciated with norite' country rock in the Sudbury district of Ontario, 

In 1966-1967 the Geophysical Laboratory continued the analysis of 
mineral relations within the sulfide systems and the relation of the sul- 
fides to adjacent silicates." The investigations of this year extended 
knowledge of phase relations of the Sudbury ores through a temperature 
ranee from about 1000° to below 400°C. Knowledge of the higher- 
temperature behavior of the combined sulfide and silicate systems is 
essential to an understanding of the manner of separation of the sul- 
fides from the silicates in the original magma, and study of the lower- 
temperature phase relations is necessary to account for the specific 
character of the final mineral assemblages. 

A. J. Naldrett and G. M. Brown studied the iron-magnesium ratio 
of the paired pyroxenes from the ore-bearing norite, hypersthene, and 
augite, the same minerals studied by Boyd and Brown in their micro- 
probe study of lamellae, to establish the temperature of cooling from 
magma of the silicates. The ratios found indicate that sulfides and 
silicates were exposed together to a temperature of at least 1100°C. 
Thus the sulfide ores very probably are formed at or above 1000 °C. 
Kullerud and H. S. Yoder, Jr., added further information about the first 
formation of the sulfide through study of the mutual solubilities of sili- 
cate and sulfide liquids. They found that they do indeed coexist in the 
liquid state. They believe that the relation of specific ore deposits to 
specific rock types results from a partitioning of metals between the 
sulfide and silicate minerals. 

J. R. Craig, Naldrett, and Kullerud examined the succession of min- 
eral assemblages that form in the solid ore mass as it cools. The wealth 
of minerals in these assemblages is illustrated by a partial list of those 
encountered in the Sudbury ores: pentlandite, chalcopyrite, cubanite, 
violarite, polydymite, millerite, covellite, sillimanite, awaruite, and 
idaite, among others. Laboratory study has shown that some of these 
minerals are formed at temperatures as low as 400 °C. Thus there now 
exists something of a model of the history of formation of one of the 
most important ore deposits in the world. 

4 A gabbro in which more than half of the pyroxene is orthorhombic. A gabbro is a 
plutonir- igneous rock of granitoid texture containing feldspar and pyroxene minerals. 
r >. Kullerud, J. R. Craig, and A. J. Naldrett were full-time participants in sulfide 
studies. Part-time collaborators included G. M. Brown, S. W. Richardson, H. S. Yoder, 
Jr., P. M. Bell, and J. L. England. In addition, G. Moh of the University of Heidelberg 
collaborated in some of the work. 


Bio geochemistry. Biogeochemistry offers an approach to still other 
problems of localized geological history that have unusual interest for 
human society. For several years P. H. Abelson and his colleagues at 
the Geophysical Laboratory have been examining the diagenesis 
(change in form) of organic matter in sediments. Heat treatment of 
sediments in the Laboratory can give clues about the organic reactions 
taking place under natural conditions. These reactions are of special 
concern because of inferences that may be drawn about the forma- 
tion of petroleum products within the earth's crust. 

This year R. M. Mitterer and T. C. Hoering demonstrated that a 
mild heat treatment of the insoluble organic matter (kerogen) in a 
Recent marine sediment obtained from the San Nicolas Basin off Cali- 
fornia produced a number of substances commonly found in petroleum. 
Exposure of a sediment to temperatures on the order of 200 °C for a 
few days caused the chemical reduction (hydrogenation or hydrogenoly- 
sis) 6 of the long, straight-chained structures in kerogen to normal, 
saturated hydrocarbons ( Fig. 3 ) . Isoprenoid hydrocarbons and porphy- 
rins, molecules with structures characteristic of those used by living 
organisms, were also produced from kerogen by such treatment. 

Model experiments, in which added olefins and alcohols were reduced 
to saturated hydrocarbons when heated with the Recent sediment 
showed that such reactions can occur. Mitterer and Hoering conclude 
that their experiments provide support for the view that kerogen, a 
relatively inert, organic high-polymer, is a major source of the hydro- 
carbons found in petroleums. 

Explosion Seismology as a "Probe" One of the few probes that can 
penetrate anything more than a relatively thin surface layer of the 
earth is a seismic wave, the result of a shock at some point within or 
on the earth. Seismic waves are, of course, produced by earthquakes, 
but they can also be produced by explosions and by various other 
methods. For many years explosion seismology has been used with some 
effectiveness to obtain information about the crustal structure of the 
North American continent, and most recently for the Coastal Plain 
and Appalachian Mountain region of the Middle Atlantic states. In 
what has been called the East Coast Onshore-Offshore Experiment 
(ECOOE) , D. E. James, T. J. Smith, and J. S. Steinhart of the Depart- 
ment of Terrestrial Magnetism have collected seismic data that they 
have now interpreted in a map of crustal thickness extending from 
central North Carolina to central Pennsylvania (Fig. 4). This is the 
first three-dimensional model of crustal structure of the eastern North 

6 Hydrogenation is the chemical reaction of a substance with molecular hydrogen in 
the presence of a catalyst. Hydrogenolysis is the cleavage of a molecule associated with 
the addition of hydrogen. 









Hydrocarbons- 225 °C 

San Nicolas Basin sediment 


















Column temperature 75-325°C 47m in 

Fig. 3. Gas chroma tog ram of normal hydrocarbons extracted with benzene-methanol from 
San Nicolas Basin, sediment heated to 225°C. Peak numbers indicate the number of carbon 
atoms in the hydrocarbon detected. 

American area based on physical observation. The map shows a crustal 
thickness varying between 30 and 60 km. The thinnest area underlies 
Washington, D. C, and adjacent Virginia, and the thickest areas are 
found directly beneath the crest of the Appalachians, showing a pro- 
nounced "root" structure beneath the mountains. The "root" is broadest 
and deepest where the highest elevations occur. Gravity values have 
been computed from this model that correlate well in size and distribu- 
tion to the actually observed regional gravity anomalies. 

James, Smith, and Steinhart infer from their data that the mean 
atomic weight of rocks in the upper mantle is lower than that of rocks 
in the lower part of the crust. They hypothesize that an upper mantle 
composed of peridotite 7 or dunite 8 and a ferromagnesian (possibly 

7 A coarse-grained rock containing a magnesium silicate (olivine) and pyroxene 

8 A coarse-grained, essentially monomineralic rock, mainly composed of olivine. Some 



























,— -* 










































amphibolite") lower crust is consistent with their observations. Their 
observations would seem to rule out eclogite, 10 a previously hypothesized 
component of the upper mantle. 

Earth Heat Flow Measurements. Steinhart, Smith, and S. R. Hart, 
also of the Department of Terrestrial Magnetism, participated in an- 
other type of investigation seeking to delineate crustal structure and 
history. This was the measurement of heat flow from the earth's crust 
at its surface. The Lake Superior basin was chosen for this investiga- 
tion partly because it had been the site of a previous explosion 
seismology study. Heat-flow values were calculated for 83 sites in the 
lake, thus providing the most comprehensive picture of local and re- 
gional heat-flow variation yet obtained for a continental area. 11 These 
measurements demonstrated also for the first time that large systematic 
heat-flow variations on the continent occur on a regional scale. A varia- 
tion in flow values by a factor of more than two was found in the Lake 
Superior measurements (Fig. 5). 

Inferences can be made about the thickness of the crustal structure 
from heat-flow measurements because the source of the heat is in radio- 
active elements concentrated in the upper crust. Low heat-flow areas 
would indicate a thin crust, and vice versa. This interpretation corre- 
lates with previous seismic observations. Thus there would seem to 
be a thick section of the crust in the central part of the lake and a thin 
section at the western end. These observations seem to confirm infer- 
ences from seismic observations that the crust below the lake is one 
of the thicker parts of the crust of North America. As Smith, Steinhart, 
and L. T. Aldrich have hypothesized elsewhere, this unusual crustal 
section may be a rift (down-faulted) structure of great age, possibly 
having occurred in Precambrian times, making it what might be called 
a fossil rift. 12 

Isotopic Rock Dating. Isotope dating is a final illustration of a method 
of seeking out events in earth crustal history. Application of this method 
has been undertaken for a number of years by a group composed of 
members of the staffs of both the Department of Terrestrial Magnetism 
and the Geophysical Laboratory. 13 

Dating techniques based on interpretation of the decay of radioactive 

9 A monoclinic silicate of calcium and magnesium and, usually, one or more other 
metals, as iron, manganese, etc. 

10 A coarse-grained mafic rock, mainly pink garnet and green pyroxene. 

11 Measurements were made from the United States Coast Guard cutter Woodrush, 
stationed on the lake. 

12 T. J. Smith, J. S. Steinhart, and L. T. Aldrich, "Crustal Structure Under Lake Su- 
perior," The Earth Beneath the Continents, Washington, D. C, 1966, p. 196. 

13 This year S. R. Hart and L. T. Aldrich of the Department of Terrestrial Magnetism, 
T. E. Krogh and G. L. Davis of the Geophysical Laboratory, and F. Munizaga and A. M. 
Stueber, Carnegie Institution Fellows, participated. 










.y o 

-o — 

.£ c 


o -a 



= >» 


Z "s 





J o 

o > 

*z o 


w pr 




o c: 


q o 


-■ o 



a> o 


"? «« 

=t "> 



£ F 

,_ _D 

*" CD 

• D 

c £ 

O > 

— d) 


in 3 


■S "2 

> £ 



£ *S 

g is 

qz v 

C Q_ 

a a 


Fig. 5. Contoured he 
indicate location and v> 


elements have faced some difficulties in field application because results 
can seem conflicting or contradictory. For example, age determinations 
on some minerals in a section of Ontario east of Lake Huron, known 
as the Grenville province, gave ages of about 900 ± 100 million years 
for a mountain-building event. On the other hand, other geological 
evidence suggested that the rocks from which these ages were obtained 
are much older. Using what is called a whole-rock rubidium-strontium 
method of dating, T. E. Krogh and G. L. Davis had established ages of 
1500-2000 million years for rocks in the area. It had been assumed that 
the 900-1000 million-year event was one in which the rocks of the region 
underwent deep burial, intense heating, and plastic deformation, of 
which they show unmistakable signs. By the application of carefully 
designed sampling techniques to the different rock bodies of the prov- 
ince, Krogh and Davis have been able to show that the 900-1000 million- 
year metamorphism was preceded by much older metamorphic events. 
Rubidium-strontium dating analysis for both granites and paragneisses 
of the Province show very close agreement at 1500-1800 million-year 
ages, indicating that the Grenville rocks were already intensely meta- 
morphosed at the time of the 900-1000 million-year mountain building. 

The Astronomer's View 

Of all the views of history the grandest, of course, is that of cosmic 
history, whose sweep defies imagination. Difficult as it is for a layman 
to conceive of a light year (5.88 trillion miles), billions of light years 
are part of the everyday thought of astrophysicists and astronomers. 
The astronomer's sole source of information is the radiation — visible, 
infrared, radio, ultraviolet, X-ray — that can be detected at the surface 
of the earth and by space probes. Yet it is a very rich source in astrono- 
mers' hands, and from it we now have a tentative conceptual structure 
of the universe and its history — comprising events on so vast a scale 
that all of human history by comparison is submicroscopic. Much that 
is contained in our present concept of the universe derives from infor- 
mation obtained on the instruments of the Mount Wilson and Palomar 
Observatories, sponsored by the Institution and California Institute 
of Technology. 

Only a relatively few years ago all astronomy, with the minor excep- 
tion of cosmic-ray detection, was optical. But for more than two decades 
radio-wave detection has been an increasingly rich source of informa- 
tion. Recently X-ray astronomy has been added with the aid of sounding 
rockets. Since about 1930, optical astronomy also has been extended 
by exploitation of the infrared part of the spectrum. As the year's work 
at the Mount Wilson and Palomar Observatories illustrates, the tend- 
ency is toward an integration of information from a source object 


through as many means of detection as possible. Optical astronomy from 
ground-based instruments, however, remains the foundation of our 
slowly emerging concepts of what the universe is and has been. 

It is most reasonable to assume that the universe is an expanding 
one — the more distant an object, the greater its velocity of recession 
from the earth. It extends at least several billion light years outward, 
but each year seems to bring the discovery of an object with a still 
greater redshift 14 than that of the previous year. A favorite hypothesis 
as to the universe's origin is that of a "big bang," or primeval explosion, 
some (possibly 10) billions of years ago. Discoveries at the Observa- 
tories gave new information on both these matters, and on others of 
cosmic interest. 

Quasi-Stellar Sources and Objects. One of the most profound mys- 
teries in the sky is that of the quasi-stellar sources (QSSs), whose 
extremely high energy output and other unusual qualities were first 
detected at the Observatories. One of them, called Parkes 0237-23, was 
shown this year to have the largest redshift measured to date. Its emis- 
sion line redshift is 2.22, as determined by spectra obtained by H. C. 
Arp and T. D. Kinman (Lick Observatory). J. L. Greenstein and 
M. Schmidt also determined two separate sets of absorption lines for 
Parkes 0237-23 as having redshifts of 2.202 and 1.956. In addition to 
their large redshifts, Parkes 0237-23 spectra indicate a relatively high 
density for the object and an unusual "double shell" in which ionized 
metals are very strong. 

A number of other studies added substantially to the facts about 
radio-quiet quasi-stellar objects (quasars or QSOs). A. R. Sandage 
of the Observatories and W. J. Luyten of the University of Minnesota 
continued a study of the faint blue objects found in high galactic lati- 
tudes. A 69-object photometric sample indicated that these blue objects 
are dominated by radio-quiet quasars and white dwarf stars. Some of 
the QSOs showed relatively large redshifts. Sandage and Luyten esti- 
mated that the density of QSOs in the star field examined, known as 
the Haro-Luyten blue-star field, may be 0.5 per square degree, substan- 
tiating an earlier prediction by Sandage that QSOs are very numerous. 

Following examination of the Haro-Luyten field, Luyten and Sandage 
extended their search for QSOs to seven other fields. Preliminary 
analysis of photometric data obtained with the 200-inch telescope for 
88 objects indicates that many of the objects in the photometric sample 
are good QSO candidates. Sandage and Luyten estimate that there are 
at least a hundred thousand QSOs up to a limit at magnitude 19.7 and 
that the number is an increasing function of limiting magnitude. 

14 The redshift is the displacement toward the red of an observed spectral line as 
viewed in a receding source. 


M. Schmidt studied the space distribution of radio-emitting quasi- 
stellar sources and found that their distribution in space increases with 
increased redshift. If the density is taken to be 15 at the distance of 
redshift 0.5, it is 100 at redshift 1.0. 

Knowledge of quasi-stellar sources was also extended in other ways. 
Schmidt obtained redshifts for 10 additional QSSs during the year. 
J. B. Oke has examined a number of the QSSs for variability. Two 
highly variable sources, 3C 279 and 3C 446, have been studied in detail. 
Changes in luminosity of as much as 0.25 magnitude per day have been 
observed for 3C 279, the visual magnitude of which has varied by 
nearly 2.0 during the past year. For both sources the spectral con- 
tinuum is somewhat redder when the object is fainter. Two observed 
spectral lines (magnesium II of 3C 279 and carbon IV of 3C 446) re- 
mained constant, independent of change in the continuum. J. Wampler 
of the Lick Observatory and Oke also studied the source 3C 273 in 
great detail with photoelectric scanners. They have made an estimate 
of electron density in the emitting region of this source, and estimate 
it to be at least 10 6 per cubic centimeter. 

Sandage also determined the redshifts of nine new radio galaxies 
as part of a program of determining redshifts of all identified radio 
galaxies from the Third Cambridge Catalogue, a project in which 
Schmidt is also participating. 

An X-Ray Star. One of the most unusual astronomical objects studied 
during the year concerned the very strong X-ray source Sco X-l, the 
position of which had been determined by the X-ray astronomy group 
of American Science and Engineering, and Massachusetts Institute of 
Technology, in a series of spectacular rocket experiments in 1966 and 
early 1967. The optical identification was made at Tokyo Observatory, 
Japan, and in a series of photoelectric measurements made by Sandage 
with the 200-inch telescope. From night to night variations of 0.8 mag- 
nitude were observed, as well as large night-to-night variations in 
spectral-line intensities. Optical brightness varied by as much as d=0.03 
estimated at about 400 parsecs 15 and the X-ray flux was estimated at 
about 10 37 ergs per second in X rays at wavelengths of 1-10 angstroms. 

Later observations of Sco X-l by J. A. Westphal and Sandage sought 
time resolutions of the object's emission for periods as short as two 
seconds. Using pulse-counting equipment designed and built by the 
Astro-electronics Laboratory of the Observatories, their further obser- 
vations showed the source to be highly variable in time intervals as 
short as one minute. Nonrandom variations with intervals of 4 minutes, 

15 One parsec — 3.258 light years. 


30 minutes, and 60 minutes were also indicated. J. Y. Jugaku found 
evidence for radial velocity changes from night to night. 

Redetermination of the Hubble Constant. One of the continuing 
problems of cosmological astronomy is the calibration of its distance 
indicators. The Hubble Constant, which gives the increase in velocity 
of an object's recession from the earth per unit of distance, 16 is the 
end result of this calibration. A preliminary value for the Constant 
was determined many years ago. As a first step in its redetermination, 
G. Tammann and Sandage have completed measurements and analysis 
of the variable stars in the galaxy NGC 2403. This is a spiral galaxy 
and was the first beyond the Local Group of galaxies in which cepheid 
variable stars were found soon after the 200-inch telescope began opera- 
tion in 1949. Since that time a series of color plates has been made 
of the galaxy over a period of years, and a photoelectric sequence be- 
tween 1962 and 1965. From this material 56 variable stars have been 
located, including 17 cepheids. Sandage has applied five methods to 
determine the distance modulus of NGC 2403, using cepheids, the 
apparent magnitude of the red supergiant variable stars, the apparent 
magnitude of the brightest resolved stars, calibration of bright irregu- 
lar blue variables from prototypes in the nearer galaxies M 31 and 
M 33, and the angular size of Hydrogen II regions. The five methods 
agree remarkably well, and a much more exact distance determination 
for this galaxy is expected when the calculations are complete. The 
distance to NGC 2403 is considered scientifically important because it 
will permit calibration study of other galaxies that will lead directly 
to certain galaxies that have measured redshifts. When that is done 
an improved value of the Hubble Constant can be calculated. 

A Blue Star Anomaly. Evidence of some interest in determining the 
validity of the "big bang" cosmological model was obtained by J. Green- 
stein and Oke in spectrograms and spectrophotometric scans of blue 
stars in the galaxies M 13, M 15, and M 92. According to the "big bang" 
model these stars should have a helium-to-hydrogen ratio of nearly 30 
per cent. The spectrographic and spectrophotometric results were in 
direct contradiction to the model's prediction. Helium I lines are very 
weak or absent in stars whose ultraviolet blue visual spectra ( UBV) 
demand their presence in the galaxies M 15 and M 92. Earlier observa- 
tions, including those of Greenstein and G. Munch, were thereby con- 
firmed. It is considered a serious anomaly for the predictions of the 
"big bang" model. 

Stellar Observations. Two further sets of observations are of particu- 
lar interest for what they tell us about the sun and our Galaxy. A. J. 
Deutsch has studied for several years a group of stars that he has 

16 Present accepted value: 100 km/sec per 10 6 parsecs (3,258,000 light years). 


called "blue stragglers." It is his hypothesis that these stars are 
metamorphs of solar-type stars. On the basis of observations made of 
the blue stragglers he has shown that if the sun's interior contains 
as much angular momentum as the blue stragglers did when they were 
solar-type stars, the radiation entering the solar hydrogen convection 
zone will maintain the sun's present rotation against the torque exerted 
by the solar wind. This supports the hypothesis advanced by R. H. 
Dicke of Princeton University that the sun's interior rotates much 
more rapidly than its surface. Deutsch also accounts for the equatorial 
acceleration observed at the sun's surface by postulating that it arises 
from the viscous coupling between a rapidly rotating interior and the 
slowly rotating hydrogen convection zone of the surface. 

E. Becklin and G. Neugebauer studied the general infrared radiating 
structure of the central region of the Galaxy and compared it with the 
infrared structure of the nucleus of Galaxy N 31. The two galactic 
centers were reported to look similar in both shape and brightness in 
their infrared structure. Thus our Galaxy has an analogue in this 

Astrophysics and Astronomy at the Department of 
Terrestrial Magnetism 

The well-known Mount Wilson and Palomar astronomy program has 
been supplemented for some years by a modest but imaginative astro- 
physical program at the Department of Terrestrial Magnetism. Pri- 
marily focused for most of its existence on radio astronomy and image 
tube development, the astrophysical activities of the Department of 
Terrestrial Magnetism broadened during the year. Optical astronomy 
observations, a radio astronomy discovery, a cosmic-ray study, and a 
nuclear physics experiment are reported as examples of the range 
of interest. 

Optical Astronomy. The optical astronomy program of the Depart- 
ment of Terrestrial Magnetism centered about the use of the DTM 
image tube spectrograph. W. K. Ford, Jr., and Vera C. Rubin report 
that a new spectrograph camera, designed by I. S. Bowen, former Di- 
rector of the Mount Wilson and Palomar Observatories, gives spectra a 
substantially improved quality as compared with earlier equipment. 
Among their observations were several probing the nature of "peculiar" 
galaxies, subjects of a study at the Mount Wilson and Palomar Observ- 
atories described last year. A commonly held hypothesis is that the 
morphology of at least some of the peculiar galaxies indicates that they 
represent galactic "explosions." Ford and Rubin's observations on the 
peculiar galaxies NGC 4038 and NGC 4027 showed that velocity varia- 
tions across these galaxies are so small that they cannot in fact be 
exploding galaxies. 



A Hydrogen "Bridge" from our Galaxy? One of the more interesting 
results of the year came from an observation of the southern sky. Sev- 
eral years ago a 100-foot parabolic radio telescope was designed by 
M. A. Tuve and his associates at the Department, and subsequently 
erected at La Plata, Argentina. This is now the principal instrument 
of the Instituto Nacional de Radioastronomia, under the direction of 
C. M. Varsavsky, a Research Associate of the Institution. F. Strauss, 
a student of Varsavsky, has commenced a search for an intergalactic 
bridge of neutral hydrogen between our Galaxy and the Magellanic 
Clouds, the nearest outside galaxies. Observations have traced an arm 
of hydrogen extending from the outer spiral arm of our Galaxy through 
at least 60° of galactic longitude. Measurements of the arm's velocity 
and position (Fig. 6) strongly suggest a connection of the arm with 
the Small Magellanic Cloud. Succeeding observations will show whether 
or not the present strong indications of a bridge to the Small Magellanic 
Cloud from our Galaxy does exist. If so, it will be the first indication 
of any long-period physical connection between this Galaxy and our 
galactic neighbors. 

/ = 270° 




■» II 








• SMC 


Fig. 6. Projection of outer arm of the Galaxy and Small Magellanic Cloud on the Galactic 
plane, showing extension of hydrogen in the direction of the Cloud. SMC, Small Magellanic 
Cloud; Kpc, kiloparsec, or 3258 light years. 

Cosmic-Ray Study. Probably the longest continued program of any 
scientist in the Institution is the cosmic-ray investigation of S. E. For- 
bush of the Department of Terrestrial Magnetism. Forbush's ioniza- 
tion-chamber observations and his data interpretations of cosmic-ray 
intensity began in 1937. They have been directed partly toward the 
diurnal variation of cosmic-ray intensity. His data now represent the 
longest series of such measurements available. Forbush's analysis of 
the 1937-1965 data, reported this year, provides a value for the average 
diurnal variation in cosmic-ray intensity (an amplitude of 0.15 per 
cent), and for the first time reliably shows that there is a cyclical 
change in the diurnal variation. The amplitude of the cyclical variations 
is about 50 per cent of the diurnal amplitude with a period of two solar 


cycles, or about 20 years. Zero values for this 20-year wave occurred in 
1988, 1948, and 1958, The 195S zero value coincided with the time given 
by H. W. Babcock for the reversal of the sun's general magnetic field. 
Forbush says that as yet there is no satisfactory theory for the 20-year 

Foil Excitation Experiments. One set of experiments at the Depart- 
ment of Terrestrial Magnetism shows that research in atomic physics 
need not always depend upon high energy "superequipment." L. Brown, 
YV. K. Ford, Jr., Vera C. Rubin, and W. Trachslin report the results 
of an experiment in which they use the foil excitation method of 
measuring ions in an excited state. The experiment was undertaken 
to test if a method could be developed for measuring the lifetimes of 
excited states, to be used to determine abundances of elements in stars. 
In the experiment a beam of sodium ions from a Van de Graaff acceler- 
ator passes through a thin carbon foil. The emerging atoms may be 
in any charge state from neutral to completely ionized. If excited, the 
atom will radiate under conditions like those of a free atom. The glow- 
ing beam that has passed through the foil is imaged on a lens and the 
image is then passed into an image tube spectrograph. The procedure 
gives a novel slant-lined spectrogram (Plate 2). The charge state of 
ions can be determined in a procedure of Trachslin's invention. The 
results from the experiment were somewhat unexpected in that only 
a few of the lines could be assigned to known transitions in sodium. 
It is therefore not yet clear whether the method will eventually afford 
data that can be used in determining element abundance in astronomi- 
cal objects, but it is evident that a new field of atomic studies relating 
to cosmological problems has been opened to investigation. 

Biochemical and Biophysical Views on Problems of Evolution 

Breathtaking as an excursion into the sweep of cosmic history can 
be, most people find new views of the history of life equally fascinating. 
At least two windows have been opened to such views in the work of 
the year — one in the study of racemization of amino acids in fossil 
shells by P. E. Hare and P. H. Abelson of the Geophysical Laboratory; 
the other, in the study of "saltatory replication" by R. J. Britten and 
D. E. Kohne of the Biophysics Section of the Department of Terrestrial 

Racemization of Amino Acids with Time. For more than a hundred 
years it has been known that biological substances may differ in optical 
properties. Commonly this is measured by the rotation of light shown 
within an instrument called the polarimeter. A substance may be opti- 
cally inactive, dextrorotatory (Z)-form), levorotatory (L-form), or 
racemized ('equal amounts of L-forms and D-forms) . Most amino acids 

Plate 2 

Report of the President 


<&»/ w? SvJf *m 

j 4 I "**** 

Plate 2. A part of the foil-excited spectrum 
of sodium. Above and below the sodium spec- 
trum are comparison lines from an iron-neon 
arc, for four of which the wavelength in ang- 
stroms is given. The unique characteristics of 
the sodium lines are the decreasing intensity of 
the lines downstream from the foil and the slant 
of the lines caused by the Doppler effect. 

5* / 7 ^jjiS|S*j&: 

I %J / imm- 

-J O 


are optically active, and one of the great biochemical puzzles is 
why the proteins of all living forms consist principally of L- 
amino acids. The universal use of the L-form has long attracted the 
attention of scientists interested in the origin and evolution of life. 
Why was one stereo-isomer chosen? 

Hare and Abelson attacked one facet of this problem during the year 
by examining the optical configuration of the amino acids extractable 
from Mercenaria shells of three geologic ages : Recent, upper Pleistocene 
(^60,000 years), and upper Miocene (10 million years). Nine amino 
acids were still detectable in the most ancient shell. As expected, the 
amino acids in the Recent shells were almost entirely in the L-form. 
But those in the Miocene shells were almost evenly divided between the 
L- and L-forms, that is, they were racemized. The Pleistocene amino 
acids were partly racemized. One of the most interesting results of 
the shell amino acid analysis was the discovery that the fossil shells 
contain almost no isoleucine. On the other hand, they contain alloisoleu- 
cine, nearly all of it in the L-form. Isoleucine and alloisoleucine have 
somewhat different chemical properties. Isoleucine, for example, is 
slightly less soluble in water. Alloisoleucine is not found in proteins. 
The apparently normal process of racemization that yields products 
of nearly identical physical-chemical behavior for most other amino 
acids, in the case of isoleucine gives products easily separable by non- 
biologic processes. 

From these results Hare and Abelson suggest that the primitive 
ocean contained a racemic mixture (equal amounts of L-forms and 
L-forms) of amino acids. The mystery of how the L-form became the 
preferred protein form still remains; indeed, it may be greater than 
before. If some unknown process produced a slight preponderance of 
either the L-form or the L-form, racemization would restore the equi- 
librium. On the other hand, once a route to the selective extraction of 
the L-form was assured, then L-form supplies could be replenished by 
racemization from the L-form as the L-form was used. 

Repeated Sequences in DNAs. The report of the Biophysics Section ir 
of the Department of Terrestrial Magnetism commences with these 
words: "A central theme runs through this report as it has in past 
years : the enormous potential for understanding the relationships and 
history of creatures through the similarity of nucleotide sequences in 
their DNA." 

R. J. Britten and D. E. Kohne report this year on their exploration 
of repeated sequences of DNA in different species representing all the 
major forms of life. In 1964 Britten and Waring discovered that a 

" E. T. Bolton, D. B. Cowie, and R. B. Roberts, Staff Members; D. E. Kohne, Staff 
Associate; D. J. Brenner and A. Rake, Fellows; S. Falkow, Visiting Investigator. 


type of mouse DNA known as satellite DNA was composed of a set of 
DNA sequences repeated with astonishing frequency. Indeed, one short 
nucleotide sequence was repeated about a million times in this DNA. 
This proved to be an extreme example of DNA repetition, but 10,000 
and 100,000 sequence repetitions appear to be more common. Britten 
and Kohne have pushed forward this year with a versatile and power- 
ful set of investigations leading to some hypotheses that may have a 
remarkable impact on our view of the evolutionary process in life at 
a very fundamental level. The DNAs investigated range all the way 
from those of viruses to those of human beings. Fifty-five different 
DXAs from the plant and animal worlds were analyzed for the char- 
acter of their nucleotide sequences. Among them only the DNA from 
viruses, from bacteria (E. coli, Clostridium perfringens) and from 
blue-green algae failed to show repetition in the genome. In some cases 
i e.g.. salmon sperm) as much as 80 per cent of the cell DNA seems 
to be made up of repeated sequences. 

In Britten's hypothesis the sets, or "families," of related nucleotide 
sequences that will reassociate with each other result from "saltatory 
replication'' — events that produce families of hundreds of thousands 
of similar nucleotide sequences in the DNA of an organism within 
a relatively short time span, perhaps of no more than one generation. 
Families of DNA sequences may be shared between or among species, 
but diverge more and more as the hypothetical time elapses after speci- 
ation increases. In short, relatedness of the DNAs decreases with time 
following species branches (Plate 3) . 

Britten and Kohne state that divergence among the members of a 
family of repeated nucleotide sequences occurs during the evolution of 
a species. They believe further that the degree of divergence may be 
a measure of the age of the family of sequences. Thus a knowledge of 
nucleotide family divergence is of obvious interest in considering the 
history of the evolution of life forms. In studies of the reassociation 
of nucleotide pairs (homology) of present-day chicken and rhesus 
monkey DNAs, Britten and Kohne's results indicate that a large frac- 
tion of the two DNAs existed as repeated DNA sequences at the time 
those species lines diverged hundreds of millions of years ago. 

In addition to the families of repeated sequences found widely dis- 
tributed among higher organisms, the DNA of all organisms includes 
a component of nonrepeated nucleotide sequences. Both the repeated 
and the nonrepeated elements change slowly with time, but Britten 
and Kohne's experimental data suggest that the average rate of change 
of the repeated sequences is less than that of the nonrepeated. 

Britten and Kohne conclude their report with the remark that "the 
wide occurrence of families of repeated sequences . . . indicates that 
sudden events may be far more important to evolution than we have 

Plate 3 

Report of the Vreniderd 

^■^^^^»MH?F?*bIv^^" : 

■ ;F. 11 



%•■ ". .v,'?; vi " 


CtSSItxfl lift} 




**« .', 




. .- 

..' , 

;p- - - - - - 

4* :&"■'» <-m' ■■:■■'.-■ 

:, ■ - * 







Plate 3. Schematic diagram suggesting the history of families of repeated DNA sequences. 
The right-hand scale indicates the period of time since the saltatory replication that produced 
each family. The left-hand scale — as a measure of divergence — represents the thermal 
stability of the reassociated pairs of DNA strands formed by the members of a family. The 
height of the peaks indicates the amount of DNA of a given age and thermal stability. 


heretofore suspected." They suggest the possibility that events can 
occur whereby a large number of new genes can arise relatively rapidly 
in the genome, 18 many of which remain unexpressed in the organism's 
phenotype, 19 adding to the reservoir of genetic potential. They note 
further that "mutation, translocation, and recombination with other 
genes would yield a whole range of potential genetic activity [that 
could] turn up at times long after . . . the saltation." Thus does the 
individual organism carry within itself not only the historical record, 
but also the makings of future evolutionary history. 

A Widening Interest in Genetics — The Department of 
Embryology as an Example 

One of the striking implications of the reports of investigators in 
biological subjects at the Institution this year is of the key position 
now occupied in wider biological realms by subjects once the domain 
only of genetics. The investigations of Britten and Kohne illustrate 
this. The work of the Department of Embryology, offering another 
equally important example of this trend, has focused on a study of 
developmental processes in the organism. J. D. Ebert, Director of the 
Department, states that even as recently as 1960 it was not uncommon 
for embryologists to take the chromosomes and genes as "constants" in 
the equation of the development of the organism. Today, however, lead- 
ing embryologists everywhere are deeply concerned with the labile role 
of genetic materials. As Ebert remarks : "Today it is a cardinal rule 
that a detailed knowledge of biochemistry and morphology of normal 
development of an organism is essential if we are to understand the 
role that genes play in its development." 

The bridging of the gap between studies of the gene and those of 
the differentiating cell in recent years has been made possible in part 
by the rapid progress in knowledge of the molecular basis for gene 
action. As Ebert states in his introduction, information contained in 
the DNA molecule is first transcribed in the developing organism into 
an RNA copy, generally within the cell nucleus. The transcribed (RNA) 
copy of the gene then moves to the cytoplasm of the cell, where the 
linear information coded in the RNA nucleotides is translated into a 
corresponding linear sequence of amino acids. These amino acids in 
turn are linked to form polypeptides and are finally converted to pro- 
teins. Even though a great deal remains for discovery before a complete 
description of the process of translation can be made, 20 enough is known 

18 A complete set of all the different chromosomes normal to a species. The genetic 
content of a cell composes a genome. 

19 The expressed characteristics of the individual organism. 

20 C. R. Woese, D. H. Dugre, S. A. Dugre, N. Kondo, and W. C. Saxinger, "On the 
Fundamental Nature and Evolution of the Genetic Code," Cold Spring Harbor Labora- 
tory of Quantitative Biology Symposia on Quantitative Biology SI, 1966, 723-736. 


to make this general model a serviceable bridge between the traditional 
domains of genetics and embryology. Furthermore, it is now clear that 
the transcription of the genetic code in every cell is highly regulated. 
There are at least three levels of "controls" of gene action. The trans- 
lation of the information contained in its DNA differs from cell to cell 
and may differ from time to time in the life history of a given cell. 
The frequency with which a given gene is transcribed into RNA may 
also vary. Finally, RNA copies of genes may or may not be translated 
into proteins. It is clear why exploration of the operation of such regu- 
lators of genie expression has become central to the study of develop- 
mental biology. 

Research in developmental biology treats, in any eukaryotic orga- 
nism."' the most complex biochemical systems known. The isolation of 
specific systems with measurable relations is not easy, and may be one 
reason that the bridge between the gene and the differentiated cell 
was so long uncrossed. However, D. D. Brown, of the Department of 
Embryology, and his colleagues believe that they have found a model 
system for study of the control of gene action in ribosomal RNA 
I rRNA) . The organism selected several years ago by Brown as particu- 
larly suitable for analyzing this system was the clawed toad Xenopus 
laevis. The wisdom of the selection and the ingenuity of the experi- 
mental technique have been amply proved by a series of significant 
discoveries during recent years. The present year has been typical. 

Brown, C. S. Weber, and J. H. Sinclair this year used a molecular 
hybridization technique employing Millipore filters to extract from 
Xenopus still further significant information about the relation of the 
genes (DNA) and ribosomal RNA. Ribosomal RNA can be separated 
into three fractions designated, according to rate of sedimentation, as 
28S, 18S, or 5S. Brown, Weber, and Sinclair found that the genes for 
the 28S and 18S fractions are highly redundant (with about 800 copies 
for each type), closely adjacent on a single chromosome. 22 The genes 
for these types are thought to be alternating. Even more interesting, 
these investigators have found that the functionally related genes for 
the 5S RNA are not physically linked to the genes involving 28S and 
18S RNA. Brown and his colleagues made this discovery by compara- 
tive analysis of DNA and rRNA during early stages of growth in an 
abnormal mutant of Xenopus lacking a nucleolus in the homozygous 

Using the same methods, Brown, Weber, and Sinclair also showed 
that somatic tissues of Xenopus with a wide range of rates of ribosomal 
synthesis contain very similar numbers of genes for the three types of 

21 An organ i nm having a true nucleus. 

22 More specifically an autosome, that is, a nonsex chromosome. 



rRNA. However, egg cells have an enormously greater number of genes 
for the 28S and 18S RNAs. Thus something of a paradox appears. 
Although somatic cells 23 synthesize ribosomes at greatly differing rates 
with complements of genes that do not vary, the egg cell displays a sev- 
eral hundredfold increase of genes for the two types of ribosomal RNA 
(Fig. 7). Experiments show that the excess genes or rDNA remain 
inert in the cytoplasm during cleavage of the fertilized egg and are 
ultimately degraded as the cells of the growing organism proliferate. 
Brown, Weber, and Sinclair's observation of the anucleolate mutant 
of Xenopus adds to the mounting evidence of the existence and nature 
of regulatory mechanisms operating throughout the development of 
an organism, "turning on" or repressing rRNA synthesis. Thus the 
anucleolate mutant failed to synthesize 5S rRNA although it was 

23 Any cell of the body except egg or sperm cells. 








Early gastrula 




Heterozygote ( l-nu) 








Tube no. 

Fig. 7. Hybridization of Xenopus laevis rRNA with various DNAs. The undotted curve in each 
chart shows the fraction of the genome homologous with 28S rRNA, 1 8S RNA, and 5S RNA 
in each type of cell. Even though the genome remains essentially the same, its rRNA expression 
differs strikingly among the cell types. 


demonstrated that the gene for this fraction was present in the anucleo- 
late cells. The absence of an activating factor, or the presence of a 
repressing factor, must be postulated to account for the behavior of 
the 5S RNA gene. If one visualizes the "construction" of a ribosome 
that requires all three DNAs, action by the linked 28S and 18S genes 
and the independently located 5S genes is required. Although the two 
sets of genes are in separate locations, they must be coordinated. The 
activator, whatever its nature, appears to be associated with the 28S 
and 18S part of the genome. As Ebert observes, the search for the nature 
of this mechanism should be very interesting. Its discovery would be 
highly significant. 

Igor Dawid, also working with Xenopus, has firmly established that 
the cytoplasm of its eggs contains substantial amounts of a DNA that is 
associated with the cell mitochondria. 24 Moreover, there is abundant 
evidence that DNA is an integral component of mitochondria in a wide 
variety of organisms. It seems possible that this may be another general 
characteristic of eukaryotes. Dawid and his colleague, David Wolsten- 
holme, state that their results can be "interpreted as suggesting the 
preservation of mitochondrial DNA during evolution." In a series of 
elegant studies they have also established that mitochondrial DNA in 
Xcnopus has the structure of a "twisted circular" molecule (Plate 4). 

Genetics at the Genetics Research Unit 

Genetics as a subject of research in the Institution goes back almost 
to its beginning. No subject in the Institution has had a longer history 
of distinguished accomplishment. Indeed, the whole question of the 
"bridge" between the genes and the development of somatic tissues, 
now opening in an exciting way for developmental biology, has been a 
subject of continued research by Barbara McClintock of the Genetics 
Research Unit for more than 30 years. 

A Control System in Maize. Again this year, Dr. McClintock reports 
on yet further analysis of one of the "bridges" between genes and their 
somatic expressions in maize : the mechanism of genie control that she 
has called the Spm (suppressor-mutator) system. Her report this year 
concerns observations of the effects of the Spm system on the gene locus 
in maize for anthocyanin synthesis — the production of the blue or red 
anthocyanin pigment in the plant body or in seed kernels. Her observa- 
tions are of special interest because they demonstrate that a single 
control system can provide an astonishing diversity of regulation in 
gene expression. 

The Spm system includes what Dr. McClintock calls component-1, 
the suppressor, and component-2, the mutator. Component-1 might be 

24 Organelles found in the cytoplasm of every cell except bacteria and blue-green algae. 

Plate 4 

Report of the President 

*' * * # ,* -« f 7 %* felt* •- *%Zf/'\ 

' ***-f" !':■/ Kf&l+LL. riff >« H *#■ v ' 4 * "- > * • ■ 


*^ m % 

Plate 4. Electron micrographs of Xenopus laevis (toad) DNA prepared by Wolstenholme. 
All micrographs are X 85,000. A-C, Molecules of native DNA from oocyte mitochondria. 
A, B, Twisted circles. C, An open circle with a contour length of 5.6 microns. 

riate 5 

Report of the President 










































































































1 5 



































considered the "switching" mechanism for the anthocyanin gene. If it 
is active the anthocyanin gene produces a uniform pigmentation; if 
it is not active no pigment is produced. However, the type and distribu- 
tion of the pigment depend on the activity of component-2. If com- 
ponents is inactive and component-1 is active the aleurone layer of 
cells of a maize kernel will be lightly pigmented. If component-2 is 
activated, deeply pigmented spots appear on the lightly pigmented back- 
ground. The size of the superimposed dark pigmentation depends on 
the stage of component-2 activation; the earlier the activation, the 
larger the number of clonal cells with dark pigment. If component-2 
is activated in the gamete all cells are deeply pigmented. Component-2 
is called the "mutator" because it induces responses that modify the 
organization of the anthocyanin locus and through it the consequent 
somatic expression of the anthocyanin gene. 

By isolating maize lines that contain one or another of these com- 
ponents in the gametes, with the mutator component in at least two 
different forms, Dr. McClintock has produced a great variety of pheno- 
typic expressions in maize kernels. Among other interesting disclosures 
she has shown that the Spm element in a cell may influence pigment 
production in neighboring cells, even when no pigment is produced in 
the cell itself (Plate 5). Dr. McClintock states that the patterns that 
result from the action of the control systems on the anthocyanin genes 
"are so varied that they defy a meaningful classification." She adds, 
"Only the fact that anthocyanin pigment ... is not vital to the plant 
makes it possible to learn about the many kinds of regulation that such 
a system can provide." She suspects that what may have been inter- 
preted as the expression of different alleles in some past genetic studies 
may represent instead different "states" of the gene loci as she has 
described them in the case of the anthocyanin genes. She concludes that 
the term "alleles" in any specific instance may be ambiguous unless 
there is a means of distinguishing between a mutant of the structural 
gene itself and a mutant that is produced by a component of a regula- 
tory system such as the Spm system. 

Phage Lambda and Other Viruses. A. D. Hershey, Director of the 
Genetics Research Unit, includes in his report this year two general 
reviews, one of the "state of the art" in the biochemical genetics of 
bacteriophages and the second, of the meaning of that state for more 
comprehensive views of genetics. In the first he summarizes what is 
known about sexual and asexual conjugation and genetic recombina- 
tion in bacteria, especially as it has been revealed in studies of 
Escherichia coli and related species. In the second he discusses the 
relations between a particular phage, lambda, and its host bacterium, 
E. coli. 


Hershey notes that at least seven mechanisms provide opportunities 
for exchange of genes in bacteria. None is typically sexual, although 
several involve conjugation between differentiated cell types. Genes are 
passed from cell to cell by mechanisms that may differ as to origin of 
transferred genes, manner of transfer, and mode of replication in the 
carrier cell line. One of the most intensely studied of these mechanisms 
is transduction, in which extra chromosomal DNA fragments or 
viruses may transfer genes from one bacterial line to another. Dr. 
Hershey has been particularly interested in description of the genome 
of the virus or phage genome. 

One class of transducing phage (e.g., coliphage PI) operates by 
picking a phage-sized section of bacterial DNA more or less at random 
and incorporating it into the phage particle. This DNA segment may 
then be transferred (transduced) to another E. coli line, where it is 
incorporated into the genome. Hershey notes that this process mimics 
sex at the population level, because all bacterial genes are transduced 
by PI with similar frequencies. 

Phage lambda, which has been a favorite material in Hershey's 
laboratory, operates very differently. It specifically transduces the genes 
responsible for galactose 25 and biotin 26 metabolism in E. coli, but no 
other genes. Another phage of the same kind, called <£80, transduces 
tryptophan-forming genes. Hershey states, "Lambda is important . . . 
because it can recombine . . . cellular and viral inheritance in ways 
that are fascinating to contemplate and, very likely, of practical im- 
portance to humans. " 

About 20 genes have been ordered thus far on a "genetic map" for 
phage lambda. This is at least half, and possibly more, of the genome 
of this phage. Hershey says that the length of the DNA of lambda is 
about 44,000 nucleotide pairs, which could accommodate 30 or 40 genes. 

Lambda has three capacities of particular interest to geneticists. Like 
a typical phage it can infect and lyse bacterial cells, destroying them 
and producing numerous phage progeny. But it can also take up 
extended residence in the bacterial chromosome, giving rise to viable 
host-cell lines of modified inheritance. These bacterial cells are said 
to carry prophage and are called lysogenic. Third, lambda can form 
transducing phage lines specific to a single gene. 

Hershey discusses what he believes to be the most satisfactory model 
of the relation of phage lambda to its bacterial host in the transduc- 
tion process. The model is that of A. M. Campbell, formerly an Associate 
in Research at the Department of Genetics and now at the University 
of Rochester Department of Biology. The model derived from Campbell's 

Galactose is a constituent of the milk sugar lactose. 
26 C 10 H 16 N 2 O 8 F. Biotin is one of the B vitamins. 



discovery, while he was at the Department of Genetics about ten years 
ago, of what are now known as the suppressor-sensitive mutants of 
lambda. These mutations are of special interest because they can be 
found for any gene that has an essential function. Hershey states that 
the Campbell model for prophage excision and insertion (Fig. 8) is 
strictly analogous to deletion and insertion in all the formal models for 
the genesis of any chromosomal rearrangement. Here is one more dra- 
matic example of the importance of detailed probing into the structure, 
the permutation of forms, the chemistry, and other qualities of what 
may superficially seem the tiniest part of the life process. As Hershey 
remarks in another connection in his report, "What we see on reflection 
is an intimation of history, another witness interrogated as to what 
living things are about." 

Among the interesting features of lambda transduction brought out 
in Hershey's review are : ( 1 ) Crossovers between phage and bacterial 
DNAs are nonequational, that is, the length of phage DNA deleted 
and of bacterial DNA inserted in the phage chromosome are unequal. 





^ A phage particles 

I 1 

A R 



I ! 

at crossover 



9 a 



Xdg phage 

Fig. 8. Structural relations of lambda wild-type phage, lambda prophage, and lambda dg 
(defective) according to A. M. Campbell's model. Genetic maps are indicated by the phage 
genes A, J, N, and R and the bacterial genes gal (galactose) and bio (biotin). Phage and 
bacterial components are indicated by shading; terminal cohesive sites by arbitrary symbols: 
pp stands for loci of permutation points; ii for loci of prophage insertion sites (loci that become 
crossover regions ip and pi in the prophage map). 


(2) Measurement of the deleted parts of the phage DNA molecules 

suggests that there are about 1100 nucleotide pairs per gene (assum- 
ing that all genes in the deleted region are known). 

The base content of three distinct sections of the lambda DNA mole- 
cule was analyzed in greater detail during the current year than pre- 
viously. Elizabeth Burgi, Anna Marie Skalka, and Hershey continued 
to use the content of guanine-cytosine as a chemical marker, and found 
that the shorter the molecular segment the less it resembles its neigh- 
bors in composition. Laura Ingraham and Hershey found that the 
measurable base-sequence similarities between lambda phage and E. 
coli DNAs were strongest near one end of the lambda DNA molecule 
and weakest in the center. They suggest that their results in this case 
argue for "a recognition device that does not depend on homology 
alone. 1 ' This could possibly be a special enzyme that directs normal 
insertion and excision of the prophage. 

Dr. Skalka and H. Echols (University of Wisconsin) identified dur- 
ing the year two genes in lambda whose primary function may be the 
control of transcription. One is concerned with nearly all messenger 
RNA synthesis during phage growth, and the other specifically facili- 
tates transcription of genes responsible for late functions. R. Werner 
demonstrated that the rate of DNA synthesis in another phage, T4, 
depends on the number of "growing points" rather than on variation in 
rate at a given point. The rate of synthesis at a given point appears to 
be constant. He suggests that his results indicate an unanticipated role 
for genetic recombination : to distribute "growing points" over newly 
synthesized DNA. 

Study Systems that Include the Environment 

The "bridge" between the gene and the growing organism has been 
mentioned as one of the ways of adding to our understanding of the 
development of life. One other way of searching for understanding 
is to study life in its environment as a cell, as a group of cells, or as 
an entire organism. 

The Environment in Embryological Study. The Department of Em- 
bryology for many years has studied a very special biological environ- 
ment, the uterine environment of the mammalian embryo. For several 
years investigators, both inside and outside the Department, have 
sought ways to sustain the embryo "living free" in culture, detached 
from its natural maternal environment. The capacity to culture a 
mammalian egg under rigorously defined conditions straight through 
from fertilization even to the time of uterine implantation would be a 
very important scientific event. D. G. Whittingham, in cooperation with 
John Biggers and R. P. Donahue of Johns Hopkins University, has 


taken the technique of embryo culture a step further than it had 
reached previously. They have succeeded in culturing a mouse zygote 
(fertilized egg) to the early two-cell stage in a simple, chemically de- 
fined medium and from the late two-cell stage to the blastocyst stage 27 
under similar conditions. He has discovered, however, that an interrup- 
tion of the culture at the early two-cell stage is necessary. The zygote 
will not develop into a blastocyst unless it is exposed to the environment 
of the fallopian tube 28 for a short period between the first and second 

Whittingham's experiments have generated some significant infor- 
mation about the biochemistry of the development of zygote into blasto- 
cyst. He found that the first cleavage division in vitro requires an 
energy source, and a very specific energy source, either pyruvate or 
oxaloacetate. Cleavage will not take place unless one of the two is pres- 
ent in the culture medium. It will not take place in the absence of an 
energy source, or in the presence of other compounds like lactate, or 
glucose, that might logically seem a source of energy. Whittingham also 
found that normally unacceptable energy sources can be used and will 
produce cleavage if the follicular cumulus cells 29 are placed in the 
medium with the one-cell embryos. The follicular cells thus appear able 
to metabolize lactate and glucose and another compound, phosphoenol- 
pyruvate, into the pyruvate necessary for the fertilized egg's first 
cleavage. Pyruvate also appears to be secreted by the oviduct, as was 
demonstrated by the cleavage of zygotes in the presence of ovarian 
fluid alone, or in an organ culture of the fallopian tube. Thus it has been 
established that pyruvate is the essential energy supplier for first 
cleavage of the zygote. It was also shown that pyruvate is made avail- 
able to the early mouse embryo from two sources, the follicular cumulus 
cells and the fallopian tube epithelium. The question of why exposure 
to the fallopian tube environment is required in the early two-cell stage, 
and only then, has not been answered. It remains a most interesting 
problem for further exploration. 

Study of Plants in Their Environment. Consideration of the environ- 
ment as part of a system for study has long been a subject for the 
Department of Plant Biology. Indeed, the Desert Botanical Laboratory, 
a predecessor of the Department of Plant Biology, was deeply concerned 
with study of the desert environment more than 60 years ago. This 
traditional interest of the Department is being carried forward dynam- 
ically today by its Experimental Taxonomy Group. Their work is well 

27 The blastocyst is a hollow sphere of cells produced by cleavage from the original 
zygote just before the migration of cells to specialized sites at which differentiation 
within the specific organism commences (gastrulation). 

28 The duct leading from an ovary to the uterus. 

29 Cells associated with the follicle within which the ovum develops. 


illustrated by the unexpected results of experiments conducted this year 
by 0. Bjorkman and his collaborators. They have found a way to make 
many plants, including some common crop plants, grow at as much as 
twice the normal rate by a very simple but theoretically startling 
expedient. In experiments with beans (Phaseolus vulgaris) and monkey 
flowers i Mi mill us cardinalis) it was found that reducing the amount of 
oxygen in the air surrounding the plants accelerates their growth. 

The average oxygen content of the atmosphere at the earth's surface 
is 21 per cent. In Bjorkman's experiments the oxygen content of the air 
around the tops of the plants was reduced to either 5 per cent or 2.5 
per cent. The roots were given air of normal oxygen content. Bean 
seedlings placed in the 2.5 per cent oxygen environment for six days 
grew 2.1 times as rapidly as they did in normal air. Growth rates for 
longer periods also were notably higher (Plate 6). Monkey flowers 
placed in air of the same oxygen content for 10 days increased their 
growth by 58 per cent. In a 5 per cent oxygen environment their growth 
increase was 90 per cent. A significant exception to the growth increase 
pattern appeared in experiments with maize. Corn grown in 5 per cent 
oxygen increased its rate of growth by only 16 per cent as compared to 
ordinary air. 

These experiments were an extension of Bjorkman's earlier work on 
the photosynthetic rates of plants maintained in air containing less 
than the normal amount of oxygen. As reported last year, Bjorkman 
found that many species of plants increased their normal rate of photo- 
synthesis by 50 per cent when the oxygen concentration of air around 
the leaves was lowered. It is Bjorkman's tentative hypothesis that the 
inhibition of photosynthesis caused by oxygen at the level of normal air 
is caused primarily "by a back-reaction between a highly reduced 
photosynthetic intermediate and molecular oxygen that increases with 
0, concentration." It is noteworthy that certain tropical grasses, in- 
cluding maize and sugar cane, which lack any major signs of photo- 
synthetic inhibition by oxygen, probably use a different "pathway" for 
the fixation of carbon dioxide from that employed by most other plants. 

C. S. French, Director of the Department of Plant Biology, states in 
the Introduction to his Departmental report that "these results open a 
new field of inquiry relating to basic differences that have evolved in 
the photosynthetic mechanism of higher plants." In addition to its 
theoretical significance, Bjorkman's discovery has some immediate 
practical value in that it will aid more rapid production of plants for 
laboratory studies. It is not too much to hope that increase in plant 
growth rate by such a simple change in environment eventually may 
have much broader utility. 

In a more traditional study Jens Clausen, retired Staff Member of 

Plate 6 

Report of the President 






















the Department of Plant Biology, has published some observations 
about the significance of plant species clusters. Clausen states that 
modern data indicate that most of the world's tree species are arranged 
in clusters of species (cenospecies) . Such clusters circle the earth within 
definite latitudinal belts, and are closely enough related to permit 
occasional genetic interchange. Examples of these clusters are the larch 
genus Larix, which includes six species: the two-needle pine cluster 
(type species, Finns silvestris) , the alders (Alnus), the birches 
(Betula) , the cypresses (Cupressus-Chamaecyparis) , the beeches 
(Fagaceae) , the white oaks {Quercus alba) , and others. Clausen argues 
for the application of a conservative concept of species and of genera so 
as not to lose the evolutionary significance of the species cluster. He 
says that "broad, deep-seated physiological and ecological relationships 
. . . built into the various heredity patterns . . . hold the [clusters] 
together even on their world-wide migrations." 

Progress in Study of Photosynthesis 

All higher green plants, the algae, some bacteria, and many simple 
marine organisms depend on photosynthesis for their existence. The 
entire life pyramid, including all higher organisms, is totally reliant 
on this process. Yet photosynthesis appears to be as recalcitrant to 
clarification at a fundamental level as it is basic to life on earth. The 
Department of Plant Biology has long been heavily committed to the 
problem of understanding photosynthesis, at the level of its highly 
complex biophysics and biochemistry. 

Green chlorophyll, a universal component of the photosynthetic ap- 
paratus in green plants, is contained in submicroscopic functional 
packages now labeled "photosynthetic units." Hundreds of chlorophyll 
molecules are known to comprise a single unit. Light absorbed by any 
one of these chlorophyll molecules is transferred from one molecule to 
another until it is caught by a particular kind of pigment molecule 
known as an "energy trap." Within that trap the energy of a light 
quantum is converted to chemical energy. Several hypotheses have been 
advanced as to the exact nature of the energy trap, and its precise 
manner of functioning is still a matter of lively scientific controversy. 

This year D. C. Fork and J. Amesz of the Department of Plant 
Biology studied the energy-trapping function in photosynthesis of a 
particular pigment known as P700. 30 P700 was discovered some years 
ago by Bessel Kok of the Research Institute for Advanced Study, Balti- 
more, a former Fellow of the Department. P700 has been consid- 
ered as one of two possible energy traps within the photosynthetic 

30 P700 is a special form of chlorophyll a that acts as a trap for system 1. 


process known as system 1. The other possible pigment was considered 
to be cytochrome /'. 

This year Fork and Amesz produced convincing evidence that P700 
and not cytochrome / is the energy trap. Experimenting with the red 
algae [Iridaea splendens, Schiaymenia pacifica, and Porphyra perfo- 
rata I they varied the ratio of the oxidized to the reduced forms of P700 
and cytochrome / by exposing the plants to varying light intensities. 
For each light intensity the fraction of both cytochrome / and P700 
present in the inactive oxidized form was measured. These fractions 
were compared with the rate of photosynthesis produced by a given 
light intensity. It was found that maximum rates of photosynthesis for 
these plants were maintained when cytochrome / was partly inacti- 
vated. On the other hand, rates of photosynthesis fell when P700 was 
partly inactivated. This is interpreted by Fork and Amesz as meaning 
that P700 is the primary energy trap for light absorbed by photo- 
chemical system 1. 

Another interesting result of the Fork and Amesz experiments was 
the discovery that a light quantum traveling into a P700 energy trap 
inactivated by oxidation may be passed on to another photosynthetic 
unit. They found that there is about a fifty-fifty chance that the travel- 
ing light quantum ( exciton ) would move on rather than being converted 
into heat and thus wasted. 

The Place of Fellows in the Institution 

Mindful that "the objects of the Corporation shall be to encourage, 
in the broadest and most liberal manner, investigation, research, and 
discovery . . . ," the Institution in recent years has evolved post- 
graduate scientific training as one of its important functions together 
with research. 

It is now a commonplace in the scientific world that an essential 
part of scientific training occurs in the postgraduate and postdoctoral 
years. Universities of course actively conduct this advanced training. 
But postgraduate education in the United States has more dimensions 
than even universities can give it. Industrial laboratories and engineer- 
ing departments of industries, government-operated research institu- 
tions — like the National Institutes of Health or the laboratories of the 
National Bureau of Standards, for example — privately managed inde- 
pendent research institutions, and others, all contribute richly to this 

Cytochrome / is an iron-containing' hemoprotein found only in plants and related 
to other cytochromes; prominent in intracellular oxidations. 

-' Carnegie Institution of Washington Articles of Incorporation, Section 2. 


Postgraduate training has become so necessary, so ingrained in our 
educational structure that postgraduate fellowships have become 
numerous; their attainment is now a sough t-f or, even an expected, 
event in the young scientist's career. The National Science Foundation 
alone supported 405 postgraduate Fellows in its fiscal year 1966, and 
285 in 1967. The National Institutes of Health supported 1237 in 1966. 
Like so many pragmatic adaptations that distinguish our cultural 
history, these fellowships have effectively met a primary need without 
involving the trauma of great formal change in our society. They are 
an admission that the fresh Ph.D. is not often a mature scientist fully 
in command of all the complex tools of his complex metier — litera- 
ture, equipment, theory, experimental technique, and professional 

The Carnegie Institution, responding to needs that its senior scien- 
tists have long understood, serves as one of these postgraduate training 
institutions. During the year the Institution's departments undertook 
reviews of their fellowship programs. Although the great value of these 
programs was already evident, the place that fellowship work gradually 
has come to assume in the life of the Institution still was something of a 
surprise when a summary was made. In the period 1952-1967 the 
Institution supported 203 Fellows for postgraduate work from its own 
funds; funds generously provided by the Carnegie Corporation sup- 
ported an additional 62. The National Science Foundation, the National 
Institutes of Health, other government agencies, and other foundations 
supported still others. Each of these Fellows was with the Institution 
at least one full year, and in many cases two, three, or more years, 
pursuing his postgraduate education. Many Guest Investigators also 
have worked with the Institution's departments during these same 
years, sometimes senior in their field, but more often younger scholars 
whose period of residence was less than a year. 

There are several observations that can be made about the Institu- 
tion's remarkably successful fellowship experience of these 15 years. 
First, the program has achieved a broad geographical distribution; 
Fellows of the Institution in this period came from 25 states of the 
United States and 33 foreign countries. Second, a very large percentage 
of the Fellows have devoted their subsequent careers to teaching or 
other academic work, or to a fundamental research in another type of 
research organization. Third, the Fellows as a group have been remark- 
ably productive in their subsequent careers, and an important influence 
in that productivity was their experience at the Institution. Finally, but 
of particular interest, nearly all of the Fellows who came to the 
Institution from foreign countries have returned to posts in their own 
lands. Of 27 foreign Fellows at the Geophysical Laboratory in the period 


mentioned, only one remained in this country at the time of the report, 
and he is expected to return home in the near future. Of 40 foreign 
Fellows who were in residence at some time during this period at the 
Department of Terrestrial Magnetism, only four have remained in the 
United States. 

Within the 15 years there were 132 Fellows at the Department of 
Terrestrial Magnetism and the Geophysical Laboratory. Sixty-five of 
the 132 now are in university teaching positions; 42 of the 65 have full 
professorships or other permanent posts, and all but a few of these 42 
are still below age 40. Fifty-four of the 132 are in posts where they 
carry on academic research; 6 have devoted themselves to applied 
arch. Thus, 90 per cent of the scholars who have been selected for 
postgraduate training at the Institution occupy key positions in the 
future progress of science. Among these scholars are a number who are 
already considered to be among the leaders in their specialities. The 
same judgment can be made for each of the other Departments of the 
Institution. The Department of Embryology, for example, has had 
about 70 Fellows in this period, including those supported by the United 
States Public Health Service and other outside sources of funds. Nearly 
all have taken up careers in teaching or academic research, and they 
include an impressive share of the younger leaders in this field. 

The productivity of these men is attested in part by the positions they 
now occupy and their scientific reputations. Another measure of pro- 
ductivity is given by papers describing the results of their experimental 
and theoretical work published in professional journals. This record has 
been truly gratifying to all whose vision endorsed and encouraged the 
fellowship program at the Institution. The Geophysical Laboratory, for 
example, compiled a bibliography of "Papers Arising in Part or Com- 
pletely from Work Performed by Visiting Staff." The 41 Fellows listed 
in this bibliography published 134 such papers between 1952 and 1967. 
Philip Abelson, Director of the Geophysical Laboratory, comments in 
his report on Fellows and Guest Investigators that they have, singly or 
in collaboration, participated in 70 per cent of the investigations de- 
scribed in the most recent annual report of the Laboratory. I can offer 
no more telling observation on the place that postgraduate training has 
in the professional life of the Institution's more permanent staff. 

The Fellows at the Institution are expected not only to participate in 
current research, but also to master pertinent subjects in which 
their previous training was inadequate. For example, Fellows at the 
Geophysical Laboratory frequently commence their work with inade- 
quate backgrounds in physical chemistry. As Abelson states, "They 
attain mastery of relevant aspects of physical chemistry during their 
tenure." At almost any time of year that might be chosen, it is likely 


that one or more seminars on special subjects of training interest will 
be going on at the Institution, including such topics as computer pro- 
gramming, genetic coding, irreversible thermodynamics, or quasars. 
Without exception the Departments of the Institution consider it im- 
portant that Fellows undergo intellectual growth during their tenure. 
One former Fellow recently wrote about his experience at the Institu- 
tion, "It has enabled many of us to acquire skills and insights not 
available in even the best-equipped graduate schools. . . ." 

There is an equivalent benefit for the Institution in this process, of 
course. Thus Ellis Bolton, Director of the Department of Terrestrial 
Magnetism, writes from the point of view of the staff, "Our Fellows 
have had . . . and will have the profoundest kind of influence upon the 
work of the Department as it progresses." The best of the fellowship 
experiences, and this includes a majority, have left both Staff Member 
and Fellow intellectually richer. 

We expect to continue to provide these opportunities for both staff 
and Fellows at a rate that members of our staff believe is feasible. No 
less than by its substantive contributions to our knowledge of the his- 
tory and structure of the universe, the Institution would wish to be 
known by these results of the compelling personal experiences of young 

Losses . . . 

It is with great sadness that I record here the death of a former 
Trustee, Elihu Root, Jr., on August 27, 1967, in New York City, at the 
age of 86. Mr. Root had just retired as Trustee of the Institution at the 
Annual Meeting on May 5, 1967. The occasion marked the first time 
since the founding of the Institution that there has been no member of 
the Root family on the Board of Trustees. Mr. Root's father, the Secre- 
tary of State and the Secretary of War, was elected in 1902 to serve on 
the Institution's original Board. He remained in that position until his 
death in 1937, when Elihu Root, Jr., was elected a Trustee. 

During his 30 years as a Trustee, Elihu Root, Jr., took an extraor- 
dinary part in the affairs of the Institution. From 1950 to 1956 he 
served as Chairman of the Board. After his work in that capacity was 
completed, he continued most actively in the affairs of the Institution, 
both on the scientific and administrative sides. Rarely indeed did he 
miss a meeting of the Executive Committee, or of the Finance Com- 
mittee, on which he was also active over many years. He was a constant 
inspiration to his colleagues on the Board, and to the Administrations 
that served with him. 

Elihu Root was born in New York City in 1881. Like his father and 


his son, he was graduated with a B.A. degree from Hamilton College in 
Clinton, New York. After earning an M.A. from Hamilton, he went on 
to Harvard Law School where he received the LL.B. degree in 1906. 

Shortly after his graduation he founded the firm of Root, Clark and 
Bird with two of his young law school classmates. One of his first clients 
as a young attorney was Andrew Carnegie, a man who, as Mr. Root 
said, "did not hold lawyers in awe." The firm of Root, Clark and Bird 
ultimately became one of the most distinguished in the nation and 
claimed many notable personalities among its clients. Mr. Root stayed 
with the firm, which eventually became Dewey, Ballantine, Bushby, 
Palmer and Wood, until 1954, when he became counsel to the firm of 
Cleary, Gottlieb, Steen and Hamilton. 

Although his law practice was a busy and stimulating one, Mr. Root 
found time to serve simultaneously as a director of several large corpo- 
rations, including the Mutual Life Insurance Company of New York 
and American Telephone and Telegraph. 

A patron of arts and letters as well as an artist himself, Mr. Root 
was a trustee of the Grand Central Art Galleries, New York's Metro- 
politan Museum of Art, and the New York Public Library. With char- 
acteristic thoughtfulness, he and his brother Edward assembled a 
notable collection of oil paintings and prints, which they donated to 
Hamilton College to be borrowed by the students for their rooms. 

Twice Mr. Root's career was interrupted by what he called "the 
unfortunate business of making war." During World War I he saw 
active service as a major in the 304th Infantry in France. And at home 
he was a pioneer in the Plattsburgh Movement, established to set up 
voluntary military training camps for civilians. By the time war was 
formally declared, it had furnished thousands of trained junior officers 
to the nation. During World War II Mr. Root served as a member of 
the U. S. Army Air Corps strategic target board, which advised on 
points where enemy economies should be attacked. For this work he was 
awarded the Medal for Merit by President Truman. 

These are but a representative few of the achievements and activities 
that occupied Elihu Root's busy life. His absence will be keenly felt by 
all who knew him, and his memory will be treasured for many years to 

With equal sadness I record here the passing of another fellow 
Trustee, another former Chairman of the Board of the Institution, and 
a dear friend — Barklie McKee Henry. He died on September 4, 1966, at 
Blue Mountain Lake in the Adirondacks. His personal interest in the 
Institution was constant and intense; his wise counsel was invaluable 
and inspiring. 


Mr. Henry became a Trustee in 1949 and at once assumed a most 
active part in guiding the Institution and shaping its philosophy in the 
context of its role as a unique research organization. In 1961 he became 
Chairman of the Board of Trustees and remained in that position until 
1965. Through all the years of his Trusteeship he brought the formi- 
dable range of his talents and abilities to bear on the fortunes of the 
Institution. No problem was too difficult for him to tackle, nor any detail 
too minute to escape his notice. He was a frequent and welcome visitor 
both in the Office of Administration and at the various departments of 
the Institution. Staff Members looked forward particularly to his visits 
and the opportunity to seek his thoughtful, understanding, and knowl- 
edgeable advice. 

Mr. Henry was born in Ventnor, New Jersey, in 1902, and was grad- 
uated cum laude from Harvard University in 1924 with an A.B. degree. 
After some further study at Balliol College, Oxford University, he be- 
came managing editor of the magazine, Youth's Companion. A year 
later he joined the Guaranty Company of New York. 

In 1930 Mr. Henry left his business career to devote himself fully to 
those things in life that he valued most. Among these was his interest 
in a rapprochement between medicine and psychiatry, which he did 
much to encourage through his work with the New York Hospital. After 
serving the Hospital in various other capacities, he became its president, 
and was influential in strengthening the psychiatric services it offered. 
The inventive side of his personality served him well in many instances. 
For example, as president of the New York Hospital he saw to it that 
the grim bars in the psychiatric ward were replaced by mesh screens, 
now in wide use. 

In the early days of the Depression he was active in the Emergency 
Unemployment Relief Organization in New York, an unusually success- 
ful self-help undertaking at the local level. He later managed the merger 
of two similar organizations to form the Community Service Society, 
the largest private family welfare agency in the United States, of which 
he became the first president. 

In 1940 he went on active service with the Navy and during the war 
commanded first a coastal mine sweeper, then a submarine chaser, and 
finally a destroyer escort. Later, he was called upon by Secretary of the 
Navy Forrestal to help with contract terminations. 

In addition to all these activities he was a trustee of the Pierpont 
Morgan Library, the American Academy in Rome, the Rockefeller 
University, the John Hay Whitney Foundation, and the Institute for 
Advanced Study in Princeton. 

A list of achievements, however, does not do justice to Barklie 
Henry's life. For a list does not intimate the manner in which these 


achievements were brought about. It does not reveal the vibrance or the 
resolution he brought to all things he encountered. 

Barklie Henry is sorely missed for his skill and his acumen, his in- 
sight and his brilliance, his understanding and his great warmth. But 
most of all he is missed as a friend and a companion, because above all 
else, he was a profoundly humane human being. 

It is with a very special sense of loss that I must record the resigna- 
tion from the Board of Trustees of Mrs. Margaret Carnegie Miller, 
which was accepted, though reluctantly and with keen regret, by the 
Board at its meeting on May 5, 1967. Mrs. Miller is deeply missed. But 
we are particularly happy that she remains near us, accessible and 
available to give of her advice to the Institution that she has known and 
watched from its very beginning. 

Although with the retirement of Dr. Barbara McClintock on June 30, 
1967, the Institution lost as a Staff Member one of the most distin- 
guished American geneticists of all time, I am particularly happy to 
record that she has consented to return as the third Distinguished 
Service Member in Institution history, and will continue her research 
program without interruption at the Genetics Research Unit in that 

The Achievement Award of the American Association of University 
Women was presented to Dr. McClintock in 1947. On April 30, 1965, 
Cornell University bestowed the honorary office of "Andrew D. White 
Professor-at-large" on Dr. McClintock, and on April 24, 1967, she was 
awarded the Kimber Genetics Medal of the National Academy of 
Sciences "for a series of brilliant and significant studies on chromosome 
structure and function. " 

And Gains . . . 

We are fortunate indeed this year in welcoming two new Trustees of 
high distinction — Eric Ashby and William McChesney Martin, Jr., both 
elected to the Board on May 5, 1967. 

Sir Eric Ashby, F.R.S., was born in Bromley, Kent, England, in 1904, 
and was educated at the City School of London and the Imperial College 
of Science, University of London. During the two following years he 
was a Commonwealth Fund Fellow at the University of Chicago and 
served as a Fellow at the Desert Laboratory of the Carnegie Institution. 

In 1931 he returned to England to become Lecturer at the Imperial 
College of Science, and Reader in Botany at Bristol University. Later 


he served as Professor of Botany successively at the University of 
Sydney in Australia and at the University of Manchester in England. 
From 1950 to 1959 he was President and Vice-Chancellor of the Queen's 
University, Belfast, Northern Ireland. 

In 1942 Sir Eric conducted an inquiry for the Prime Minister of 
Australia into the enlistment of scientific resources during World War 
II. He was thus instrumental in organizing scientists for the war effort 
in Australia. In 1945 and 1946 he was Counsellor and Charge d'Aff aires 
at the Australian Legation in Moscow. 

His contributions to education are unique. He has pioneered the 
development of higher education in Africa, India, and Australia, as well 
as in Europe. From 1958 to 1960 he was Chairman of the now famous 
"Ashby Commission" from which the blueprint for higher education in 
Nigeria was derived. 

Sir Eric is at present Vice-Chancellor of Cambridge University. He 
continues to serve as Honorary Advisor to the Nigerian National Uni- 
versities Commission and as a member of the Council of the University 
College of Makerere, Uganda. 

William McChesney Martin, Jr., at present Chairman of the Federal 
Reserve Board, began his career in the Bank Examination Department 
of the Federal Reserve Bank in St. Louis in 1928. A year later he joined 
A. G. Edwards and Sons, also in St. Louis, and was a partner in that 
firm from 1931 until 1938. In 1931 he also became a member of the New 
York Stock Exchange, and in the same year edited and published 
Economic Forum, a scholarly journal published quarterly. He remained 
with the Exchange until 1941, serving successively as Member, Chair- 
man of the Board, and the youngest President in the history of the 

In 1941 Mr. Martin joined the Army, and by 1945 had risen from 
private to colonel. In that year he was appointed to the Board of Direc- 
tors of the Export-Import Bank, and in the two succeeding years served 
as Chairman of its Board. In 1949 he became the United States Execu- 
tive Director of the International Bank for Reconstruction and Develop- 
ment, and also served for a year as Assistant Secretary of the Treasury. 
In 1951 he was appointed Chairman of the Federal Reserve Board. 

Mr. Martin has received honorary degrees from Yale, Amherst, and 
Harvard, and from Washington University, St. Louis. He serves cur- 
rently as Trustee of Yale University, The Johns Hopkins University, 
and the Foreign Service Educational Foundation. 

The Institution eagerly anticipates a long and pleasurable associa- 
tion with both Sir Eric Ashby and Mr. Martin. 


It gives me the greatest pleasure to record the several honors that 
have come to Staff Members during the past year. 

Dr. Allan R. Sandage of the Mount Wilson and Palomar Observa- 
tories was awarded the Gold Medal of the Royal Astronomical Society 
for his fundamental work on stellar evolution and the history of the 
Galaxy, and for his further work in providing optical data basic to 
modern cosmology. In addition, Dr. Sandage was invited to be the 
Sigma Xi National Lecturer in October 1966, the Holiday Lecturer 
for the American Association for the Advancement of Science in 
December 1966, and the Vanuxem Lecturer of Princeton University in 
February 1967; and he was the recipient of honorary Sc.D. degrees 
from the University of Chicago and the University of Illinois. 

Dr. Alfred D. Hershey of the Genetics Research Unit received the 
honorary degree of Doctor of Science from the University of Chicago on 
May 5, 1967, at a special convocation celebrating the seventy-fifth 
anniversary of the University. 

Dr. James D. Ebert of the Department of Embryology was elected to 
membership in the National Academy of Sciences, and was also elected 
to membership on the Board of Scientific Overseers of the Jackson 

Robert P. Kraft of the Mount Wilson and Palomar Observatories 
was an invited lecturer for the American Astronomical Society at the 
summer meeting at Yerkes Observatory in June 1967 on Stellar Ro- 
tation, Stellar Evolution, and the Conservation of Angular Momentum. 

On December 16, 1966, Dr. Donald D. Brown of the Department of 
Embryology was named "Maryland's Outstanding Young Scientist of 
the Year" by the Maryland Academy of Sciences. 

The American Society of Zoologists elected Dr. David W. Bishop of 
the Department of Embryology as Chairman of their Division of 
Development Biology. 

Reports of Departments 
and Special Studies 

Department of Terrestrial Magnetism 

Committee on Image Tubes for Telescopes 

Department of Plant Biology 

Mount Wilson and Palomar Observatories 

Geophysical Laboratory 

Department of Embryology 

Genetics Research Unit 

of Terrestrial Magnetism 

Washington, District of Columbia 

Ellis T. Bolton 

L. Thomas Aldrich 
Associate Director 

Carnegie Institution Year Book 66, 1966-1967 


Introduction 7 

Geophysics 7 

Geomagnetism 8 

Cosmic-ray program 8 

Electrical conductivity studies of the mantle under the Andes ... 12 

Explosion seismology 13 

Time-term analysis — ECOOE 13 

Wide-angle reflections and amplitudes from ECOOE 24 

Late arrivals and amplitudes 25 

Earthquake seismology 28 

Q for P waves in the mantle 28 

Shear-wave absorption in the upper 130 km 29 

A path for high-frequency shear (S) waves in the upper mantle ... 31 

Converted waves from the interface at a depth of 410 km .... 32 

A search for small, deep earthquakes in the Andes 35 

On the spatial distribution of earthquakes near San Juan, Argentina . 37 

Isotope geology 42 

Potassium, rubidium, and strontium in ultramafic rocks and minerals . 42 

Geological history of the Grenville province 44 

Heat flow 52 

Technique and results 52 

Astrophysics 57 

Optical astronomy 57 

Radio astronomy 59 

Northern hemisphere 59 

Southern hemisphere 61 

Nuclear physics 62 

Elastic scattering of polarized protons on nuclei of spin zero .... 62 

Elastic scattering of polarized protons on deuterons 64 

Atomic physics 66 

Atomic and ionic spectroscopy with foil excitation 66 

Biophysics 68 

An instructive glossary . 68 

Repeated nucleotide sequences 73 

Time course of DNA reassociation 76 

Precision of sequence matching and fractionation 78 

Rate of divergence of nucleotide sequences during evolution .... 80 

Saltatory events of replication 83 

Chromosome pairs 86 

A source of divergence? 87 

Intramolecular heterogeneity of the DNA of temperate bacteriophages . 88 

Nucleotide sequence relationships among phages <£80 and A, and E. coli . 95 

DNA-DNA reactions involving identical DNAs 96 

Heterologous DNA-DNA reactions 96 

Qualitative aspects of microbial DNA duplexes 106 

Memory and learning mechanisms 118 

- Cited 124 

Bibliography 125 

Personnel 129 

Department of Terrestrial Magnetism 

i iiiii 

NGC 4038 









II IIIII 1 111 

Spectra of peculiar galaxy NGC 4038, taken with the DTM image tube spectrograph at Kitt 
Peak National Observatory. Line emission arises from bright knots in galaxy; these knots may 
be identified by insets at right showing orientation of spectrograph slit across galaxy. Emission 
lines going completely across spectra are due to night-sky radiation. 


The thrust of the Department of we eschew cooperation with others, 
Terrestrial Magnetism's efforts con- or fail to rise to the opportunities 
tinues to emphasize fundamental presented by sometimes very large 
work in physics. Our activities con- undertakings such as the seismic 
cern astrophysics, biophysics, and studies of the East Coast reported 
geophysics. Within each of these very here. Whenever we become involved, 
encompassing areas we have chosen however, the individual scientist is 
to explore limited problems that are intimately involved in every phase of 
both significant and approachable by the activity from, for example, dig- 
a small staff. We continue also to ex- ging holes for seismometers, to writ- 
press ourselves through personal re- ing programs for the computer and 
search in which the individual bringing wisdom to bear on the 
scientist, perhaps in concert with a meaning of the results. That this is 
colleague or two, becomes deeply im- a satisfying and valid pursuit, and 
mersed in his quest and remains a successful way of research in all 
unbuffered from gathering the ma- the areas we investigate, is keenly 
terials he needs or from using the illustrated in the scientific reports 
tools he requires. This is not to say that follow. 


L. T. Aldrich, R. Cabre, S.J., M. Casaverde, G. L. Davis S. del Pozo, L. Fernandez, S.J., 
S. E. Forbush, E. Gajardo, A. A. Giesecke, Jr., S. R. Hart, D. E. James, T. E. Krogh, 
H. Marconi, F. Munizaga, P. N. S. O'Brien, A. Rodriguez, G. Saa, I. S. Sacks, R. Salgueiro, 

T. J. Smith, J. S. Steinhart, A. M. Stueber, 
S. Suyehiro, and F. Volponi 

The geophysical description of our 
planet involves many of the mathe- 
matical and physical tools at man's 
disposal. Our report describes the 
efforts of a dozen scientists — Staff 
Members, Fellows, and Research As- 
sociates of the Department — to use 
these tools creatively. The report 
year started auspiciously with the de- 
livery of a "small" digital computer 
system (IBM 1130). It has been a 
much-used part of the efforts de- 
scribed below and has extended our 
capability most significantly. 

The range of topics undertaken by 
this group is in itself a demonstration 
of the varying needs of modern geo- 
physics. The statistical study just 
completed of the diurnal variation of 
the cosmic-ray flux bombarding the 
earth demonstrates the great value of 

a 30-year collection of carefully made 
observations, faithfully obtained and 
reported by colleagues in Christ- 
church (New Zealand), Cheltenham, 
Fredericksburg (United States), and 
Huancayo (Peru). The time-term 
analysis of some 800 seismic observa- 
tions made during the summer of 
1965 demonstrates the complexity of 
the requirements for describing the 
structure of the earth's crust over a 
large area, our good fortune in hav- 
ing friendly observers to obtain the 
data, and the need for an interested 
government agency to support the 
activities of the North American Ex- 
plosion Seismology Group. 

Intensive efforts in measuring 
thermal gradients, thermal conduc- 
tivity, and the bottom temperature 
regime of Lake Superior have pro- 



vided a new appreciation of the vari- 
ations in this parameter in sup- 
posedly stable shield areas of the con- 
tinents. For the first time, from these 
measurements one can plot meaning- 
ful contour maps of the rate at which 
heat tlows from the earth's surface. 
The study will be extended to the 
remainder of Lake Superior this 
summer with the helpful collabora- 
tion of the U.S. Coast Guard. 

In the area of isotope geology a 
continuing study of the North Amer- 
ican Precambrian shield is reported 
and isotope data are given for ultra- 
mafic rocks in limited areas of North 

Significant results have been ob- 
tained from the operation of a net of 
four seismic stations with high dy- 
namic range and wide frequency re- 
sponse. The net extends from New 
Guinea through Chile and Peru to 
Washington. The original purpose of 
the net has yet to be realized, but 
already definitive evidence for a ve- 
locity discontinuity in the mantle 
under South America has resulted 
from a study of deep-focus earth- 
quakes. This net will have been 
doubled by the end of August 1967. 


Cosmic-Ray Program 

S. E. Forbush 

Tht variation, with a period of two 
solar cycles, in the cosmic-ray diurnal 
anisotropy, 1937—1965, and the super- 
posed variations correlated tvith mag- 
netic activity. Annual means, 1937- 
19 65, of the cosmic-ray diurnal ani- 
sotropy component, in the asymptotic 
direction 128° E of the sun, contain 
a well determined wave with a period 
of 20 years, which is twice the solar 
cycle period of 10 years for the in- 
terval 1937-1 905. This variation is 
independent of magnetic activity, 
and, when removed, the residual var- 
iations combined vectorially with 

those in the asymptotic component 
38° E of the sun (in which there is 
no 20-year wave) give variations 
which are principally in the asymp- 
totic component 90° E of the sun. 
Yearly means of this resultant com- 
ponent are well correlated (r = 
+ 0.75) with magnetic activity U 
and, on the average, vanish for U = 
0. U is the absolute value of the 
southward geomagnetic component of 
the so-called equatorial ring current 
(Forbush 1 ) . The amplitude of the 20- 
year wave is 60% of the amplitude 
for the 1937-1965 average diurnal 
anisotropy 90° E of the sun. The 
wave passes through zero in the 
middle of 1958 near the time shown 
by Babcock 2 for the reversal of the 
sun's general magnetic field. 

In the Archimedean spiral streams 
from the sun, Ness and Wilcox 3 found 
a decided tendency for the magnetic 
field, in the plane of the ecliptic, to 
be oriented parallel or antiparallel to 
the theoretical direction (135°) pro- 
posed by Parker. 4 This is approxi- 
mately the direction of the cosmic- 
ray diurnal anisotropy component 
with the 20-year variation. Wilcox 
and Ness 5 found that, within sectors, 
the field directions were consistent 
with the outward or inward solar 
fields over the solar region from 
which the sectors originated. These 
facts may bear on the mechanism 
underlying the 20-year variation in 
the cosmic-ray diurnal anisotropy. 

These results are based on the 
pressure-corrected ion-chamber data 
gathered at Cheltenham-Fredericks- 
burg and at Huancayo for the period 
1937-1965, and at Christchurch for 
the period 1937-1961. To eliminate 
the so-called local temperature effect 
which remains after correcting to 
constant barometric pressure, the de- 
viation of each yearly mean vector 
from a 25-year mean, in the 24-hour 
LMT harmonic dial, is obtained for 
each station. These yearly deviation 


vectors for Huancayo and for Christ- fitted to N_i 6 , the yearly deviations 

church are normalized to those for from the 25-year mean, of the diurnal 

Cheltenham by a clockwise rotation anisotropy 128° E of the sun. The 

of 37° and 20°, respectively, and by amplitude of the 20-year wave is zero 

multiplying respective amplitudes by in P_ 10 , the yearly deviations from 

0.91 and 1.00. Components of these the 25-year mean of the diurnal ani- 

yearly deviation vectors at Chelten- sotropy 38° E of the sun. The devia- 

ham and of the normalized deviation tions (N- ie — W) of points from the 

vectors for Huancayo and for Christ- wave in Fig. 1 are correlated (r — 

church are taken on a right-handed +0.75) with U and with yearly 

set of axes, P e and N e , with the P means of P_ 1C (r = +0.65), which 

axis making an angle (9, in degrees, are also correlated with U (r = 

positive clockwise, with the observed 0.59). The components parallel, P 36 , 

25-year average vector for Chelten- and perpendicular, N 36 , to the asymp- 

ham, which has its maximum at 13.8 totic direction 90° E of the sun are 

hours local solar time in the 24-hour obtained from P_ 16 and (N- 1G — W). 

harmonic dial. The mean value of N 3G is zero, and 

An analysis of variance for the 25 the correlation between 2V 36 and U is 

values of P and N at the three sta- negligible. 

tions shows that the variance among The correlation (r = +0.75) be- 

years for P and for N is at least 20 tween P 36 and U is shown by the 

times the residual variance, definitely open circles in Fig. 2. Making use of 

indicating that the diurnal anisotropy these correlations, the values of iV_ 16 

was not constant from 1937 to 1961. in Fig. 1 are corrected to U = (no 

The residual variance corresponds to magnetic activity). These corrected 

a standard deviation of 0.017% for values of AT_ 16 are well fitted in Fig. 3 

the statistical uncertainty of single by the 20-year wave used in Fig. 1. 

yearly values of P and of N at each This fit indicates that the 20-year 

station. The corresponding standard variation is independent of magnetic 

deviation estimated from the varia- activity. Using the above correla- 

bility, within months, of daily mean tions, and assuming that the total 

24-hour waves at Cheltenham is diurnal anisotropy 90° E of the sun 

0.014%. It may be compared with vanishes for U = (i.e., no mag- 

the value 0.011% that would result netic activity), the resulting ampli- 

if the only uncertainties were those tude of the diurnal anisotropy 90° E 

due to the finite sampling or "count- of the sun, averaged during the 

ing rate" of the detector. This shows period 1937-1965, is 0.12%. 

that the so-called temperature effect As indicated above, the 1937-1965 

is effectively eliminated by using the mean of N 36 is zero and the correla- 

deviations of the yearly means of the tion between yearly means of N 36 and 

diurnal variation from a 25-year of U is not significant. Furthermore, 

average. It also shows the excellent the variance of annual means of N 3G 

agreement between the values of P is only 23% of that for P 36 . This was 

and N from the three stations. Thus the basis for assuming heretofore 

yearly values of P e and N e , averaged that the variability of P 36 arises from 

for all stations, are used hereafter; variations, from the 25-year average, 

for three stations (1937-1961) and of the component in the diurnal ani- 

for two (1961-1965) the respective sotropy 90° E of the sun. Similarly, 

standard deviations are 0.010% and the time of maximum of the 1937- 

0.012% for P e and for N e . 1965 average vector, say C 1S (ampli- 

Figure 1 shows the 20-year wave tude 0.12%), for the total diurnal 






Fig. I. Twenty-year wave fitted to deviations, N_u, of yearly means from the 1937—1961 
average of the diurnal anisotropy component in the asymptotic direction 128° E of the sun. 

anisotropy is taken as 18.0 hours 
local asymptotic time. For each sta- 
tion, the local solar time of maximum 
of the vector C 18 determines the addi- 
tive correction required for geomag- 
netic deflection if, in fact, the local 
asymptotic time of maximum for C 1S 
is 18.0 hours. 

Those corrections, in hours, are: 
for Godhavn (not otherwise used 
herein), 1.2; Cheltenham, 1.7; Christ- 
church, 2.7; and Huancayo, 4.1. The 
values are reasonably compatible with 
those obtained from calculations of 
^omagnetic deflections by Venkate- 
san and Dattner" and by McCracken 
et al. 7 This correlation indicates that 
the local asymptotic time of maxi- 
mum for CV-, is near 18.0 hours. 

For Huancayo the difference, in the 

24-hour harmonic dial on local solar 
time, between the observed 1937- 
1965 average diurnal variation vector 
and the vector (d 8 /0.91) gives the 
average vector for the so-called tem- 
perature effect. This finding is in fair 
agreement with preliminary values 
from a few yearly mean differences 
between the diurnal variation in the 
ion chamber and that from Simpson's 
IGY neutron monitor. 

These results comprise a self-con- 
sistent, statistically satisfactory ac- 
count of the variability of yearly 
means of the diurnal anisotropy and 
of its average for the period 1937- 

Monte Carlo experiments with 
filters. It has been indicated by some 
investigators that using digital com- 







P,.= -12.9+ 0.29 U n 
36 r= +0 .75 ° 

Legend: o Annual means 

• 3-Year means 
Except for 1937-38 




U inX 

Fig. 2. Correlation between magnetic activity, Uq, and the deviations P^6> of yearly means from 
the 1937—1961 average, of the diurnal anisotropy component in the asymptotic component 90° E 
of the sun. 

puters to "filter" cosmic-ray data 
would improve the signal-to-noise 
ratio for the diurnal variation. In 
collaboration with Dr. Martin Pome- 
rantz and Dr. S. P. Duggal of the 
Bartol Research Foundation, Monte 
Carlo experiments to test this sug- 
gestion were conducted on the Bartol 

Bihourly values were computed 
from a synthetic 24-hour variation 
having constant phase and amplitude. 
To each of these bihourly values, 
comprising about a "year," was added 
a random variate from a population 
with zero mean and known variance. 
With the computer these values 
were "filtered" by each of three dif- 
ferent filters designed to remove 
noise and to pass the synthetic diur- 
nal variation. For each "day" the 
24-hour wave was determined by har- 
monic analysis of bihourly values. 

Statistical analyses of the results (in- 
volving about one year of synthetic 
data for each filter) demonstrated 
that the reliability of the average 
wave was the same whether the data 
were filtered or not. Thus, unless the 
data contain persistent waves with 
frequencies that are not harmonics 
of the 24-hour wave, filtering does 
not improve the results obtained for 
the 24-hour wave. Since there is no 
evidence for such waves in cosmic- 
ray data, filtering results in no im- 

Observations and reductions of 
data. Cosmic-ray ionization chambers 
were operated throughout the report 
year at Huancayo and at Fredericks- 
burg, Virginia. Scalings and reduc- 
tion of records have been maintained 
current for both stations. 

Cooperation in operation of cosmic- 
ray meters. Grateful appreciation is 






1940 1950 1960 

Fig. 3. Twenty-year wave from Fig. 1 fitted to N_i6, corrected to Uq — 0. 

expressed to the U.S. Coast and Geo- 
detic Survey and the staff of its mag- 
netic observatory at Fredericksburg 
for efficient operation of the meters 
during the past report year, and to 
the Government of Peru and the 
Director and staff of the Instituto 
Geofisico del Peru for making cos- 
mic-ray records from Huancayo 
available. / 

Electrical Conductivity Studies 
of the Mantle under the Andes 

S. E. Forbush, M. Casaverde, R. Salgueiro, 
S. del Pozo, and L. T. Aldrich 

The report year has been largely 
a period of preparing equipment and 
recording sites for observations. 
Some data have been obtained on lo- 
cations distant from the Andean pla- 
teau, but the reduction of this infor- 
mation is still to be completed. The 
field season beginning July 1, 1967, 
is expected to provide data from the 
ten portable Askania variographs, 

three DTM versions of the vario- 
graph developed at Gottingen Uni- 
versity, and two or three la Cour in- 
struments with increased sensitivity 
in their D components. This equip- 
ment will be thinly deployed over an 
area of about 12° latitude by 12° 
longitude in Peru and Bolivia. It is 
hoped that the study may also be ex- 
tended into northern Chile this year. 
This array of stations, designed in 
consultation with Dr. U. Schmucker, 
should provide the broad outlines of 
the region of anomalously high elec- 
trical conductivity as a minimum. If 
a sufficient number of stations can be 
operated simultaneously, they will 
also provide an extensive survey of 
the equatorial ring current. 

The Askania instruments have 
been revised at the Department to 
have a more positive and accessible 
drive system, which will also provide 
as an integral part of the system a 
12-hour switch to give positive indi- 



cation of the D trace. The chronome- 
ters for all the instruments have been 
revised to provide hour marks and 
15-minute marks using low enough 
power to be operated by the main As- 
kania battery. This revision of the 
chronometers has included a new 
waterproof housing to maximize their 
usefulness at jungle stations. The 
DTM version of the Gottingen vario- 
graph has been designed to provide 
minute marks for easy digitizing 
(and reduced power requirements) 
as well as hour and 15-minute marks. 
Each timing system is independent, 
so the measurement of this crucial 
element is relatively certain. 


T. J. Smith, J. S. Steinhart, 

P. N. S. O'Brien, and D. E. James 

For the past several years, much of 
the effort at the Department in ex- 
plosion seismology has been directed 
toward the effective utilization of 
time-term analysis in the hope of 
producing realistic representations of 
crustal structure. Dissatisfaction 
with traditional plane-layered models 
as a means of describing the crust 
led us to apply time-term analysis to 
data from the Lake Superior experi- 
ment (see Year Books 62, 63, 64) . 
This method proved so successful in 
delineating crustal structure that an 
even more ambitious time-term anal- 
ysis was planned for the East Coast 
Onshore-Offshore Experiment 
(ECOOE). A description of the ex- 
periment and a map showing the lo- 
cations of shots and stations were 
given in Year Book 65. Reexamina- 
tion of the Lake Superior results al- 
ready reported (see Year Book 64) 
has continued. 

Time-Term Analysis — ECOOE 

D. E. James, T. J. Smith, and J. S. Steinhart 

This study was focused upon the 
problem of obtaining a three-dimen- 
sional representation of crustal struc- 

ture beneath the Middle Atlantic 
states and investigating the extent to 
which Bouguer gravity anomalies 
predicted by the seismic structure 
can be correlated in size and shape 
with observed gravity anomalies. 

Subcrustal velocity. Provision is 
made in the time-term analysis to 
compute subcrustal velocity. It is, 
however, impossible to determine ve- 
locities by the time-term method 
when the profile is completely or al- 
most completely unreversed. In the 
strict sense, only the ECOOE South- 
ern Line observations were single- 
sided; the Northern Line, while heav- 
ily unbalanced toward shots at sea, 
does have many observations of the 
Schuyler shots and hence is partially 
reversed. Thus, the two shot se- 
quences represent somewhat different 
problems and are treated as two dis- 
tinct groups of data. 

The subcrustal velocity for the 
Northern Line was found to be 8.15 
km/sec by the usual time-term 
method. Because there is no direct 
way of computing the Southern Line 
velocity by time-term analysis, we in- 
vestigated the behavior of time terms 
when the refractor velocity was con- 
strained to incorrect values. By this 
means we sought to develop a set of 
criteria by which we could determine 
the refractor velocity for a single- 
ended profile. A dipping-layer time- 
term model, originally scaled to Lake 
Superior P 1 structure, was con- 
structed and the time terms were cal- 
culated using different constrained 
velocities. The result is shown in Fig. 
4. Station 16 was used both as a shot 
point and as an observing site, and 
the shot was observed by sites 1 
through 15. As the velocity is con- 
strained to incorrect values, the time 
terms undergo a distortion which 
produces a discontinuity between the 
shots and the stations. 

The Southern Line time terms 
were computed for the range of con- 




Model 1S66 
* Shots 
o Stations 

25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 
f 500 

V n ■ 5.5 km/sec 
15.07 km 

7 6 5 4 3 2 1 400300200100 

7.95 km 


V,= 6.5 km/sec 

50 km 

Fig. 4. Constrained velocity set of time terms for Lake Superior model for velocities of 6.30, 

6.50, and 6.70 km/sec. 

strained velocities from 8.06 to 8.50 
km sec. The time terms of the four 
on-line stations nearest shore and a 
selected set of on-line shots were 
plotted in the same fashion as the 
model time terms. The Schuyler time 
term is taken to be 3.67 seconds on 
the basis of the Northern Line re- 
sults. The fact that the Schuyler time 
term is fixed is central to this tech- 
nique. The results for the entire set 
of constrained velocities are shown 
in Fig. 5. It is clear from the figure 
that a demand for time term conti- 
nuity from shot sequence to stations 
sharply limits the range of acceptable 
velocity values. 

On the assumption that the struc- 
ture is smoothly varying from sta- 
tions to shots (less than 30 km sep- 
arate the most easterly station from 

the most westerly shot), a velocity 
close to 8.25 km/sec appears to be 
most appropriate for the Southern 
Line data. Although this velocity 
value depends upon the normalizing 
value of the Schuyler time term, it 
varies only between about 8.22 and 
8.28 for possible Schuyler time terms 
as computed for the Northern Line. 
It is clear that a velocity of 8.15 km/ 
sec is incompatible with the condition 
of shot-station time-term continuity. 
The 0.10 km/sec difference in P n 
velocity between the Northern and 
Southern Lines is not particularly 
surprising. Herrin and Taggart 8 have 
shown that large changes in P n veloc- 
ity do occur over comparatively small 
areas, and there is no a priori reason 
for expecting the P n velocities for the 
Middle Atlantic states to be particu- 



Coast Line 
Fig. 5. Constrained velocity set of time terms for Southern Line in-line shots and stations. 

larly homogeneous. The P n velocities 
for the final time-term solutions are 
given in Table 1, together with the 
total number of observations and the 
rms residual for all observations. 

Structural distortion of time terms. 
Asymmetry of the recording network 
will clearly produce distortions in the 
time terms if the structure being in- 
vestigated is sufficiently complex. To 
examine the size and nature of these 
distortions, the single-layer geomet- 
rical model shown in Fig. 6 was con- 
structed to the scale of the East Coast 
experiment. Least-time ray paths 
were measured graphically to obtain 
a set of travel times corresponding 
to first-arrival refracted events, and 
these synthetic observations were 

then subjected to time-term analysis. 
The resulting model time terms are 
also shown in Fig. 6 along with the 
"true" time terms given by 

where H is the depth measured along 
a normal from the surface site to the 
interface. There are regions in which 
the normal is not unique, but the 
admissible choices yield results that 
are identical for purposes of this 
study. It will be seen from the figure 
that station 109, which models the 
Schuyler site, has been taken as a 
shot point as well as a receiving sta- 
tion and thus serves as a normaliza- 
tion point for the model results. 
This simple study indicates clearly 

TABLE 1 . Velocity, Number of Observations, and RMS Residual 
for Northern and Southern Line Observations 

Shot Sequence 


No. of Obser- 

RMS Residual, 

Northern Line 
Southern Line 

8.25 ±0.05 (est.) 





ECOOC P„ Model 01o6 

True time terms 

Computed time terms 
normalized to Schuyler) 

* bhots 



5 E 

V =6.50 km/sec 


V n = 8. 20 km/sec 

50 c 

-• — o- 

117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100 13579 

I 100km I (Schuyler) 2468, ° 



Fig. 6. Structural model and synthetic time-term results showing structurally induced time-term 

distortions for single-ended profiles. 

an approximate algorithm to compen- 
sate for structural distortion in cases 
where a similar geometry is appli- 
cable and calculated time terms are 
available. Each of the three segments 
a, b, and c of the time-term graph 
can be associated with a segment of 
the interface whose slope is deducible 
from the slope of the corresponding 
time-term segment. The depth of the 
interface a' is easily determined, 
since the associated time terms have 
been subjected to no structural dis- 
tortion. The depth of the b' segment 
can be calculated by requiring that 
the intersection of the time-term seg- 
ments a and b correspond to the in- 
tersection of the interface segments 
a/ and b'. The final segment c' is po- 
sitioned similarly under the criterion 
that the intersection of b and c be 

related by refraction paths to the in- 
tersection of b' and c'. The details of 
this procedure, though difficult to de- 
scribe, are straightforward and well 
defined, and in the model study they 
yield a corrected structure that is in 
satisfactory agreement with the orig- 
inal travel-time model. From the cor- 
rected structure so derived, it is now 
possible to deduce true time terms 
and determine the corrections to be 
applied to the calculated values. Ob- 
served departures from the idealized 
piecewise linear interface can be in- 
terpreted as local perturbations upon 
the approximating model. 

The true set of time terms in Fig. 
6 represents an idealized sampling of 
calculated ECOOE time terms over 
an approximate east-west section in 
the central part of the area. By the 



procedure described above, a set of 
true ECOOE time terms can be found 
from the calculated time terms. It is 
evident that there are many possible 
schemes for taking any one calculated 
time term into the set of true time 
terms; however, the unique peak in 
the time terms can be translated in 
only one way. For the particular set 
of calculated ECOOE time terms, the 
observed peak goes into the true peak 
by means of a translation vector 
whose components are a spatial offset 
of 72 km shotward and a time-term 
decrease of 0.46 second. All time 
terms associated with the "root" 
structure have been offset by this 
amount with the exception of the 
time terms near the eastern discon- 
tinuity, which have been offset by 40 
km but are otherwise unaltered. The 
structure in the northwestern part 
of the area was sufficiently different 
from that implied by the initial model 
that a second structure was modeled. 
Here the translation vector has a spa- 
tial offset of 65 km shotward and a 
time-term decrease of 0.52 second. 
The estimated time-term error due to 
structure is less than ±0.2 second. 
The modified time-term map from 
which crustal thicknesses have been 
derived is shown in Fig. 7. 

Mean crustal velocities. Because of 
the paucity of P t data for the East 
Coast experiment, it has not been 
possible to determine mean crustal 
velocities directly from the ECOOE 
data by the time-term method (see 
the section ''Late Arrivals and Am- 
plitudes" for a discussion of East 
Coast mean crustal velocities deter- 
mined from wide-angle reflections) ; 
however, a number of crustal veloci- 
ties determined by other investigators 
from profile analysis are shown in 
Fig. 7. Some of these values are from 
earlier Carnegie data. 9 ' 10 For the re- 
maining velocity values, we are in- 
debted to members of the U.S. Geo- 
logical Survey and to Brian Lewis of 

the University of Wisconsin. From 
these few mean crustal velocities, ve- 
locity contours have been drawn on 
the assumption that the lines of equal 
velocity approximately parallel the 
strike of the Appalachian structure. 
The data are adequate for the south- 
ern part of the area of study and do 
not indicate striking lateral changes 
in mean crustal velocity, save for the 
overall decrease that occurs between 
the crest of the Appalachian Moun- 
tains and the coastal area. For the 
northern and central parts of the 
area, there are virtually no data, and 
the assumption that the velocities are 
constant parallel to the structure is 
a major source of uncertainty. 

A number of isolated mean crustal 
velocities have been measured since 
the computation of the crustal thick- 
nesses presented in this report. In 
the central and northern parts of the 
area they indicate agreement with 
our values, but there is some sugges- 
tion that the crustal velocities of the 
Piedmont, Coastal Plain, and conti- 
nental shelf in the southern part of 
the area are lower than those used in 
the calculations of crustal thickness. 

The crustal structure that has been 
calculated from the modified time 
terms is shown in Fig. 8. Crustal 
thicknesses based on time terms are 
contoured by solid lines. The dashed 
contours are crustal thicknesses that 
have been inferred from other seis- 
mic work or are based upon time- 
term values calculated at the edges of 
the area studied. The dashed lines 
are included only because gravity 
studies discussed later in this report 
require closed contours. The mean 
crustal velocities used in the calcula- 
tions are contoured in Fig. 7. Because 
mean crustal velocities are not well 
known over the area studied, it is 
expected that significant deviations 
from the contoured velocity values 
may occur locally. Time-term varia- 
tions caused by local deviations in 





















































• ail CTUSta] velocity result in spuri- 
ous changes in crust al thickness. For 
areas, particularly along the 

at, this effect may be important. 
The structure contoured in Fig. 8 
a number of Interesting features. 
.1 thicknesses range from 30 
in one area on the Coastal Plain 
I Piedmont to about GO km at one 
nt under the Appalachian Moun- 
neral, crustal thicknesses 
th the Piedmont are not rapidly 
chai . An exception to this pat- 

vident in the area around 
>n, D.C. In this area, there 
many data, and crustal thick- 
are shown to change very 
idly over the Piedmont and Coastal 
Plains. The trend of comparatively 
large crustal thicknesses along the 
st appears firmly established. 
Time terms are few on the western 
k of the Appalachian Mountains, 
: the contours drawn in that area 

The most striking feature of the 
map is the prominent "root" struc- 
ture beneath the Appalachians. This 
• tends to lie almost directly be- 
neath the trend of highest elevation 
in the Appalachian Mountains, and 
appear.- to be broadest and deepest 
under the mountains in the southern 
1 of the region where the highest 
elevations occur (about 5000 feet). 
In the northern part of the area 
Med, the thickest part of the crust 
Is to lie slightly to the east of the 
of highest elevation. 
G >ity anomalies a nd seismic 
crui structure. Over the years, 
many attempts have been made to 
ncile seismic results with gravity 
M In general these attempts 
have not been very successful, partly 
■ it is difficult to make mean- 
ingful direct comparisons between 
mic profile models and gravity 
anomalio . 

By using gravity computations 
based upon the crustal structure de- 

rived from the time terms, we have 
attempted to ascertain the range of 
velocity-density relationships that 
yield an acceptable correlation be- 
tween gravity anomalies predicted 
from seismic structure and the ob- 
served anomalies. In the absence of 
detailed information about the veloc- 
ity structure of the crust, the choice 
of a velocity-depth relationship is 
somewhat arbitrary. There are, how- 
ever, a number of constraints im- 
posed upon the velocity-depth func- 
tion: (1) It must yield the mean 
crustal velocities from which depths 
were calculated. (2) It must be com- 
patible with the results from existing 
profiles. (3) The velocity for the up- 
per part of the crust (below the first 
2-3 km) should be near 6.10 km/sec, 
as this value has been found for much 
of the East Coast area, 9 ' 10 ' 15 " 17 al- 
though there remains some discrep- 
ancy between this value and the ve- 
locity values determined by shallow 
refraction work. Within the limits of 
the constraints, we have assumed as 
a model of velocity increase within 
the crust V = 6.10 + 0.017Z (see 
"Late Arrivals and Amplitudes"), 
where Z — depth in kilometers. 
A subcrustal velocity of 8.20 km/sec 
is assumed over the entire map area. 

Bouguer gravity values for various 
velocity-density relationships have 
been computed at 50-km grid inter- 
vals over the map area. The gravity 
values have been adjusted so that the 
computed anomaly is equal to the ob- 
served anomaly at Chesapeake Bay 
(grid point 1100, 500 in Fig. 9A), 
and this crustal section is taken to be 
the reference section for the area. 
Predicted gravity values in the region 
of the Coastal Plains have been cor- 
rected for the effect of the low-density 
sedimentary rocks. 

It remains only to determine the 
velocity-density relationship that best 
fits the East Coast area. Three 
straight-line velocity-density relation- 


ships cited in the literature were crease may be systematic in the lower 
used: Woollard's relationship, dp/dV part of the crust, or it may simply 
— 0.233; Birch's constant mean represent a discontinuity in the mean 
atomic weight relationship, d„/dV = atomic weight between the rocks of 
0.316; and Thompson and Talwani's the lower crust and those of the up- 
(1966) approximate gabbro-dunite per mantle. The second alternative is 
line, dp/dV = 0.194. All the relation- suggested particularly by the fact 
ships are of the form p = a + bV, that the line connecting values for 
where p = density, V = compres- gabbroic rocks to those of dunites 
sional velocity, and a and b are con- and peridotites cuts across lines of 
stants. Bouguer gravity anomaly mean atomic weight and has a slope 
values for these velocity-density re- dp/dV ~ 0.20. Thus, because the best 
lations have been computed at 50-km fit between computed and observed 
grid intervals over the area studied, gravity values is obtained from a 
The results establish that any dp/dV of about 0.19 to 0.23, one con- 
straight-line relationship with a value elusion might be that gabbroic rocks 
of dp/dV greater than about 0.23 or make up the lower part of the crust 
less than about 0.19 yields predicted and peridotite, the upper mantle. As 
gravity anomalies that are respec- the slope dp/dV between gabbroic or 
tively too large or too small relative less mafic rocks and eclogite is well 
to the observed anomalies. The grav- above 0.23, it appears unlikely that 
ity values obtained by using the the upper mantle consists of any sig- 
velocity-density relationship derived nificant fraction of eclogite. 
by Thompson and Talwani (1966) Discussion. From this study we 
are contoured in Fig. 9 (A) over most may summarize as follows: (1) A 
of the area of study; with the excep- difference of 0.1 km/sec in the meas- 
tion of the edges and local anomalies, ured P n velocity exists between the 
the discrepancy between computed Southern and Northern Lines. (2) A 
and observed gravity values is less three-dimensional model of crustal 
than 20 mgal. A simplified form of thickness has been obtained which 
part of the Bouguer gravity map of shows that crustal thicknesses vary 
the United States is displayed in Fig. from about 30 km in some areas over 
9(B). the Coastal Plains and continental 
The fact that the Woollard curve shelf to nearly 60 km in a few locali- 
has about as large a value of dp/dV ties along the crest of the Appalach- 
as can be tolerated raises interesting ian Mountains. (3) Three-dimen- 
questions and imposes some strict sional gravity computations based 
limitations upon the range of possible upon the seismic structure yield grav- 
materials that constitute the lower ity values that correlate well in size 
part of the crust. On the basis of and shape to the observed regional 
Birch's results, the low value of dp/ Bouguer anomaly for values of dp/dV 
dV suggests that the mean atomic between 0.20 and 0.23. (4) The fact 
weight must decrease in the region of that 0.23 is about as large a value of 
the lower crust and upper mantle. dp/dV as can be accommodated by the 
Recent work by Knopoff 18 indicates data, strongly suggests that the mean 
that the slope Birch derived for lines atomic weight of subcrustal rocks is 
of constant mean atomic weight is lower than that of rocks in the lower 
too low; if this proves to be correct, part of the crust. A mafic (possibly 
the decrease in mean atomic weight amphibolite) lower crust with an up- 
must be even greater than that im- per mantle composed of peridotite or 
plied from Birch's curves. The de- dunite is one model consistent with 
























-£ O 


<L> . 


Jr> Q- 






D C 

o o 


M- > 

o 5 





£■ o 

*> o 

o fc 


>- 00 


0) • 
3 3 


05 ^ 
3 ^ 

o -Q 

* -o 

M- * 

° a 

c E 

o o 

•s u 





0) t/i 

> <u 

■o S 

<D CO 


a <d 


co 3 





cQ .T 

. .«+- 

o o 

. a 

U) D 

LE £ 



this conclusion: models in which ec- 
logite is the rock type of the upper 
mantle are not. 

Whether the measured differences 
in P., velocity for this area are the 
result of genuine velocity inhomoge- 
neity in the mantle or simply the con- 
sequence oi different average ranges 
of observations is a question difficult 
to resolve. The data indicate, how- 
ever, that, even if the apparent P n 
velocity increases with distance, the 
increase must be very small. More- 
over, there is some suggestion from 
previous profile studies that real lat- 
eral subcrustal velocity variations do 
exist in the area between the South- 
ern and the Northern Lines. Work of 
Tatel et al. 9 in the area of the Ches- 
apeake Bay and the Patuxent River 
indicated a P n velocity near 8.1 km/ 
sec. whereas preliminary results from 
the North Carolina experiment of 
1962 suggested a P n velocity near 8.3 
km sec. :9 Thus the data appear to 
favor lateral velocity inhomogeneity 
in the upper mantle, the velocities 
being near 8.15 km/sec for the 
Northern Line and near 8.25 km/sec 
for the Southern Line. 

Wide-Angle Reflections and 
Amplitudes from ECOOE 

P. A 7 . S. O'Brien 

The records were divided into two 
groups, those from the Northern Line 
of shots and those from the Southern 
Line. For each group the times t of 
the late arrivals were squared and 
plotted against the square of the ob- 
serving distance x. 

For the Southern group the plotted 
points fell on two parallel straight 
lines whose intercepts on the t 2 axis 
were in the ratio of 4/1. This pro- 
vides almost incontrovertible evi- 
dence that the two sets of points 
relate to a primary and a first-order 
multiple from a single plane reflec- 
tor. The estimated value of the mean 
overburden velocity is 6.30 km/sec, 

and this should be quite reliable un- 
less there are extreme, systematic 
variations in the near-surface veloci- 
ties. The mean reflector depth was 
31.8 km. Because the primary reflec- 
tions were observed out to a range 
beyond 400 km, it is possible to place 
an upper limit to the mean velocity 
gradient in the crust, for the ray- 
leaving the shot point must have been 
reflected at the point 200 km away 
from the shot and 32 km beneath it. 
If the velocity gradient had been 
more than 0.01 sec -1 , no ray leaving 
the shot point would have reached 
that reflection point. 

The few records from the North- 
ern Line examined showed mainly 
multiple reflections — the observing 
ranges were greater than for the 
Southern Line — which gave a mean 
overburden velocity of 6.50 km/sec 
and a mean depth of 34 km. The few 
primary reflections observed were 
fully consistent with these values. 

The measured amplitudes were 
highly variable. However, it was quite 
clear that P n could not be a simple 
head wave, as the recorded amplitude 
was 30 times that calculated for a 
range of 200 km and 300 times as 
great as that calculated for the ex- 
treme range of 700 km. The simplest 
way to explain these large amplitudes 
is to involve a velocity gradient in 
the uppermost part of the mantle. 
A gradient of 5 — 10 X 10^ 3 sec -1 
would be sufficient. 

The amplitude of P n decreased by 
a factor of 2 when the distance was 
increased from 200 km to 700 km. 
If this attenuation was due entirely 
to absorption, the Q of the upper 
mantle would be about 1100. If Q has 
a lower value than this, the velocity 
gradient beneath the Mohorovicic 
discontinuity must increase with 

The velocity gradients quoted are 
inconsistent with those assumed in 
the time-term analysis, but no at- 


tempt has yet been made to remove of refracted rays so that the time 
the inconsistency. terms are independent of azimuth 

and refer to a point on the refractor 
Late Arrivals and Amplitudes almost directly beneath the station. 

In the structure obtained by Smith 
et al. 21 the maximum variation of the 

P. N. S. O'Brien 

Selected records from the Lake time term with azimuth was about 
Superior Experiment of 1963 and the 1.0 second (12.5 km) and the maxi- 
East Coast Onshore-Offshore Experi- mum offset of the refractor point 
ment of 1965 were reexamined, pri- from the station was about 80 km. 
marily to identify reflections and to It therefore seemed desirable to re- 
measure P-wave amplitudes. The analyze the first arrival times with- 
identification of reflections is impor- out the assumption that each station 
tant because analysis of their time- had a unique time term. A more im- 
distance curves enables an estimate portant reason for the reanalysis was 
to be made of the mean velocity over- that the presence of structure en- 
lying the reflector. If the reflector is abled the mean overburden velocity 
also a refractor giving first arrivals, to be estimated from the first arrival 
it is possible to convert the first ar- times themselves. A direct check 
rival time terms into depths without could therefore be obtained on the 
having to use qualitative arguments very high velocity used in the con- 
to place limits on the overburden ve- version of time terms to refractor 
locity. Measurement of P-wave am- depths. 

plitudes is important because ex- The method for obtaining over- 

pected values may be calculated for burden velocity may be understood 

the crustal models proposed from in- by reference to Fig. 10. The refrac- 

terpretation of the time-distance tor structure is reflected in the travel 

data; and if they are not consistent times at stations offset toward the 

with those measured, the model is shot point by an amount h tan Q c , 

obviously unsatisfactory. where h is the refractor depth and 6 C 

Neither experiment was designed is the critical angle. By matching up 

for either of the above purposes; the time-distance graphs obtained 

nevertheless, the results of this re- from shooting in opposite directions, 

examination have been informative, an estimate may be made of 2h tan 6 C , 

Some work was also done to finish and this, together with the estimated 

the analysis and interpretation of a delay time (h cos C /V, where V is the 

few seismic records obtained in the mean overburden velocity) , enables h 

Italian Alps. 20 and V to be determined. 

The Lake Superior Experiment of It was possible to match up three 
1963. The continuing work on these separate structures on the upper re- 
records falls into three main parts: fractor; they gave estimates of the 
reanalysis of the first arrival time- mean overburden velocity of 5.13, 
distance data; analysis of the late 5.44, and 5.50 km/sec. If it is as- 
arrival time-distance data ; and anal- sumed that the overburden is not 
ysis of the amplitudes. homogeneous but consists of a layer 

Reanalysis of the first arrival of 4.0 km/sec (sediments) overlying 

time-distance data. These data have one of 6.1 km/sec (volcanics), the 

already been exhaustively analyzed mean vertical velocities are reduced 

by the time-term method. 21 ' 22 But this to 4.81, 5.03, and 5.28 km/sec. There 

method assumes that the refractor was only one well-defined structure in 

has a plane boundary within the cone the arrivals from the Mohorovicic 



■2h, tan 9 


r* — 2 h, tan? c — H 


Ground surface 


Fig. 10. Use of offset in obtaining mean overburden velocity. 

discontinuity. Matching it up indi- 
cated that the mean velocity beneath 
the upper refractor was about 6.68 
km sec. 

The reinterpretation of the first 
arrivals has not yet been completed 
in detail, but the final structure 
should not depart much from that 
shown in Fig. 11. It should perhaps 
be emphasized that the reinterpreta- 
tion was based on exactly the same 
time-distance information as the 
original time-term analysis and that 
any differences relate simply to the 
added constraints of the present 
method, namely, that delay times 
should refer to positions on the re- 
fractor and not to positions on the 
ground surface, and that refractor 
velocities should be obtained only 
from reversed shooting across the 
same portion of the refractor. The 
second constraint accounts for the 
fact that the upper-mantle velocity is 
estimated to be 8.21 and not 8.07 
km /sec as found by the method of 
time terms. 

Analysis of the late arrival time- 

distance data. The records were so 
dominated by reverberation in the 
lake that it was not often possible to 
pick a clear onset to any late arrival. 
However, nearly all records taken at 
distances less than about 450 km 
showed one, and sometimes two, late 
arrivals whose amplitude was usually 
5 to 10 times that of the first few 
cycles of the first arrival. 

The only records from the west end 
of the lake that were examined were 
those from Victor, Yankee, X-ray, 
and Tyro. All of them exhibited a 
late arrival which was clearly a re- 
flection from the Mohorovicic dis- 
continuity and whose time-distance 
curves indicate a mean crustal veloc- 
ity of 6.75 km/sec. The records from 
X-ray and Tyro also showed an ear- 
lier late arrival which probably re- 
fers to an intracrustal discontinuity. 
Its time-distance curves indicate 
mean velocities of about 6.0 km/sec 
for X-ray and 6.3 km/sec for Tyro 
and mean depths of about 12 km and 
23 km, respectively. However, allow- 
ance for the considerable thickness 




Range in km from Otter 

o g -2 
o .5 o 



t — r 




t — r 


35 5.28 km/sec 45 48 28 5.03 km/sec 2221 56 59 18 16 4.31 km/sec 10 




f 30 


?0- —- -0? 

6.71 km/sec 






8.21 km/sec \ 

Southern line 

Ray path L— *-R 

Ray path R— »~L 

Main line 

Time-term analysis (Smith et a!., 1966) 

Fig. 1 1 . Lake Superior crustal structure from first arrivals. 

of sediments would reduce the esti- 
mated depths and slightly increase 
the estimated velocity of the material 
immediately above the reflector. 

So far, Woodsman and Otter Cove 
are the only stations at the eastern 
end of the lake whose records have 
been examined. They show a reason- 
ably late arrival which is most simply 
interpreted as coming from a reflec- 
tor whose depth increases from about 
15 km near the Keweenaw Peninsula 
to about 25 km at a position two 
thirds of the way from the tip of the 
Peninsula to Otter Cove. This inter- 
pretation is still highly tentative and 
needs to be checked against records 
from other stations. The records also 
show a very few later arrivals that 
are consistent with an identification 
as reflections from a depth of about 
53 km. 

Analysis of the amplitudes. The 
amplitudes of the late arrivals were 
broadly consistent with their inter- 

pretation as reflections, but the first 
arrivals from the Mohorovicic dis- 
continuity were too large to be due 
to head waves from a refractor with 
a constant velocity. Velocity gra- 
dients of about 5 X 10 -3 sec -1 would 
be sufficient to explain the observa- 

Bancroft 23 has presented P-wave 
spectral amplitudes as a function of 
range and has clearly shown that 
maximum amplitudes were obtained 
for shots fired at depths in the range 
130-159 meters. O'Brien 24 has used 
the spectral amplitudes to estimate a 
Q value of 830 ± 30 for the upper 
crust and has shown that the opti- 
mum shooting depth corresponds to 
the depth at which the bubble pulse 
"frequency' ' becomes equal to the res- 
onant frequency of the water layer. 
This depth, for Nitramon WW-EL 
fired on an acoustically hard bottom, 
is equal to 143 W - 18 meters, where W 
is the charge weight in tons. 


EARTHQUAKE SEISMOLOGY determine the absorption of compres- 

. a; ; 5. ,/.. /.. Fernandez, sional wave energy in the upper half 

H. Marconi, A. Rodriguez, f the mantle. To this end, the fre- 

°> amI F - quency spectrum at the Department, 


at a distance of 50°, and that at 

O for P W WES IN the Mantle Cuzc0 > p f u > at a stance of 8°, were 

I. S>. Sacks The amplitude ratio of two fre- 

The large For avian earthquake of quencies A and f 2 at a seismograph 
r 17. 1966, has been studied to A is 

mM .4(A) S(f 2 ) R(f 1 ) -.(/,-Aln 
M, = _ • • exp 

A(f,) S(f,) R(f 2 ) QiVt 

where S(f) is the source function, The frequency independent param- 

R(/) is the receiver function includ- eters such as the geometrical effect, 

ing seismograph and crustal transfer spherical spreading, and focusing, 

functions, /\ is the distance from cancel out in this equation ; i.e., abso- 

earthquake to seismograph, and Vi is lute amplitudes are not considered, 

the effective velocity along the path. If we now take the ratio of M for 

The ratio /• V, equals t lt the travel two stations 
time to station 1. 

S(A) R 1 (f 1 ) /-tt(A-A)A 


3A S(A) RAf.) \ ft / _ g 1 (/ 1 )Jg,(/,) 

M 2 5(A) Ro(h) I -*(f*-f i)U \ Ri(f*)R*(fi) 

— . — • exp 

S(f- 2 ) R 2 (f 2 ) 

/-7T (A-/x) A 
exp [ , 

\ Qi ) \ Q 2 / (1) 

If identical seismographs are used at 14-second window was used. The 

the two stations, the R(f) factors broadband tape channel was played 

are mainly the crustal transfer func- through Kronheit band-pass filters, 

tions. Fernandez has calculated and To get detail in areas where the spec- 

measured transfer functions for var- trum is very steep, e.g., at 1.3 cps, 

ions crusts. The amplitude of R is three filters were used in series. The 

roughly proportional to the imped- mean amplitude in the time window 

a nee contrast (p ] V 1 /p 2 V 2 ) at the was determined by means of a pla- 

Moho discontinuity and is usually nimeter. The filters were considerably 

about 0.8. Therefore, R,R 2 probably broader than the equivalent filters 

varies between 0.6 and 1.5 over the that would have been used with con- 

frequency range. The above range of ventional Fourier analysis over the 

R I: is based on a time window of same time window. The spectrum ob- 

in finite extent. As the time window is tained is therefore somewhat 

reduced, the amplitude of R de- smoothed, showing the gross struc- 

creases. Kanamori 28 has shown that ture but not the microstructure. 

use of a relatively short time window, The two spectra are shown in Fig. 

about 10 sec, reduces the amplitude 12. The energy arriving at Cuzco has 

of R considerably. In the analysis of components above the ground noise 

the Peruvian earthquake example a up to 11 cps, but the signal at the 






Frequency (cps) 

Fig. 12. Spectra of the compressional wave arrivals from the Peruvian earthquake of Octo- 
ber 17, 1966, recorded at DTM in Washington, D.C. (upper) and at Cuzco, Peru (lower). Only 
the slopes of the curves are significant, not their positions. The absolute amplitude at Cuzco is in 
fact considerably larger than at Washington. 

Department dropped below the noise 
at 3.5 cps, owing partly to the greater 
attenuation of the longer path and 
partly to local noise at this station. 
The dominant feature in the DTM 
spectrum is the steep decline beyond 
1.1 cps and the flattening out beyond 
1.5. There is some indication of a 
similar feature but of smaller ampli- 
tude on the Cuzco spectrum, which 
suggests that source radiation might 
account for some of the effect. Equa- 
tion 1 cannot be solved unless some 
estimate is made of Q lt the Q on the 
path to Cuzco. The simplest model 
would be a homogeneous model in 
which Qi = Q 2 . In this case the 
mantle Q = 1750. It is suggested 
from the work of Sumner 26 that the 
Q p for the upper couple of hundred km 
may be nearer 1000, in which case the 
mantle Q p — 1500. These figures are 
based on the mean slope of the spec- 
tral data. Confidence limits on this 
determination would be misleading at 
this stage because the two seismo- 
graphs and the earthquake were not 
along the same great-circle path and 
we have therefore made the assump- 
tion that the ratio [S(/i)]/[S(/ 2 )] is 
independent of azimuth. 

Shear Wave Absorption in the 
Upper 130 km 

sSKP-SKP. A technique is de- 
scribed that enables the Qs structure 
of a region to be determined by means 
of the spectral data from a single 
well-situated vertical component seis- 
mograph. The factor by which the 
shear- wave amplitude decreases at 
frequency / is exp (-irft/Qs) where t 
is the travel time to the observation 
point and Qs is a dimensionless, fre- 
quency-independent number. The ray 
paths of the phases used are shown 
in Fig. 13. The spectrum of the sSKP 
wave train is divided by that of the 
SKP arrival. The ray paths are very 
similar for all the path below the 
earthquake, and any path effects or 
receiving station effects therefore 
cancel out. The spectrum of sSKP is 
modified by three effects that do not 
affect the SKP spectrum: 

1. Possible differences in the radi- 
ation of the earthquake in the SKP 
and sSKP directions. The directions 
are nearly antipodal, and the radia- 
tion is therefore expected to be simi- 

2. Filtering effect of the earth's 



Fig. 13. Ray paths of sSKP and SKP. The two paths below the earthquake are very similar. 

crust at the point of reflection. This 
effect is small, provided that the fre- 
quency window is adequate. The 
transfer function (amplitude versus 
frequency) of a single-layer crust, 
considering the interference between 
reflections and multiple reflections 
from the Moho and the surface, has 
spectral peaks that repeat every V/2h 
cps where the V is the shear velocity 
in the crust and h is the crustal 
thickness. The effect, which can be 
seen in Fig. 14 at the lower frequency 
end of the spectrum, is to superim- 
e an oscillation on the Q effect. 
3. The effect of anelasticity, i.e., 
the greater absorption of high fre- 
quency energy due to the finite Q of 
the region between the earthquake 
and the surface. The ratio of the 
spectral amplitudes over a frequency 
range A/ of the two phases is R = 

exp (-irAft/Q) , where t is the time 
the sSKP phase spends in the region 
above the earthquake. If the fre- 
quency range A/ is large enough, 
the absorption will dominate the first 
two effects. Plate 1 shows the verti- 
cal-component short-period and long- 
period records from the Port Moresby 
standard seismograph for an earth- 
quake at a depth of 130 km in 
South America. The spectral ratio of 
the SKP and sSKP phase is shown 
in Fig. 14. It is necessary in any 
analysis of this type to be sure that 
all the frequency components plotted 
are in fact above both ground noise 
and digitizing noise. An equivalent 
time section of the record just before 
the sSKP arrival (which is the 
smaller one) was therefore analyzed 
to determine the background noise. 
All the points appearing in Fig. 14 



Q s for upper 130 km 

0.8 0.9 1.0 1.1 1.2 


Fig. 14. The spectral ratio of the arrivals shown in Fig. 15. The frequency spectrum of sSKP 
is divided by the frequency spectrum of SKP. Points are plotted only where the signal-to-noise 
ratio at that frequency is greater than 2/1. The open circle at f = 0.45 cps has a S/N of 1.5. 
The square X, and closed circles at f > 0.8 cps refer to the time window used — the square, 7.5 
sec; the X, 1 sec; and the closed circle, 1 5 sec. 

are those with a signal-to-noise ratio 
greater than 2/1. The noise in this 
case is mainly the coda of the SKP 
arrival, as can be seen in the short 
period (SP) record of Plate 1. A 
rather poor point, S/N ^ 1.5/1, was 
all that could be determined in the 
frequency range 0.2 - 0.8 cps. The 
time window used was 44 sec on the 
LP record. Three time windows were 
tried on the SP record — 7.5, 10, and 
15 sec. The resultant spectra were 
sensibly the same for the three win- 
dows, suggesting that interference 
effects from reflections were small. 
The Q value satisfying the data is 
300. The uncertainty in this value 
may be seen from the lines represent- 
ing different Q values drawn in Fig. 
14. All points lie between 200 and 
400. The value is considerably higher 
than that reported by Anderson et 
al. 27 from surface waves, which was 
110 for the same depth. It does, how- 

ever, apply to a different region from 
that considered in their report. 

A Path for High-Frequency Shear 
(S) Waves in the Upper Mantle 

The Qs for the upper 600 km of 
the earth was determined to be 160 
from a study of shear waves reflected 
and refracted at the core as reported 
in Year Book 62. It was suggested 
that not all of this upper region had 
this low Q, but that a relatively thin 
zone, possibly of the order of 100- 
200 km thick, had a Q value well be- 
low 100, while the remainder of the 
region had a much higher Q. It 
seemed plausible that the "earth- 
quake free" region in South America 
between depths of 350 and 500 km 
might be associated with a region of 
low rigidity, owing possibly to par- 
tial melting, and that this region 
would have a low Q for S waves. 
A number of broadband, wide-range 



seismographs were installed to study 
the Q structure in greater detail. 

As was reported in Year Book 65, 
the spectrum of the direct S wave 
recorded from a deep earthquake 
showed severe attenuation of the 
higher frequencies, as would be ex- 
pected from a path with a Qs of 160. 
Later in the 5 train, however, higher 
frequencies do appear. Recordings of 
a deep earthquake made with three 
different frequency passbands are 
shown in Plate 2. The compressional 
arrival maintains its envelope at all 
frequencies; i.e., the duration of the 
arrival at the lower frequencies is 
similar to that at the higher fre- 
quencies, as would be expected from 
a simple source and a simple path to 
the seismograph. The low-frequency 
shear arrival is also very simple and 
of similar duration to the P. At the 
high frequency, there is no trace of 
this clear S wavelet at the expected 
arrival time, but 15 seconds later 
some high-frequency activity starts 
and persists without significant am- 
plitude variation long after the low- 
frequency trace amplitude has be- 
come very small. The low-frequency 
S amplitude drops by more than a 
factor of 3 after 10 seconds, whereas 
the high frequency shows no diminu- 

tion for well over a minute. The 
following tentative conclusions are 

There is a path for high-frequency 
& through an otherwise highly ab- 
sorptive region. Oliver, Sykes, and 
Isacks of the Lamont Observatory 
recently reported (Trans. Am. Geo- 
phys. Union, March 1967) such a 
path from the deep earthquakes in 
the Fiji region to their station NIU. 
If such a path exists in South Amer- 
ica it presumably reaches the higher 
Q region at some distance from To- 
conce. Assuming that it travels then 
to the station at a velocity of 4-5 
km/sec, the S (direct) — S (high- 
frequency) time of 15 seconds is 
equivalent to a distance of 60-75 km 
from the station. The long duration 
of the high-frequency late-arriving S, 
however, is unexplained. 

Converted Waves from the Inter- 
face at a Depth of 410 km 

A rather clear arrival has been 
recognized between the P and S ar- 
rivals from South American deep 
earthquakes. It has been identified as 
a shear-to-compressional wave con- 
version at a depth of about 410 km. 
Figure 15 shows the ray paths. The 


,nferf °ce at + An 

Earthquake depth > 400 km 
Fig. 1 5. Ray paths of direct P, direct S, and converted S-to-P arrivals. 



identification was made as follows: 
Waveform comparison. The direct 
shear waves from a deep source have 
the higher frequencies severely at- 
tenuated by the low-Q region through 
which they travel. In Year Book 62 
we reported a mean Qs value of 160 
for the upper 600 km. The waveforms 
could be meaningfully compared only 
at periods longer than 2 sec. Plate 
2, a and &a is a radial component 
seismogram which shows P, the new 
phase, and S clearly. The peaks and 
troughs of the new phase are marked 
A through D in Plate 2, a and 2, aa 
and the correlated peaks on the S 
phase are similarly marked (2, a). 
Plate 3 is another example of the 
phase, again from a radial component 
record. The S waveform has been 
determined both from direct S and 
ScS as recorded on the Cuzco broad- 
band large dynamic range seismo- 
graph, and found to be similar to the 
converted phase wavelet shown in 
Plate 3. The similarity to the S 
wavelet suggests that the phase was 
radiated as S. Since it arrives well 
before the direct S, it must have 
traveled a considerable portion of 
its path as the higher-velocity P. 

It arrives at the seismograph as P, 
appearing only on the vertical and 
radial components but not on the 
transverse. The position of the inter- 
face relative to the hypocenters was 
determined by studying earthquakes 
occurring at varying depths. The time 
difference between P and the SP 
phase decreases as the earthquake 
depth decreases, indicating that the 
interface is above the earthquakes 
studied, which ranged between 526 
and 649 km. Figure 16 shows theo- 
retical travel-time differences be- 
tween the SP phase and P for various 
depths over the distance range for 
which the phase exists. The travel 
times have been constructed by means 
of the Jeffreys-Bullen tables. It is 
known that the difference from the 
Jeffreys-Bullen travel times is quite 
large for shallow earthquakes in this 
general region, but the agreement is 
surprisingly good for deep earth- 
quakes. Figure 17 shows these resid- 
uals for two deep earthquakes. The 
earthquakes were located with tele- 
seismic data, i.e., few of the stations 
whose residuals are shown in the 
figure were used to locate the earth- 
quakes. Since the theoretical SP 

Distance, degrees 

Fig. 16. Theoretical travel-time differences between the direct compressional arrival and the 
S-P converted wave. Curves are shown for three earthquake depths. The travel times for other 
depths can be interpolated using the differences which are marked between the curves. 





— 1 




• • 

° o 


May 13 d 02 h 23 m 
(19.3s, 63. 8w) 


ii i a d 

Co $ ° 

H = 589, M = 

5.1 ' 




'©" T o _ 



July 30 d 02 h 11" 
(22.8s , 63. 7w) 
H = 526,M = 4.5 




Distance in degrees 

Fig. 17. Travel-time residuals of two deep earthquakes used in the converted S-P wave study. 
Residuals are shown from the Jeffreys-Bullen travel times. Dots are readings from high-elevation 
seismographs which generally have high positive residuals. The earthquakes were located by 
stations mainly at teleseismic distances. The residuals shown should therefore be independent of 
the hypocenter location procedure. 

travel times were calculated from the 
Jeffreys-Bullen tables for depths 
greater than 400 km only, they may 
be moderately reliable. Table 2 shows 
the residuals of the observed SP 
phase from the theoretical. The 
agreement is satisfactory. 

Figure 18 shows the relative am- 
plitude of this converted phase as a 
function of angle incidence. Table 3 
gives the angle of incidence for vari- 
ous epicentral distances and depths. 
The amplitude reaches a maximum 
just before the phase disappears. In 
the record at a distance of 12°, which 
has an angle of incidence at the inter- 

face of 34°, and therefore near max- 
imum amplitude, the SP phase is 
considerably larger than direct P. At 
a distance of 14°, the phase has dis- 
appeared. The angle of incidence is 
greater than 34° and, as can be seen 
from Fig. 21, the converted phase 
does not exist. 

As yet, there are not sufficient data 
to determine the velocity contrast at 
the interface from the amplitude be- 
havior of the SP phase. Preliminary 
indications are that it must be of the 
same order as the Moho discontinu- 
ity, i.e., a velocity contrast of the 
order of 10%. Frequencies as high as 

TABLE 2. Residuals from Theoretical SP-P Travel Times, Based on an 

Interface Depth of 413 km 

















W. Brazil 












S. Bolivia 






S. Bolivia 








15 20 25 

Angle of incidence, degrees 




Fig. 18. The relative amplitude of incident shear (S) to refracted compressional (P) waves 
(after McCamy, Meyer, and Smith, 1962). 28 The different curves are for different velocity con- 
trasts. Curve 13 has a contrast of 17%; curve 15, of 11%; and curve 17, of 6%. Preliminary 
data favor curve 15. 

TABLE 3. Angle of Incidence at Interface for Different Depths and 

Epicentral Distances* 

Angle of Incidence, 

Epicentral Distance, degrees 

Depth 667 


540 km 





The depths are the same as used in the theoretical travel times in Fig. 16. 

2 cps in the SP phase (see Plate 2, c) 
suggest a very sharp interface with a 
transition zone thickness of a few- 
kilometers only. 

A Search for Small, Deep 
Earthquakes in the Andes 

Shigeji Suyehiro 

To identify small, deep earthquakes 
in South America a quadripartite net 

about 1 km in size was set up in 
Toconce, Chile, and one in Cuzco, 
Peru. During the period of observa- 
tion the U.S. Coast and Geodetic Sur- 
vey identified two fairly large, deep 
earthquakes in northern Argentina 
within about a month by teleseismic 
observations, but neither foreshocks 
nor aftershocks associated with these 
earthquakes were observed by the 



quadripartite net. The lack of small, 
deep earthquakes during the period 
of observation may indicate that 
there are comparatively few small, 
deep earthquakes in comparison with 
shallow earthquakes. 

In the course of the quadripartite 
observation in Toconce, systematic 
azimuthal distortions of apparent ve- 
locity and direction of propagation 
were noted. 

The apparent velocity is plotted 
against the direction of propagation, 
both having been reduced to the level 
plane. Frequently, earthquake waves 
traveling from south to north give 
much lower velocities than those 
traveling in the opposite direction. 
The distortion cannot be attributed 
to an actual distribution of earth- 
quake hypocenters, such as deep ones 
to the north and shallow ones to the 
south, which would give the observed 
result without assuming any local 
anomaly. Nor can it be attributed to 
the difference in weathering at quad- 
ripartite points because all four seis- 
mometers were installed on the same 
solid bedrock. Actually the deep 
earthquakes in northern Argentina, 
which should give an apparent veloc- 
ity of 12-14 km/sec, gave only 7-9 
km/sec. Shallow earthquakes to the 
north, which should give low veloci- 
ties of the order of 8 km/sec, gave, 

without exception, the much higher 
apparent velocity of more than 20 
km/sec. In Fig. 19, therefore, the 
ratio of observed to calculated appar- 
ent velocity is plotted for earthquakes 
of more than 300 km in epicentral 
distance located by the U.S. Coast 
and Geodetic Survey or by G. Saa, 
S. J., using the Carnegie local net. The 
calculation is based on the Wadati- 
Oki travel times, and nearer earth- 
quakes of less than 300 km in epi- 
central distance were not plotted, for 
a small shift in epicenter will greatly 
change the calculated apparent ve- 
locity. Figure 19 indicates that the 
observed distortion is caused by some 
local structure beneath the array 
rather than by any actual distribu- 
tion of hypocenters. Figure 20 shows 
the distortion observed in the direc- 
tion of propagation. 

The simplest way of explaining 
such azimuthal distortion is to as- 
sume an interface of 30° in dipping 
angle where velocities in the surface 
are 4.5 km/sec, and those of the 
second layers, 6.5 km/sec. 

This dipping interface of the order 
of 30° must be local. To interpret the 
crustal structure under a certain 
area, a much larger array is required. 
Small arrays must be used for study- 
ing local seismicity rather than local 










° n 

Q_ U ' 

£- 0. 

























■ a 


-180 -140 -100 -60 -20 20 60 
W N 

Azimuth (degrees) 

100 140 180 

Fig. 19. Ratio of observed to calculated apparent velocity plotted against azimuth in To- 
conce. Calculation was made from epicentral distance and depth, based on Wadati-Oki ve- 
locity distribution, for reported earthquakes, epicentral distance of which is more than 300 km. 




































.— to 

■o 5 









— <■ 








• N 








160° -120° -80° 

■40° 0° 40 c 



80° 120° 160° 

Fig. 20. Deflection of observed from calculated direction of propagation plotted against 
calculated direction of propagation in Toconce for located earthquakes of A > 300 km by U.S. 
Coast and Geodetic Survey or by Carnegie net. 

On the Spatial Distribution of 

Earthquakes Near San Juan, 

F. Volponi and H. Marconi 
The problem of the spatial distri- 
bution of seismic activity has been 
discussed in numerous papers, many 
of which had another purpose of 
study. The following report examines 
some experimental data on this sub- 
ject obtained in a restricted region 
on the eastern flank of the Andean 
cordillera that divides Chile and Ar- 
gentina. This region lies between 31° 
and 32° south latitude and 68° and 
69°30' west longitude (Fig. 21). Two 
large earthquakes have occurred here 
— one, of magnitude 7.8 on the Rich- 
ter scale, on January 15, 1944, and 
another, of magnitude 7, on June 10, 
1952. The present activity of the area 
may be judged by the fact of the 
occurrence of 12 perceptible events 
in the first 6 months of 1966. 

The equipment for the study con- 
sisted of single, vertical, Wilson- 

Lamison type seismometers with elec- 
tronic amplification and visible re- 
cording. Amplification was 10 5 -10 6 at 
10 cps. Timing was obtained from 
crystal chronometers, and second 
marks were registered on the trace, 
making it possible to measure abso- 
lute time with a precision of 0.1 sec. 
The stations were located (as indi- 
cated in Fig. 25) at Zonda, Pie de 
Palo, Hualilan and Leoncito. The 
principal station at Zonda also con- 
tained many other seismometers of 
different constants and amplifica- 

Also, as early as 1958 an excessive 
number of seismic events compared 
with the total number recorded had 
been observed with values of S — P 
of about 12 sec. As seen in Fig. 22 (A) 
the frequency distribution of S — P 
values for a period of 18 months 
shows a large excess in the interval 
10 to 14 sec. These data were ob- 
tained on a vertical Benioff-type seis- 
mograph with T — 0.8 sec and with 



70° 65° 60° 

Fig. 21. Location of zones studied. 

a magnification of 28,000 for signals 
at 1 Hz. The magnitude range of the 
recorded events was from 2.5 to 4.5. 

The curve in Fig. 22(B) was ob- 
tained from a long-period (T = 50 
sec) horizontal seismograph with a 
magnification of 35,000 for signals 
of period of 1 Hz and 70,000 for 2-Hz 
signals. It is seen that a slightly 
larger number of events (166) in the 
magnitude range from 2 to 4 were 
recorded than in the earlier study. 

Finally, the curve in Fig. 22(C) 
gives the distribution resulting from 
the consideration of 537 events of 
magnitude between 1 and 3 recorded 
on the Carnegie vertical seismograph 
with T„ = 1 sec and magnification 
280,000 at 5 Hz. 

The three curves are very similar. 
The figure shows a small concentra- 
tion of activity in the range 5 to 20 
km (S — P = 1 to 4 sec) followed 
by an almost total absence of events 
before beginning the more important 
concentration for S — P values be- 

tween 10 and 13 sec. Not only does 
this maximum occur at 12 sec, but 
also the percentage of events with 
this S — P time is nearly identical 
in all three distributions. 

The events occurring in the shorter 
S — P interval were well recorded at 
Zonda; a number were also recorded 
at Pie de Palo, but only a few were 
registered at Hualilan or Leoncito. 
The seisms of the second group (S — 
P = 12 sec) , in contrast, were well 
recorded at all four stations with 
only a slight variation in intensity. 
The gain at each station was ad- 
justed according to the noise level 
encountered, but the sights were spe- 
cially chosen to permit them to oper- 
ate at the highest possible gain. 

Two properties of the events with 
S — P = 12 sec were studied: 
their spatial distribution, and evalu- 
ation of the constants m and K, in 
the Ishimoto-Iida relationship 










30 - 




I 30 






July 1959 to Feb. 1961 
119 Events 
M = 2.5 - 4.5 

T = 0.8 second 


i i I i i i r 

i i i i 



ii i i 

Jan. to June 1966 
166 Events 
M = 2 -4 

T = 50 seconds 


Jr "^-T 


I I I I 

I I I l 


Jan. to June 1966 
537 Events 
M= 1 -3 

T = 1 second 

S - P (seconds) 

Fig. 22. Frequency distribution of S-P values at Zonda station. 

where n is the frequency of seismic 
events and A their maximum ampli- 
tude. Figure 23 shows a plot of fre- 
quency versus amplitude for ampli- 
tudes from 1 to 25 mm at Zonda sta- 
tion. If the events are grouped in 
amplitude at 5-mm intervals and 
plotted as log N versus log amplitude, 
Fig. 24 results. The linear fit of the 
figure implies the validity of the 
Ishimoto-Iida relationship. From 
equation 2 

log N + mlogA = log If (3) 

where K and m are constants ; m is 
the slope of the line in Fig. 24 and 
has the value 1.6, giving a value of 
K x = 3.6 X 10- 4 . 

Of the 340 seisms with S — P 
times in the range of 10-13 sec, 
about 40 had data that permitted the 
precise location of their hypocenters. 
Examples of these data are shown in 
Plate 4. The precision of the hypo- 
central location is believed to be with- 
in 2 km in any direction. The plot 
of the determinations is shown in 
Fig. 25. Their locations fall in a nar- 



E 10- 




I l i I I — I I I I l l I l 


1 I I I I I I I I I I I "T 


J_J I 1 L_L 

M i l l I i i i 



5 10 15 20 

Amplitude (mm) 

Fig. 23. Plot of frequency versus amplitude at Zonda station (amplitude 1—25 mm] 

4 6 8 10 

Log of the amplitude 

Fig. 24. Verification of formula n = KA^ 71 and emission of the constants m and K. 







• Zonda 

Piede Palo 

20 40 Km 

I —i 



West longitude 

68 c 
















• • • 
• • • 





• •• • 
• • 


• • 

• 9 

Fig. 25. Distribution of hypocenters from seismograms shown in Plate 4. 

row band about 30 km in the north- In the vertical direction they are all 
south direction and cover the full 100 located roughly between 70 and 100 
km of the net in the east-west direc- km, and it is presumed that this ac- 
tion. The band probably extends far- tive volume is part of the upper 
ther, both to the east and to the west, mantle. 



In summary, a volume of seismic 
activity has been studied near San 
Juan, Argentina, which has the fol- 
lowing* properties: 

1. The trace amplitude, frequency 
of occurrence relationship can be 
characterized by the Ishimoto-Iida 

seisms mm 1 - 6 

nA 1G = 34 X 10- 

km 3 year 

2. The volume studied is 30 X 30 
X 100 km 3 and lies with its axis nor- 
mal to the Andean cordillera between 
Chile and Argentina. 

3. About 30 ^c of all the local 
earthquakes recorded in this area 
100 x 100 X 100 km 3 lie in this 
active volume. 

4. Xo surface expression can be 
correlated with this activity in the 
upper mantle. 


S. R. Hart, T. E. Krogh, G. L. Davis, 

L. T. Aid rich, F. Munizaga, and 

A. M. Stueber 

Potassium, Rubidium, and 

Strontium in Ultramafic Rocks 

and Minerals 

A. M. Stueber 

Ultramafic rocks are of importance 
in the earth sciences because they are 
commonly considered to be the domi- 
nant material of the earth's mantle. 
Furthermore, ultramafic rocks are 
generally regarded as having some 
sort of genetic relationship with ba- 

The determination of the potas- 
sium, rubidium, and strontium con- 
tent as well as the strontium isotopic 
composition of ultramafic rocks and 
minerals from a small area of west- 
ern North Carolina has been under- 
taken. Prior work of this nature on 
ultramafic rocks has been primarily 
in the form of survey studies which 
have made significant contributions 
but at the same time have posed sev- 

eral equally significant and as yet 
unanswered questions. It has been 
shown, for example, that the Sr 87 / 
Sr 86 of alpine ultramafic rocks of 
wide geographic distribution are 
rather variable and definitely higher 
than the Sr S7 /Sr 86 of basalts, but the 
possible variability of Sr S7 /Sr 86 in a 
single alpine ultramafic intrusion has 
not been investigated. Similarly, it 
has been shown that the strontium 
concentrations of ultramafic rocks 
are generally low, but at the same 
time variable over as much as an 
order of magnitude. However, the 
distribution of the Sr, as well as the 
K and Rb, within the several mineral 
phases of ultramafic rocks has not 
as yet been established. Finally, it ap- 
pears that there is no general pro- 
portionality between Sr 87 /Sr 86 and 
Rb/Sr of alpine ultramafic rocks. The 
possibility that such a proportionality 
does exist within a single intrusion, 
and that an Rb-Sr isochron age can 
therefore be determined, remains to 
be tested. 

A detailed and systematic sampling 
program was carried out on five 
closely related alpine ultramafic in- 
trusions in western North Carolina: 
Balsam Gap, Dark Ridge, Addie- 
Webster, Buck Creek, and Day Book. 
Samples were selected on the basis 
of variable mineral composition and 
variable degree of serpentinization. 
More than 100 grams of each of these 
samples were ground with care to 
avoid contamination and were inti- 
mately homogenized in an attempt to 
improve reproducibility of K, Rb, and 
Sr analyses, and mineral separations 
were carried out on a few selected 
samples. The results of the analyses 
which have been completed thus far 
are presented in Table 4. 

It is clear that the concentrations 
of K, Rb, and Sr are quite low in 
olivine; enstatite tends to contain 
more K and Rb, but not significantly 
more Sr, than olivine; diopside will 



TABLE 4. Potassium, Rubidium, Strontium, and the Isotopic Composition of Strontium 
mafic Rocks and Minerals from Western North Carolina 











Webster pyroxenite 











0.71 05 

Webster enstatite 












. . . 

Webster diopside 












Addie peridotite 












. . . 

Addie olivine 












. . . 

Addie pyroxenes 






(enstatite + diopside) 







Dark Ridge peridotite 






• • • 






• • • 







Dark Ridge olivine 






. . . 







Dark Ridge enstatite 












Dark Ridge diopside 













* Estimated uncertainty in Sr 87 /Sr 86 after normalizing to Sr 86 /Sr 88 = 0.1194 is ±0.001. 

accept relatively large amounts of K, 
Rb, and Sr. Indeed, diopside gener- 
ally has an order of magnitude more 
Sr than either enstatite or olivine. 
This is reasonable in view of the 
tendency for Sr to substitute for Ca, 
which is present in significant 
amounts only in the diopside. 

In some cases the reproducibility 
of K and Rb concentrations is quite 
good, whereas in others there seems 
to be room for improvement. Because 
K and Rb vary sympathetically (as 
indicated by the rather good repro- 
ducibility of K/Rb over a wide range 
of values), and because differences 
between duplicate analyses are sim- 
ilar on a percentage basis at both 
high and low concentration levels, it 
is thought that differences in dupli- 
cate analyses for K and Rb, at least 
in mineral separates, are the result 
of inhomogeneous distribution of K 

and Rb within the minerals. How- 
ever, differences in duplicate analyses 
for whole rocks may be the result of 
inhomogeneous distribution of min- 
eral phases within the sample, in 
spite of the attempts to eliminate this 

From these analyses of mineral sep- 
arates it seems reasonable to conclude 
that the concentrations of K, Rb, and 
particularly Sr, in an ultramafic rock 
are very critically controlled by the 
amount of diopside present in the 

The K/Rb of these rocks and min- 
eral separates are plotted as a func- 
tion of K concentration in Fig. 26. 
Stueber and Murthy 29 have shown 
that the K/Rb in ultramafic rocks 
generally vary between 200 and 500 
and seem to be unrelated to K con- 
tent. Such a conclusion would apply 
in a general way to the data pre- 






! 1 
Dark ridge peridotite 
Addie peridotite 

1 1 



V- ebster pyroxenite 


En* D! 


_ t 

-i ^" 



1 >^ 


-/dT 0£ 


En | | 

IV \ 



K, ppm 

300 700 

Fig. 26. Potassium-rubidium ratio as a func- 
tion of potassium content of minerals from 
ultramafic rocks from western North Carolina. 

sented here. However, if each ultra- 
mafic rock is considered as a separate 
system it can be seen that the K/Rb 
invariably increase with increasing K 
content. Such behavior is opposite to 
the trend observed in igneous rocks 
in general, and is similar to that re- 
ported by Hart 30 for the ultramafic 
samples from St. Paul's rocks. More- 
over, for each ultramafic rock re- 
ported here, the K/Rb increases from 
olivine to enstatite to diopside, which 
is the normal trend of differentiation. 
Ordinarily Rb, because it has a larger 
ionic radius than K, is enriched in 
magmatic liquids relative to K as dif- 
ferentiation proceeds. The apparently 
opposite behavior for the ultramafic 
minerals may represent an equilib- 
rium distribution for K and Rb. On 
the other hand, it may be due to the 
effects of partial fusion, since alpine 
ultramafic rocks can be regarded as 
residual. Partial fusion of a rock or 
mineral results in a decreased K/Rb 
in the liquid and an increased K/Rb 
in the residue. Selective partial fu- 
sion of diopside or enstatite to pro- 
duce basalt might result in a residual 
ultramafic rock in which the K/Rb 

are higher in the diopside and the 
enstatite than in the olivine. 

The Sr S7 /Sr 86 determined thus far 
fall in the same general range as 
those previously reported for ultra- 
mafic rocks. There is no indication of 
strict proportionality between Sr 87 / 
Sr S6 and Rb/Sr. In the four analyses 
from Webster and Addie, which have 
very similar and low Rb/Sr, there is 
a suggestion of proportionality, and 
this is now being investigated. It is 
significant that the Dark Ridge sam- 
ples, with distinctly higher Rb/Sr, 
have distinctly higher Sr 87 /Sr 86 than 
the Addie-Webster samples. The sam- 
ples from Dark Ridge, however, show 
a very close similarity in Sr 87 /Sr 86 in 
spite of significant differences in 
Rb/Sr. The results obtained thus far 
are too few to permit final conclu- 
sion, but they are tantalizing, and 
further work is in progress. 

Geological History of the 
Grenville Province 

T. E. Krogh and G. L. Davis 

Age measurements this year pro- 
vide the basis for a major revision 
of the geological time sequences in the 
Grenville province. The real age of 
the rocks in this area is uncertain be- 
cause age determinations on minerals 
yield values of about 900 :± 100 m.y. 
(Stockwell, 1964), 31 whereas certain 
other geological arguments suggest 
that the rocks are much older. In 
earlier work we have determined age 
values between 1500 and 2000 m.y. for 
rocks from this area using the whole- 
rock rubidium-strontium method. 
Now a major proportion of our effort 
is directed toward evaluating the 
effect of the 900-m.y. event on the 
whole-rock system. Certain innova- 
tions in sampling techniques have 
made it possible to draw definite con- 
clusions about the volume of the rock 
that remained a closed system with 
respect to migrations of dating ele- 



ments during the 900-m.y. event. 
Most of the rocks in the region are 
paragneisses and have definite char- 
acteristics indicative of having under- 
gone deep burial, intense heating, and 
plastic deformation. It has been 
commonly assumed that these char- 
acteristics were established during 
the 900-m.y. event. Through applica- 
tion of our sampling techniques we 
have been able to show that some of 
these characteristics were established 
during an earlier metamorphism 
about 1500-1800 m.y. ago. 

Our investigation has been two- 
fold. In certain cases we have studied 
smaller and smaller details in order 
to evaluate the dating technique. In 
others it has been necessary to meas- 
ure widely spaced samples to learn 
about regional geological patterns. 
In detailed studies of two granite 
bodies a good fit of points to isochron 
plots for 1500 and 1700 m.y. was 
accomplished. This establishes the oc- 
currence of two discrete events and 
eliminates the possibility that these 
ages were the result of a younger, 
say 900-m.y., event acting on rocks 
with a much older true age. Our 
studies of widely spaced samples com- 
plement the more detailed ones, as 
we have found evidence that two 
granites in other parts of the region 
were emplaced about 1500 m.y. ago 
and that another major granite body 
formed about 1700 m.y. ago. Simi- 
larly, studies of paragneiss samples 
collected from a single outcrop yield 
age values for the most part between 
1500 and 1800 m.y., as do samples 
collected over much of the northwest 
Grenville area, and over the eastern 
Grenville area along the St. Lawrence 

The French River area as an ex- 
ample of the application of the 
method. Measurements on whole-rock 
samples from a single granite body 
in the French River area of Ontario 
yield a whole-rock isochron age of 

1725 m.y. (Fig. 27). Circled points 
shown on the diagram, representing 
a special rock type, will be discussed 
later. The relatively small standard 
deviation of 16 m.y. in the slope of 
the line through this series of plotted 
points has considerable significance 
both in the evaluation of the assump- 
tions of the whole-rock method and 
in a geological sense. An explanation 
of the significance and of the mean- 
ing of the mineral data also shown 
in Fig. 27 will require a digression 
into certain fundamental concepts of 
Rb-Sr dating. In this figure unnum- 
bered points represent data for 
whole-rock samples. Sample Ont 4-0, 
whole rock (WR), is joined with a 
broken line to the data point for 
apatite in this rock. The data point 
for biotite would lie far to the right 
of the diagram if it were shown. In 
the whole-rock dating method, rep- 
resentative samples split from 5 to 50 
kg of rock are analyzed for total Rb 
and Sr as well as for strontium iso- 
topic composition. The age is deter- 
mined by comparing the relatively 
small amount of radiogenic Sr 87 gen- 
erated in the rock with the amount 
of radioactive parental Rb 87 present 
in the rock. Because we have no way 
of directly measuring the amount of 
Sr 87 present when the rock was 
formed, we assume that at that time 
the abundance of Sr 87 relative to one 
of the stable Sr isotopes (Sr 86 by con- 
vention) was the same in all the sam- 
ples. The assumption is verified if the 
analytical data, expressed on a plot 
of Sr 87 /Sr 86 versus Rb/Sr, lie on a 
straight line. 

The geological significance of a 
high degree of fit to a straight line 
lies in the implication that all samples 
did in fact contain isotopically iden- 
tical Sr at one time. The isotopic 
composition of the initial strontium is 
shown at the intercept of the line 
with the vertical axis. Any argument 
suggesting that a particular granite 



• I 

0.790 - 




co 0.750 









ONT 4-0, WR ■ 
APPARENT ISOCHRON 1725 ± I6m.y. /V _ 





0.4 0.6 



Fig. 27. Isochron plot for the French River, Ontario, granite. Solid line is the whole-rock iso- 

chron. Broken line is the mineral isochron. 

is older than the indicated isochron 
age must include some proposed 
mechanism by which the isotopic 
composition of Sr can be made iden- 
tical in all parts of the sampled body. 
This requirement is most severe when 
vast volumes of granite act as a 
buffer to external chemical and phys- 
ical effects. Thus, if a major granite 
mass can be shown to have a high 
Rb-Sr ratio over an extensive region, 
this entire region must be purged of 

radiogenic strontium in order to elim- 
inate traces of an earlier history. In 
contrast to major volumes of whole 
rock, minerals need only to exchange 
their strontium over a distance of 
millimeters in order to lose all specific 
trace of a past history. 

Data points for apatite, whole rock, 
and biotite (Fig. 27) lie approxi- 
mately on a straight line, and, if an- 
alyzed, other minerals should lie on 
or near the same line because apatite 


and biotite are probably the least and ical interactions occurred at the time 
the most reactive minerals in the of last mineral equilibration or dur- 
rock. This relationship establishes ing an earlier event, 
that the minerals and whole rock all In the French River granite, nar- 
achieved an identical isotopic compo- row (5-10 cm) amphibolite dikes 
sition at a time indicated by the slope occur at one outcrop area. The vari- 
of the line through the data points, ation of Rb and Sr in a suite of 
The Sr 87 /Sr 86 of the strontium com- samples cut from a section taken per- 
mon to all minerals was 0.738 (Fig. pendicular to a dike shows that the 
27), and the time of establishment of Rb content drops to half and the Sr 
this isotopic composition was about content doubles along the edge of the 
1000 m.y. ago. The slight difference dike (0-2 cm) over that present in 
between the apatite whole-rock age granite 7-15 cm from the dike. Iso- 
value and the biotite whole-rock age topic data for these samples (Fig. 27, 
value probably indicates that biotite circled points) demonstrate that corn- 
continued to lose radiogenic stron- plete isotopic mixing did not take 
tium after the apatite had ceased to place between the samples at the time 
gain radiogenic strontium. Any min- of late mineral recrystallization 
eral bias introduced in sample split- (1000 m.y.). These Rb and Sr vari- 
ting will move the data points up or ations developed either during an 
down this mineral isochron and will earlier metamorphic event or during 
thus introduce scatter into the whole- the intrusion of the dike, 
rock data. French River paragneiss. Chemi- 

The occurrence of mineral iso- cally layered rocks that are thought 
chrons demonstrates intergranular to be of sedimentary origin but now 
migration of strontium but tells are intensely recrystallized and de- 
nothing about the distances involved formed into paragneisses are com- 
during the migrations. There is no mon in the French River area. One 
requirement, for example, that the outcrop exposed approximately 2 
isotopic composition of strontium was miles south of the granite was suit- 
ever identical in all grains of a single ably layered for studying isotopic 
mineral in one rock. migration. Specifically, we hoped to 

Evaluation of the size of the closed determine the time at which the 

system. To evaluate the distances in- gneiss formed from its sedimentary 

volved in isotope migration we have precursor. Block samples represent- 

obtained samples in which the min- ing a continuous section of approxi- 

erals, rather than being in a three- mately 18 inches of rock perpendicu- 

dimensional mosaic, are concentrated lar to the layering were obtained, 

into discrete layers. Monomineralic Within each layer of granitic gneiss 

layers are not easily obtained; how- the Rb concentration drops by about 

ever, only layers with finite differ- a factor of 3, whereas the Sr content 

ences in their Rb/Sr are required, increases by about 50%, as the edge 

We have often found that chemical of each biotite-amphibolite layer is 

interactions have occurred between approached. The abrupt drop in Rb 

rocks with different chemical compo- and increase in Sr correspond to the 

sitions to form layers with interme- occurrence of a plagioclase-rich mi- 

diate or contrasting characteristics, crocline-free layer adjacent to each 

Insofar as the layers can be shown amphibolite layer, 

to be reaction zones formed during Further studies of these variations 

intense heating of the region, we are are in progress, but at this point it 

able to determine whether the chem- is reasonable to conclude that we are 



examining the products of major 
chemical interactions that occurred 
at the time of formation of this 
gneiss. With this degree of chemical 
interaction it is a reasonable assump- 
tion that the isotopic composition of 
Sr would be the same in all parts of 
the section at the time of the re- 
actions. If the reactions occurred dur- 
ing the 1000-m.y. event, data points 

for all parts of the gneiss would lie on 
a line with a 1000-m.y. slope. If, how- 
ever, the gneiss formed during an 
earlier event, further modification of 
the isotopic relationships during the 
1000-m.y. event are to be expected. 

As shown in Fig. 28, samples 1A 
and 2B, each approximately 1 cm 
wide, are located at an abrupt Rb/Sr 
discontinuity; in Fig. 29 data points 


Fig. 28. French River, Ontario, paragneiss. Diagram showing changes in concentration of 
rubidium and strontium along a section normal to the layering. The vertically dashed layers 
are granitic gneiss; the crosshatched layers are biotite-amphibolite. 




0.750 - 






0.730 - 

1622 m.y. ± 70 m.y. 






0.4 0.6 



Fig. 29. Isochron plot for the French River paragneiss. 

for these two samples define a line 
with a slope comparable to that for 
minerals in the rock. The other 
samples of gneiss do not lie on a 
single line with this slope but rather 
lie approximately along the isochron 
transferred from Fig. 27 for the 
French River granite. An arithmetic 
composite sample (1A, 2AB, SI) also 
appears to lie on this line, suggesting 
that the enclosed volume may have 
been closed to migrations of the dat- 
ing elements during the 1000-m.y. 

event. We conclude that the gneiss 
formed during an earlier metamor- 
phism and did not form during the 
late heating event demonstrated by 
the apatite in sample 66-88. If we 
propose limited isotopic mixing dur- 
ing a time when the major phases 
remained stable we should expect low 
points on the Rb/Sr profile of Fig. 28 
to gain, and high points to lose, ra- 
diogenic Sr 87 . Thus samples 2AB and 
1C do lie below samples with compa- 
rable Rb/Sr in Fig. 29. Similarly, the 


samples occupying- the valleys in the (Year Book 65, pp. 380-383). Whole- 
profile of Fig*. 28 do show an increase rock data for three locations from 
in radiogenic Sr over that which the each of the two dominant granite 
higher samples would have had 1725 types present are shown in Fig. 30. 
m.y. ago. Also, a mineral isochron Samples from three separate out- 
could be passed through samples 3Z> crops of the felsic granite (Ont-1, 
and 3D as if they had locally ob- 65-165, and 65-162-0) define the same 
tained an identical isotopic composi- line as the three samples from differ- 
tion approximately 1000 m.y. ago. ent parts of the biotite-hornblende 

A least-squares line through all the granite, 65-199, 201, and 202. Again, 
granitic gneiss samples yields an age the minerals have undergone isotopic 
value of 1622 ± 70 m.y.; another exchange long after the time of 
isochron between three samples of formation of the rock. As noted on 
gneiss (not plotted on Fig. 29) from the diagram, sample 65-165 is a corn- 
other parts of the outcrop defines a posite; it was made in the field by 
line for 1430 ± 26 m.y. The three taking samples approximately 5x5 
high points on the isochron yield age X 2 cm at 1-meter centers over ap- 
values between 1750 and 1900 m.y., proximately 5 square meters of 
assuming an initial Sr S7 /Sr 86 of 0.704. freshly blasted road cut. Results for 
Clearly, the metasediments must be a 5-kg sample from this outcrop an- 
older than 1725 m.y., the age of the alyzed in duplicate last year lie near 
granite that presumably intrudes but not on the line defined by the 
these gneisses. The data demonstrate other whole rocks. The implication of 
that a paragneiss subjected to re- this test of sampling technique is that 
gional metamorphism after the time an average of many possibly open 
of its formation may yield a variety subsystems collected over a large 
of age values. Our best estimate of area of homogeneous material may 
the time of formation of this gneiss is approximate a closed system better 
obtained by considering the samples than a single block sample, 
farthest removed from variations in Another composite, 162-0-9, shown 
the Rb/Sr profile shown in Fig. 29 in Fig. 30, was made in the labora- 
and by eliminating from considera- tory by combining hand-specimen- 
tion such samples as 3/ and 1A, sized samples collected at 2-meter in- 
which are susceptible to additions of tervals across a freshly blasted road 
radiogenic strontium. Sample 66-88 cut. Results for this composite, and 
represents about 30 kg of rock from the 10-kg whole-rock sample from the 
the thick layer 87-3; hence this sam- same outcrop, are essentially identi- 
ple and sample 3Z) probably define cal. A contact between the felsic 
our best approximate age at about granite (65-162) and a biotite-horn- 
1750-1900 m.y. for an assumed ini- blende granite occurs on this outcrop, 
tial ratio of 0.704. The spread in ages A study of the Rb and Sr concentra- 
possible for a single outcrop is in- tions across the contact indicates a 
structive in interpreting single whole- gradual drop in Rb from 170 to 95 
rock paragneiss age values reported ppm and an increase in Sr from 140 
later in this section. to 220 ppm within a distance of about 

Lake Muskoka granite and other 10 cm. Such a gradual variation 

coeval granites. Additional work has could occur for a magmatic contact 

been completed on a granite body that or, alternatively, might have devel- 

occurs on the eastern shore of Lake oped during the metamorphism of 

Muskoka, approximately 150 km the granites. The rocks themselves 

southeast of the Grenville front show the effects of having recrystal- 








co 0.760 




1 1 1 1 1 1 — r 

i — i — r 

65-162-0 WR 

APPARENT ISOCHRON 1497 ± 10 m.y. 










■ ONT 1-2 

" ONT I 




i 65-202 


J I L 

J I L 

J I 

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 


Fig. 30. Isochron plot for a granite at Lake Muskoka, Ontario. 

lized under directed stress with the 
development of lenticular pods, a few 
millimeters thick, of quartz and feld- 
spar or of mafic minerals. Here the 
foliation is parallel to the contact. 
Data for two samples on the estab- 
lished Rb/Sr gradient are shown in 
Fig. 30. They were sawed from a 
block containing the contact, and 
were 10 cm apart. Because their 
points do not define a line whose slope 
is parallel to that of the mineral iso- 
chron the rock was not isotopically 
mixed during the last recrystalliza- 
tion (700 to 800 m.y. ago). They do 
lie within the error of our measure- 
ments, on the line denned by widely 

spaced samples from other parts of 
the body. Therefore, the Rb/Sr gra- 
dient of which the samples are a part 
must have been established approxi- 
mately 1500 m.y. ago. 

It is a reasonable assumption that 
the reerystallization exhibited by 
these rocks would be accompanied by 
isotopic mixing of strontium, over a 
scale of at least a few centimeters. 
If so, then the metamorphism of the 
granites occurred approximately 1500 
m.y. ago. 

In two other localities ages within 
1% or 2% of those obtained in this 
study have been found. Two whole 
rocks from a single large outcrop 


near Round Lake approximately 35 provide the most comprehensive pic- 

km southwest of Pembroke, Ontario, ture of local and regional variation 

yield two-point isochrons of 1510 yet obtained for a continental area. 
m.y. A single sample from an outcrop 

south of Sandridge, Ontario, yields Technique and Results 
a whole-rock age of 1475 m.y., when 

an initial ratio 0.703 is used. In addition to the determinations 
Pa ragncissc >,s. Age values for para- on the 15-km grid, to be discussed 
gneisses and granitic rocks in On- later, a number of special observa- 
tario and Quebec can be calculated tions were made to assess the accu- 
from single whole-rock analysis if we racy and precision of the method 
assume an initial Sr S7 /Sr S6 . For ex- more objectively than had previously 
ample, for the French River para- been possible. The repeatability of the 
gneiss, age values between 1500 and measurements was tested by making 
1900 m.y. can be calculated. Five out two successive drops at a location off 
of 6 paragneiss samples collected in the Keweenaw Peninsula; seven days 
a region approximately 300 X 150 km later approximately the same site 
lying east of Georgian Bay yield age was reoccupied and a third value ob- 
values in that range. Three samples tained. The first two observations 
of paragneiss from the area between produced identical heat flow values of 
the Saguenay River and Sept Isles, 1.11 hfu (1 hfu = 1 microcalorie of 
Quebec, on the southeast region of flux per square centimeter per see- 
the Grenville, yield age values be- ond), and the third yielded 1.06, 
tween 1550 and 1650 m.y. Three indicating with little doubt that the 
granitic rocks with age values be- method gives results replicable with- 
tween 1000 and 1300 m.y. occur in in 5%. 

this area, as well as three others with It has been suggested that a rea- 
age values between 1500 and 1600 sonable evaluation of the absolute 
m.y. These values indicate that pre- accuracy of the deep-lake technique 
Grenville rocks occur in the eastern could be obtained by direct compari- 
as well as the northwestern parts of son with results from conventional 
the Grenville Province, and that the borehole determinations. To this end, 
age patterns in the two areas may a profile of 5 observations was con- 
be similar. structed. This profile extended into 

the lake from a point near the bore- 

HEAT FLOW hole gites of Birch ( 195 4)32 and Roy 

S. R. Hart, J. S. Steinhart, and T. J. Smith (1963) 33 on the Keweenaw Peninsula. 

The program to use modified As may be seen from Fig. 31, all the 

oceanic techniques for the measure- values lie in the range from 1.06 to 

ment of terrestrial heat flow in deep 1.17 hfu and, taking into account the 

lakes was continued during the year, observed local variation, are readily 

and roughly half of the proposed re- compatible with the Birch and Roy 

gional survey of Lake Superior was values of 0.93 and 0.95 hfu. There is 

completed. More than 90 measure- some indication that the observation 

ments of the bottom temperature gra- nearest the shore (5 km) is being 

dient were made from the U.S. Coast disturbed by lake edge effects. The 

Guard Cutter Woodrush with the only other land observation in the 

Department's 6-meter thermistor area is in the Porcupine Mountains, 

probe, largely in the western and cen- where the reported heat flow of 1.07 

tral part of the lake. The heat flow hfu (Roy, 1963) agrees very well 

values calculated for 83 of these sites with the nearest lake result of 1.02 





• D 
O > 

























hfu obtained about 20 km to the 
north of the Roy station. 

Note has been made (Year Book 
6.4) of the necessity for correcting 
the observed temperatures to account 
for the seasonal variation in the bot- 
tom lake water. A daily bottom tem- 
perature recorder located between the 
Bayfield and Keweenaw Peninsulas 
in 640 feet of water was operated 
from October 1965 to October 1966 to 
obtain the Lake Superior annual tem- 
perature cycle displayed in Fig. 32. 
Perturbations to the sediment tem- 
peratures caused by this external flux 
have been calculated both for a ho- 
mogeneous half-space model and for 
a model consisting of a single layer 
overlying a half -space. (For details 
of the latter calculation, see Lachen- 
bruch, 1959. 34 ) As one might expect, 
corrections to the observed tempera- 
tures derived from the single-layer 
model come closer to linearizing the 
gradients than those derived from 
the simpler half-space model. For this 
reason the heat flow values in this 
report have been computed from the 
bottommost portions of the gradients 
corrected under the single-layer as- 
sumptions. Conductivities for both 

the model and the heat flow calcula- 
tions are based on measurements 
made on the individual cores taken 
with each observation. It is hoped 
that a model, now under study, which 
allows the continuous variation of 
conductivity with depth, will permit 
a still more realistic correction for the 
annual cycle to be applied, thereby 
making available for use those values 
of the gradients at shallower depths. 
The validity of the annual correc- 
tion does not, however, depend solely 
upon an appropriate choice of a 
model relating thermal conductivity 
to depth in the mud. Figure 33 dis- 
plays temperature measurements for 
four of the cores. Also shown are the 
measurements after correction for 
the annual temperature cycle. The 
single layer over a half-space model 
is used. The four cores cover a range 
of water depths, a range of time dur- 
ing the experiment, and include one 
of the most successful corrections and 
one of the least successful. It seems 
there is a depth effect on the annual 
temperature cycle of the bottom water 
such that measurements in water 
shallower than that in which the an- 
nual cycle was measured received 

Time connecting observations 
Smoothed cycle 


Jan Feb 

Mar Apr May Jun July Aug 

Fig. 32. Daily variation of bottom water temperature in Lake Superior. AAeasured offshore 
of Porcupine Mountains in water depth of 640 feet. 




f 3.0 

:£ 4.0 




• Observed 

Corrected for 
annual change 

3.4 3.5 3.6 3.7 3.8 3.5 3.6 3.7 3.8 3.5 3.6 3.7 3.6 3.7 3.8 3.9 

Temperature, °C 

Fig. 33. Corrected and observed temperatures at four sites in Lake Superior. 

insufficient correction and those from 
very deep water were overcorrected. 
It is not immediately clear whether 
only the amplitude of the bottom an- 
nual cycle changes with water depth 
or whether the phase changes as well. 
Review of a large number of the 
measurements suggests that there is 
no simple, precise answer to this 
question, but a first-order correction 
for depth may be possible. 

The thermal conductivity and wa- 
ter content of each of the cores were 
measured at selected points with 
spacing varying from 100 to 10 cm 
to provide reliable interpolation to 
any point of interest in the sediment 
column. Conductivities were deter- 
mined with a needle probe by a 
method similar to that of von Herzen 
and Maxwell (1959) 35 ; water content 
was analyzed by conventional wet 
and dry weighing. The resulting plot 
of conductivity versus water content 
based on about 140 pairs of data 
bears a qualitative resemblance to the 
graph presented by Ratcliffe (I960) 36 
for ocean sediments, but there is a 
measurable absolute difference be- 
tween our derived relationship and 
that of Ratcliffe. Moreover, the scat- 

ter of the data about a best-fitting 
curve is several times as great as 
would be expected from estimates of 
the largest uncertainties in the meas- 
urements. The absolute discrepancies 
can be explained by differences in the 
sediments used in the two studies, 
but the significant scatter of the data 
carries the more disturbing sugges- 
tion that any average relationship 
between conductivity and water con- 
tent may, for a specific sample, be in 
error by as much as 4% or 5%. 
Additional inquiries into this problem 
are planned for the coming field 

As a note of interest it has been 
observed that the bottom muds of 
Lake Superior lose their moisture at 
a truly astonishing rate. A sample 
from core 10 (Fig. 34) with water 
content of about 50 % lost almost 1 % 
of its initial weight each 20 minutes 
when placed in an open dish at room 
temperature. The clear implication is 
that care must be given wet cores 
when their water content is a matter 
of quantitative importance. 

Discussion. The range of published 
continental heat flow values is from 
0.7 to about 2 hfu; rarely, in areas 




°> 15.3 





* 15.1- 


T4 # 9- in open dish 


X % 

x x 

Core no. 10-2.5 
Air drying at 25 C C 



10 20 30 40 50 60 70 80 
Time, min 

Fig. 34. Weight loss due to water loss as 
a function of time for a sample of Lake Supe- 
rior mud. Note very high loss rate. 

of recent tectonic activity, flux of 2.5 
hfu has been observed. The small 
number of published values for the 
continents has made it difficult to 
determine how the range of conti- 
nent-wide variation compares with 
variation in a small geographic area. 
In the oceans variations in heat flow 
of a factor of 2 or more have been 
observed in traverses of only a few 
hundreds of kilometers, and there is 
a suggestion of changes of less dra- 
matic amplitude in a profile of conti- 
nental deep-hole measurements made 
in New England by the Harvard 
group. From the 1966 measurements 
(Fig. 31) we are able for the first 
time to get a clear idea of the extent 
of regional variation of heat flow in 
a continental area. It has sometimes 
been stated in the past that heat flow 
in the stable shield areas, of which 
Lake Superior province is a part, ex- 
hibits remarkably uniform heat flow, 
and for the few measurements pre- 
viously published the variation was 
not large. We now see that there is 
a variation of more than a factor of 
2 even within Lake Superior. Because 
of the unusual crustal structure 
found in the Lake Superior region 
(Year Book 6U\ see also The Earth 
beneath the Continents, Steinhart, 
Smith, eds., 1966), this is a place 

where variation would be likely. The 
significant finding is that for the first 
time on the continents large and sys- 
tematic heat flow variations have 
been exhibited on a regional scale. 

In the range of observed values 
from about 0.5 to 1.2 hfu, or a little 
higher, the most interesting are the 
low ones, since the lowest values pre- 
viously reported anywhere on the 
continents were about 0.7 hfu. The 
values on the high end of the range 
encompass all the values usually 
found on the stable continental shield 
areas, although those between 1.2 and 
1.3 are a little higher than the typical 
shield average, which is about 0.8 or 
0.9. The low values occur along the 
western end of the lake in the same 
region that one very low value was 
found and thought to be anomalous 
in the preliminary measurements of 
1963 (Year Book 63). It may be 
noted that, if average radioactive iso- 
tope concentrations are taken for 
many of the typical crystalline rocks 
of the Canadian Shield, and their 
heat flow generation is considered, a 
layer 15-20 km thick would be suffi- 
cient to produce a measured heat flow 
of 1 hfu. 

It is known, however, both from 
the seismic velocity structure and 
from other considerations, that ra- 
dioactive elements are concentrated 
upward within the crust, so that, 
although the bulk of the heat flow is 
the result of radioactivity within the 
crust, as much as 20% of the heat 
may be generated below the Mohoro- 
vicic discontinuity. The values near 
0.5 could then be explained by simply 
requiring that the normal shield 
rocks (granites, gneisses, etc.) form 
only a thin layer in this area, if they 
are present at all, and that rocks 
containing about one half the usual 
amount of radioactivity comprise the 
principal thickness of the crust. The 
commonly occurring rocks of low 
radioactive concentrations are the 


mafic rocks, such as basalt and gab- be rather uniform over the area, we 

bro, and they could be expected also might expect that heat flow could be 

to exhibit high seismic velocities. In directly correlated with the thickness 

the area of the low values such high of the crust. Comparing the heat flow 

seismic velocities were found in the map (Fig. 31) with the seismic pro- 

1963 seismic experiment (Year Book file (Year Book 6U) , it is clear that 

6A) . On the shore to the west of this in a general way there is a correla- 

area the Duluth gabbro and Kewee- tion. The thickest parts of the crust 

nawan basalts are the principal sur- off the Keweenaw Peninsula produce 

face rocks and extend to considerable, the highest values, and the compara- 

though unknown, depths. Thus, it tively thin crust at the western end 

appears that there is no great diffi- of the lake produces the lowest 

culty in providing a mechanism that values. In detail, however, the agree- 

accounts for low values, although it ment does not appear as convincing, 

is harder to explain the extent of the and the more detailed comparison 

local variation in heat flow. Materials will, of necessity, await the additional 

with very high seismic velocities are data to be obtained in 1967. Perhaps 

found underlying the whole region the revisions in the model of the 

of Lake Superior at comparatively crustal structure in the Lake Supe- 

shallow depths. If the concentrations rior region reported this year wil 1 

of radioactivity are really so low, provide for a better fit to the heat 

then the variation in values must be flow anomalies. This possibility will 

explained by lateral structural be examined along with other geo- 

changes that are not altogether clear physical comparisons when the heat 

in the seismic results. If we assume flow of the eastern part of the lake 

the concentrations of radioactivity to has been measured. 


L. Brown, W. K. Ford, Jr., C. Petitjean, Vera C. Rubin, W. Trachslin, 
K. C. Turner, M. A. Tuve, and C. M. Varsavsky 

The astrophysics program at the foil excitation. The results of these 
Department is a combination of in- diverse investigations are reported 


teracting interests and activities, below. 

Basic problems of galactic dynamics 

are being studied with observations 

from the 21-cm hydrogen line re- W. K. Ford, Jr., and Vera C.Rubin 

ceiver and from the image tube spec- Optical astronomy at the Depart- 
trograph. Thus, although the tech- ment during the report year has 
niques involved are quite different, centered about the use of the DTM 
many of the same underlying ques- image tube spectrograph. Early in 
tions provide motivation for both the report year a semisolid //2.25 
investigations. On the other hand, the schmidt spectrograph camera, de- 
DTM Van de Graaff generator has signed by Dr. I. S. Bowen and fabri- 
been used for extensive and detailed cated by Davidson Optronics, was de- 
experiments on scattering of polar- livered. This camera has produced 
ized protons in order that the nature spectra of substantially better quality 
of the nuclear force may be better than cameras used earlier, with no 
understood, as well as for exploratory significant loss in speed. The spectro- 
work in atomic spectra produced by graph was used by Ford and Rubin 



on about 20 nights, at the 84-inch 
Kitt Peak telescope, the 72-inch Per- 
kins telescope of the Ohio State and 
Ohio Wesleyan Universities at Lowell 
Observatory, the 31-inch Fan Moun- 
tain telescope of the University of 
Virginia, and the 36-inch Belts ville, 
Maryland, telescope of the Goddard 
Space Flight Center, NASA. Obser- 
vations were started on three contin- 
uing programs: (1) radial velocities 
and line intensities of emission re- 
gions in the Andromeda galaxy; (2) 
radial velocities of galaxies and com- 
pact blue objects in the Virgo cluster 
of galaxies; and (3) studies of veloc- 
ity fields in selected galaxies of inter- 
est. In addition, spectra of suspected 
M dwarfs in the Pleiades cluster were 
obtained in collaboration with Rev. 
M. F. McCarthy, S.J., a visitor from 
the Vatican Observatory. Most of the 
results discussed below are based on 
very few observations and are there- 
fore preliminary. 

M31. To study the detailed dynam- 
ics of a galaxy, it is necessary to have 
accurate radial velocities from re- 
gions all across the galaxy. Recent 
21 -cm hydrogen line observations by 
Burke, Turner, and Tuve (Year Book 
63, p. 322) have stimulated renewed 
interest in comparing optical and ra- 
dial velocities in M31. As a start on 
a program to obtain velocities of 
higher accuracy than previously 
available, four emission regions in 
the south-preceding end of M31 
(Baade's numbers 4, 7, 8, and 22) 
were observed. For each region one 
plate was obtained with the region 
held fixed on the slit. A velocity was 
obtained from several measures along 
H«, and the probable error listed be- 
low measures the internal agreement 
on a single plate. The observed ve- 
locities of —461 ± 13, —538 ± 6, 
—450 ± 78, and —588 ± 8 km/sec, 
correspond to velocities of — 314, 
—291, —200, and —295 km/sec in 
the plane of that galaxy, in good 

agreement with the rotational veloci- 
ties determined from the 21-cm 
neutral hydrogen observations by 
Burke, Turner, and Tuve (1964). 37 

Observations of several globular 
clusters indicate that the absence of 
any sharp features in the red spectral 
region makes it impossible to obtain 
accurate radial velocities from plates 
centered in the red. An attempt will 
be made in future work, therefore, 
to include observations of globular 
clusters in the region containing the 
H and K lines of Ca II. 

Virgo cluster. Although the Virgo 
cluster of galaxies contains several 
thousand galaxies whose velocities 
could be obtained, fewer than 50 ve- 
locities are known. Hence all conclu- 
sions about velocity dispersion, mass, 
and equipartition of energy are based 
on a biased sample, because to date 
only the brightest members have been 
examined. The speed offered by the 
image tube system can be used effec- 
tively on telescopes of moderate size 
to substantially increase the size of 
the sample. Radial velocities of Virgo 
cluster galaxies may thus be deter- 
mined. We also wish to sample the 
spectra of compact blue objects in the 
cluster region to determine whether 
quasi-stellar objects occur with high 
spatial density in regions of high 
galaxy density. 

Of 10 galaxies observed, 5 have 
strong emission lines in their spectra, 
and velocities have been determined: 
NGC 4298, V = 1110 km/sec; NGC 
4388, V = 2652 km/sec; NGC 4639, 
V = 962 km/sec; NGC 4654, V = 
1048 km/sec; and IC 3453, V = 
2555 km/sec. The other galaxies have 
only weak absorption features, and 
all will be reobserved. From the list of 
compact blue objects in the Virgo 
cluster region (Rubin, Moore, and 
Bertiau) 38 no. 197 and TON 97 have 
been chosen for observation. Both 
objects have spectra without emission 
lines, and are probably stars in our 



galaxy. Last year we reported obser- 
vations of nos. 56 and 138. Only no. 
56 has emission lines and is probably 
extragalactic, even though its meas- 
ured velocity is V — —225 km/sec. 

Studies of selected galaxies. NGC 
3389: A supernova was discovered 
in NGC 3389, an S + galaxy in Leo, 
on February 28, 1967, by A. D. 
Chuadze at the Abastumani Astro- 
physical Observatory, U.S.S.R. We 
obtained two spectra of the super- 
nova on March 7 and March 8. The 
supernova is of Type I, with broad 
unidentified emission features resem- 
bling features in supernova studied 
earlier by Minkowski. No hydrogen 
emission is observed. A third spec- 
trum, with both the supernova and 
the brightest regions of the galaxy 
on the slit, reveals Ha emission with- 
in the galaxy. For this position angle, 
the velocity variation with the galaxy 
has been measured. 

NGC 4038: The peculiar galaxy 
NGC 4038 has been observed on 11 
plates. For each of these, the slit of 
the spectrograph was oriented along 
a particular axis of interest. Spectra 
from three position angles are shown 
in the frontispiece. At the right of 
each the position of the slit is indi- 
cated. The most notable features are 
the wide variation in the intensity 
ratio H/V(0 III) 5007 A in different 
regions of the galaxy, and the small 
velocity variations across the galaxy, 
amounting to only a few hundred 
kilometers per second, thus eliminat- 
ing any possibility that NGC 4038 is 
an exploding galaxy. 

Within 0.7° of NGC 4038 on the 
sky is a somewhat similar peculiar 
galaxy, NGC 4027. Velocities meas- 
ured on a single spectrum are ap- 
proximately the same as those in 
NGC 4038. A small wisp of material 
near NGC 4027, but not connected to 
it, is also moving with a velocity sim- 
ilar to that of NGC 4038 and NGC 

NGC 3227: NGC 3227 is a Seyfert 
galaxy with an elliptical companion, 
NGC 3226. We have obtained a num- 
ber of spectra of both objects and 
have used a wide range of exposure 

Short exposures show that the 
broad nuclear emission is due to 
discrete clouds with velocities which 
differ by several hundred kilometers 
per second from the central velocity. 
On longer exposures narrow emission 
lines extend 500 pc outside the 
nucleus. Bordering the nucleus, emis- 
sion clouds are seen to the blue of 
the narrow emission. Emission re- 
gions far from the nucleus (3500 pc) , 
with relative velocities of 150 km/sec, 
will enable an accurate mass to be 
determined for NGC 3227. 


K. C. Turner, C. M. Varsavsky, and 
M. A. Tuve 

Northern Hemisphere 

K. C. Turner 

Considerable emphasis was placed 
on equipment development during the 
report year. Receiver improvements 
have been devised which keep the 
gain of our hydrogen spectrometer 
from changing more than a few per 
cent over several hours. The receiver 
"front end" exhibiting this stability 
is awaiting shipment to the joint 
CIW-IARA observatory at Peyreyra. 
This new instrument, tested for sev- 
eral weeks on our Derwood telescope, 
shows a system noise temperature of 
about 250°K. 

A new digital recording system 
offering improved flexibility and ac- 
curacy has been installed at the Der- 
wood observatory, as well as six 
extremely narrow-band filters, only 
0.5 km/sec wide. 

The surface of the 30-meter tele- 
scope at the Avery Road station is 
now complete, and interferometric 
observations are planned for the com- 
ing year. 



-20 -10 +10 +20 

Radial velocity in the local standard of rest (km/sec) 

Fig. 35. Velocity-galactic latitude plot at / = 0. 

Northern hemisphere observations 
have been made in the region north 
of the galactic center (I = 0, 5 < b 
< 50) to investigate the nature of 
the ''central bulge" of the galaxy. 
This region is also visible at the 
CIW-IARA Observatory, and so will 
permit measurements to be made to 
a common scale in both hemispheres. 
Figure 35 shows velocity-latitude 
contours for I — 0. It can be seen 
that the principal feature is very 
narrow. A measurement with the new 
narrow-band channels shows that the 
peak has a full width at half maxi- 
mum of only 5.5 km/sec. This feature 

is so large (about 30° in I) and yet 
has so low a velocity dispersion as to 
suggest that it is quite nearby. (Its 
angular size corresponds to a dimen- 
sion of about 8.6 kpc if the feature 
is located above the galactic center.) 
This object extends from about I = 
356° to at least I = 12° at b = 30°, 
but careful study of the profiles re- 
veals the existence of at least two 
main peaks about 4 km/sec apart. 
Many more observations will be re- 
quired to find out what is going on 
in this region. 

Some preliminary observations 
have been made following the edge 



of the moon as it passes through the 
galactic plane. With this occultation 
technique, 5 minutes of arc H I clouds 
of 100° brightness should be detect- 
able at the 5<r level with a 60-foot 
dish and a 250 °K receiver. 

Southern Hemisphere 

K. C. Turner and C. M. Varsavsky 

Although work in the southern 
hemisphere has been mainly observa- 
tional, the second dish has continued 
to progress steadily, and should be 
usable for observations in the coming 

After a month's work on the crys- 
tal mixer receiver in June, a noise 
temperature of about 900°K was ob- 
tained, and observations were begun 
on the southern rotation curve, em- 
phasizing the improved accuracy ob- 
tainable with our narrow-bandwidth 
filters (2 km/sec wide). These obser- 
vations generally confirmed the ear- 
lier Australian work, although finer 
detail was observed with our greater 

The most exciting work of the year 
has certainly been that of F. Strauss, 
a student of Varsavsky. The purpose 
of his research program is to search 
for an intergalactic bridge of neutral 
hydrogen between our galaxy and the 
Magellanic Clouds. So far he has ob- 
tained 80 line profiles taken over an 
area of about 250 square degrees. 

Near the galactic plane the profiles 
show 3 to 5 separate peaks that cor- 
respond to different clouds or spiral 
arms. The peak with the highest ve- 
locity corresponds to the outer spiral 
arm of our galaxy. This arm has 
been traced without interruptions 
through the whole range of galactic 
longitudes observed, i.e. from I = 
225° to I = 285°. 

Figure 36 is a plot of the velocity 
of this arm, with respect to the local 
standard of rest, versus galactic lon- 
gitude. Figure 37 shows the position 

Fig. 36. Velocity of outer arm versus galactic 

of this spiral arm in the galactic co- 
ordinate system I, b. In both figures 
the approximate position and velocity 
of the center of the Small Magellanic 
Cloud are shown. Figure 38 is a pro- 
jection on the galactic plane of both 
the spiral arm and the Small Magel- 
lanic Cloud. To calculate the distances 
of the observed points to the galactic 
center, Schmidt's model was used 
with the parameters recommended by 
the International Astronomical Un- 
ion. For distance to the Small Magel- 
lanic Cloud we adopted the value of 
46 kpc used by the Australians. It 
can be seen that the three plots show 
a remarkable tendency of the spiral 

Fig. 37. Position of outer arm in galactic 





/ = 2-0" 






x Galacric 


• SMC 


Fig. 38. Projection of outer arm and Smal 
Magellanic Cloud on the galactic plane. 

arm toward the Small Magellanic 
Cloud, both in velocity and in spatial 
position. The brightness temperature 
of the peak varies from about 80°K 
at / = 225" to 5°K at I =285°. The 
search will be continued when the in- 
stallation of the parametric amplifier 
is completed. It is not yet possible to 
know whether this spiral arm reaches 
the Small Magellanic Cloud. 

Other observations at the IARA 
include a search, by W. G. Poppel, 
for intergalactic hydrogen in the con- 
stellation of Microscopium, where a 
region of reduced incidence of distant 
galaxies has been observed, and an 
investigation by S. Garzoli of the 
region near the galactic plane from 
298° < I < 312°. This region con- 
tains the coal sack, a well-known dark 
cloud. Both these investigations are 
in the preliminary stages and are 
representative of the richness of op- 
portunity offered by the southern sky 
and of the enthusiasm of our south- 
ern colleagues. 

L. Drown, W. Trdchslin, and C. Petitjean 

We have continued the study of 
the elastic scattering of polarized 
protons, using the University of 
Basel's polarized ion source; but the 
emphasis of the study was changed 
from target nuclei of spin zero to 
those of spin one, specifically deuter- 

ons. The completed analysis of our 
measurements for spin-zero nuclei 
r -C and 1(; has yielded a unique set 
of accurate phase shifts for each re- 
action. Dr. R. G. Seyler of Ohio State 
University has derived the formulas 
describing proton-deuteron scatter- 
ing, and we have programmed and 
checked them with the DTM com- 
puter. Analysis will take place this 
summer at Ohio State University. To 
clear up an ambiguity in the struc- 
ture of 41 K (significant because of its 
role in the theory of isobaric ana- 
logue resonances) , we have measured 
polarization effects in 40 Ar(p,p) 40 Ar. 
This was undertaken as an interest- 
ing sidelight at the request of Pro- 
fessor D. Robson of Florida State 

Elastic Scattering of Polarized 
Protons on Nuclei of Spin Zero 

The characteristics of the elastic 
scattering of protons from nuclei of 
spin zero were explained in Year 
Book 65 (p. 73), and the results of 
proton-alpha scattering were dis- 
cussed in some detail. During the 
year we analyzed our polarization 
measurements of carbon and oxygen 
in the same way in order to obtain 
more accurate phase shifts. Taking 
note of the complications of the 
phase-shift analysis of proton-deu- 
teron scattering, we made a careful 
study of ambiguities in the analysis 
of p- 12 C and p- 1G 0. By ambiguity we 
mean that more than one set of phase 
shifts can reproduce the experimental 
results. In these two examples there 
is none, providing that angular distri- 
butions of both cross section and po- 
larization are used, but if only one 
kind of data is used there is consider- 
able ambiguity. Specifically, 9 sets of 
phase shifts were found that repro- 
duce well the cross-section data for 
oxygen alone, and three of them fit 
better than the set that fits both 



cross-section and polarization meas- 
urements. The analysis of our polar- 
ization data alone is also ambiguous 
but has fewer sets of phase shifts 
that satisfy the experimental results 
than the analysis of cross-section 
data alone. 

A contour map of polarization in 
p- 12 C scattering (Fig. 39) is illustra- 
tive of the nature of the measure- 
ments. This map, which is the same 
kind as used last year for proton- 
alpha scattering (Year Book 65, Fig. 
36), shows the polarization expected 
for any given scattering angle and 
energy. Two resonances, which cause 
very rapid changes in polarization as 
a function of energy, are not shown, 
as the map is intended primarily for 
use by experimenters in analyzing 

polarized protons, and regions of 
rapid variation are not useful for 
that purpose. 

Professor Robson's suggested ex- 
periment was (as mentioned) the 
measurement of polarization as a 
function of energy for a fixed scatter- 
ing angle in 40 Ar(p,p) 40 Ar. The re- 
sults, shown in Fig. 40, established 
the spin of a closely spaced group of 
resonances in 41 K, which appear as 
a single state when observed with the 
low energy resolution of our equip- 
ment. The high resolution studies at 
Duke University, 39 which show the 
structure of this group of resonances 
in great detail, succeeded in deter- 
mining the orbital angular momen- 
tum to be I = 1, but the total angu- 
lar momentum could be either 1/2 or 



Ep (MeV) 


Fig. 39. Contour map of polarization in p~ 12 C scattering. The polarization is given in units 
of 0.01, and the contour interval is 0.05. The ordinate is laboratory scattering angle in de- 
grees, and the abscissa is laboratory proton energy in MeV. 



-0.2 — 


Ep (MeV) 


Fig. 40. Polarization as a function of energy for 40 Ar(p / p) 40 Ar at laboratory scattering 
angle of 135°. The location of resonances found by a high resolution experiment 39 are shown 
as a series of vertical marks just above the bottom edge of the graph, and the location of the 
single level used for the calculated curves is shown by an arrow. Points with error bars give 
the experimental values measured with an energy resolution of 0.015 MeV. The full curve gives 
the theoretical value for angular momentum 3/2, the dashed one for 1/2. 

3/2. The angular momentum of each 
state in the group should be the same 
according to the theory of isobaric 
analogue resonances. On this assump- 
tion we could establish the total an- 
gular momentum to be 3/2 and not 
1/2, as is clear from reference to the 
two theoretical curves in Fig. 40, 
calculated from the single-level for- 
mula by Professor Robson and J. L. 
Adams of Florida State University. 

Elastic Scattering of Polarized 
Protons on Deuterons 

Except for the scattering of nu- 
cleons on nucleons, the elastic scatter- 
ing of nucleons on deuterons is the 
simplest interaction, in terms of the 
number of particles involved, open to 
experimental investigation with our 
Van de Graaff machine. There is 
strong experimental evidence that the 
polarization effects of nucleon-nu- 
cleon scattering, e.g. Year Book 6U 
(p. 312), are two orders of magni- 
tude smaller than those we can ob- 

serve with the polarized ion source, 
but measurements reported in Year 
Book 65 (p. 77) show that this is not 
true for proton-deuteron scattering. 
Collisions of protons and nuclei with 
spin zero show the nuclear force to 
have central and spin-orbit depend- 
ence. By using deuterons as targets 
we hope to observe in addition a de- 
pendence on the relative orientation 
of the two spins. 

The analysis follows the form used 
for spin-zero nuclei as reported in 
Year Book 65 (p. 73), which in turn 
parallels the previous proton-deu- 
teron phase-shift work of Christian 
and Gammel, 40 who assumed the po- 
larization of the interaction to be 
identically zero. The scattering by 
spin-zero nuclei is described with a 
two-by-two matrix (the M matrix) 
and by spin-one nuclei with a six-by- 
six matrix. The details of these calcu- 
lations are not suitable for this re- 
port, but the basic ideas are. The 
appalling complication of the calcula- 



tions — about nine times more labo- 
rious than for spin-zero nuclei — can 
be done by the computer, leaving* the 
physicist free to study the effect that 
altering one or more phase shifts has 
on experimentally measurable quanti- 
ties. These calculations are exact in 
that they assume only the validity of 
quantum mechanics and the short 
range of nuclear forces. The accuracy 
with which one determines the phases 
depends on the accuracy of the ex- 

The phase shifts may be considered 
to be a convenient form of nature's 
parameters, and they have a simpler 
relationship to nuclear forces than 
other quantities, frequently in a form 

TABLE 5. Set of Phases for S, P, and D Waves 

*- 2 S,/ 2 


4 S ,/,-- 


2 P 


4 P,/2 
4 P 3 /2 



- 4 D 





4 D , 

sin Sz = 

/ 00 

fc /o£/(r)^^ 1 (r)'^; i (r)dr 




where Si is the phase shift, k the 
wave number, U(r) the potential de- 
scribing the nuclear force, and Ui (r) 
the incident and Vi(r) the scattered 
wave functions. The phase shifts 
have their simplest form in the scat- 
tering of spin-zero particles, e.g. 
alpha particles, by nuclei of spin-zero. 
A single phase shift is required for 
each partial wave; the familiar 
spectroscopic notation of S, P, D, etc., 
denotes orbital angular momenta of 
I = 0, 1, 2. . . . The scattering of 
protons by nuclei of spin zero re- 
quires phases above S to be split ac- 
cording to the possible values of the 
total angular momentum /. They then 
become S, P 1/2 , P 3/2 , D 3/2 , D 5/2 , etc. 
For the scattering of protons by 
nuclei of spin one the set of phases 
given in Table 5 is necessary for S, 
P, and D waves. The superscript de- 
notes the multiplicity, i.e. the maxi- 
mum number of substates possible, 
which is here only quadruplet or 
doublet, and the subscript denotes the 
total angular momentum. Mixing be- 
tween certain states is allowed (a 
phenomenon forbidden for spin-zero 

target nuclei), as denoted by the 
arrows. The analysis of Christian and 
Gammel has neither splitting of quad- 
ruplet or doublet phases nor mixing 
parameters; they describe the inter- 
action with only 2 S, 4 S, 2 P, 4 P, 2 D, *D, 

The observed effects of polariza- 
tion are small at low energies, sug- 
gesting that they may be produced by 
small changes in the phases of Chris- 
tian and Gammel. We can in fact 
explain the polarization measure- 
ments in this manner. The polariza- 
tion resulting from a splitting of the 
4 P phases by 1.4° is shown in Fig. 41 
together with measured values. An 
equally good fit can also be had with 
the parameters mixing quadruplet 
and doublet states, but this predicts 
a distribution for polarized deuterons 
scattered by protons that disagrees 
in sign with measurements made 
at the University of Wisconsin by 
P. Extermann, 41 whereas quadruplet 
splitting is in good agreement. No 
other simple variation of splitting or 
mixing parameters gives a reason- 
able fit to experiment. 










D (p , p) D 
E p = 3.00 MeV 



Fig. 41. Polarization as a function of scattering angle for D[p,p)D at laboratory energy 
of 3.00 MeV. The full curve shows the polarization resulting from splitting the 4 P-phases by 1.2°, 
viz. 4 Pi 2 = 21. 6°, A ?z/2 = 22. 8°, and 4 P$/2 = 24.0°; the other phases, which are un- 
split, have values from Christian and Gammel. 40 

Mixing between S and D states can- 
not be determined by simple polari- 
zation experiments, but can be found 
by measuring the change in polari- 
zation of a polarized beam by scatter- 
ing. The experiment should be accom- 
plished this summer and fall. It 
requires two successive scatterings of 
the protons — the first by deuterium, 
and the second by helium. Helium 
analyzes the polarization of the scat- 
tered protons and thereby determines 
the change in polarization caused by 
the first scattering. Preliminary tests 
of equipment have been made for this 


L. Brown, W. K. Ford, Jr., V. C. Rubin, 
and W. Trachslin 

Atomic and Ionic Spectroscopy 
with Foil Excitation 

Our interest in foil excitation was 
reported in Year Book 65 (pp. 76- 

77) with a description of the impor- 
tant ideas behind the method. We 
have since tested the apparatus by 
observing the foil-excited spectrum 
of sodium that had been accelerated 
to a speed of 2.75 X 10 8 cm/sec. The 
experiment was successful, if evalu- 
ated by the amount of information 
obtained and the performance of the 
equipment. Seventy-seven spectral 
lines were observed whose wave- 
lengths could be measured with an 
accuracy of about 0.5 A. The in- 
crease in quantity and quality from 
earlier experiments comes from the 
speed of the image tube spectrograph. 
The results puzzle us, however, in 
that only 5 or 6 of these lines can yet 
be assigned to known transitions in 
sodium. Because of this limitation the 
present goal of the work — the meas- 
urement of lifetimes of excited states 
for use in determining abundances of 
the elements in astronomical objects 



— may need modification. Whether 
this goal can be reached and whether 
there are other uses for these exper- 
iments will not be clear for a while. 
In the experiment a beam of so- 
dium ions passes through a very thin 
carbon foil. On emerging from the 
foil an atom of sodium may be in any 
charge state from neutral to com- 
pletely ionized and, if excited, will 
radiate under conditions approxi- 
mating those of a free atom. The 
beam glows after passing through the 
foil, and its intensity decreases down- 
stream from the foil. A lens images 
the glowing beam onto the slit of a 
spectrograph containing an image 
tube, as shown in Fig. 42. In our 
experiments the spectral range 3900 
A to 6700 A is observed with a dis- 
persion of 22 A/mm. The spectrum 
so obtained differs from conventional 
laboratory spectra in two ways. The 
lines, each of which is a picture of 
the beam in a single wavelength, de- 
crease in intensity at different rates 
with increasing distance from the 
foil; the lines are not perpendicular 
to the base line, but slant. Both these 
characteristics are illustrated in 

Plate 5, a portion of the sodium 

The light incident on a given point 
A in the line on the plate originates 
at a specific point B in the beam, and, 
since the light intensity of the beam 
decreases according to the distance 
from the foil, the variation will be 
recorded as an apparent wedge- 
shaped spectral line. The slant of the 
spectral lines results from Doppler 
shift. The light will be shifted in 
wavelength depending on the velocity 
of the radiating atom relative to the 
lens. The changing angle between the 
direction of the beam and the direc- 
tion to the lens causes a different 
Doppler shift at each point along the 
line. This effect is used to determine 
the charge state of the ions in a pro- 
cedure invented by Trachslin. An 
electric field of 15 kV/cm is applied 
perpendicular to the direction of the 
beam and parallel to the line of sight. 
Neutral atoms are unaffected by it, 
but ions are accelerated in the direc- 
tion of the field to a speed propor- 
tional to their charge and the dis- 
tance from the foil. This velocity 
component, which is perpendicular to 

Image of 
beam at 
graph slit 

Fig. 42. The imaging (not to scale) of the glowing sodium beam onto the spectrograph slit. 
Light passing from point 8 in the beam to point A in the plate is shifted in wavelength because 
the radiating atom at B has a velocity component away from the lens. Because of the finite 
aperture of the lens, the light at A results from a range of angles A#, and hence of relative 
velocities. The width of the lines in the recorded spectrum is approximately proportional to the 
diameter of the lens. 



the original direction of the beam, is and hence the slant of the line on the 

small compared with the total speed, plate, enough to identify the charge 

but. since it is in the direction of the state, 
line of sight, it shifts the wavelength, 


E. T. Bolton, D. J. Brenner, R. J. Britten, D. B. Cowie, S. Falkow, 
D. E. Kohnc, A. Rake, and R. B. Roberts 

A central theme runs through this 
report, as it has in past years: the 
enormous potential for understanding 
the relationships and history of crea- 
tures through the similarity of nu- 
cleotide sequences in their DNA. 
Some of the work — studies of brain 
function and memory — does indeed 
lie outside of areas guided by this 
theme. Even there, measurement of 
gene expression by means of messen- 
ger RNA homology to DNA has a 
plausible role. A second theme is also 
present, derived from the measure- 
ment of the temperature at which 
nucleic acid strand pairs dissociate. 
This method has been used both for 
the classification of DNA and as an 
index of the degree of divergence in 
viral, bacterial, and animal systems. 

As a subject moves from the period 
of discovery toward development a 
vocabulary grows with it, usually 
confused in its early stages. Nucleic 
acid sequence pairing is no exception. 
The following glossary is an attempt 
to relieve some of the growing pains. 
It has several purposes: One is to 
supply an introduction to the non- 
specialist; a second is to present some 
personal views on the usages of cer- 
tain words. A few current results are 
summarized, and there is, naturally, 
an attempt to clarify the meaning of 
some words. It is hoped that those 
who feel strongly will criticize con- 
structively and will correct as well 
as contribute additional words or 
phrases, with definitions. 

Apologies are offered for stylistic 
license, for theft of words from 
other specialties, and for the narrow- 

ing down of meanings and usages. 
We proceed on the ground of at- 
tempted clarity. 

R. J. Britten 

binding — 1. The reassociation of DNA 
fragments with immobilized DNA. 
2. The adsorption of nucleic acids on 
hydroxyapatite or cellulose nitrate fil- 

cation concentration — The concentra- 
tion of cations determines the dimen- 
sions of the charge cloud around the 
negatively charged nucleic-acid mole- 
cules and the repulsive Coulomb forces. 
The cation concentration and the tem- 
perature are used to control the con- 
formation or secondary structure of 
nucleic-acid molecules. Symbolized by 
k representing the molarity of mono- 
valent cations. Divalent ions are im- 
mensely more effective and must be 
considered separately. 

Cerenkov counting — P 32 is assayed in a 
scintillation counter at high gain by 
the light emitted by passage of j3 par- 
ticles through water. Samples can be 
counted without drying, they are not 
adulterated, and valuable materials 
can be reprocessed. 

complementarity — The fraction of a set 
of nucleic acid strands which will pair 
with a second set, at a given criterion. 
(Apologies to Niels Bohr.) See crite- 
rion ; pair. 

complementary — Said of a pair of nu- 
cleic acid strands when all the nucleo- 
tides in sequence can be paired in the 
Watson-Crick sense — A:T and G:C. 
Partially or imperfectly complemen- 
tary are acceptable usages. See strand. 

concatenation — The formation of chains 
of fragments by the reassociation of 
DNA strands that typically terminate 



at various places in the nucleotide se- 
quence. Thus reassociated pairs termi- 
nate in two single-stranded ends that 
can reassociate with other complemen- 
tary fragments to continue the chain. 
This situation occurs for all DNA ex- 
cept whole linear molecules from a 
nonpermuted virus. See network; re- 

C t (kot) — The product of the con- 
centration of DNA or RNA and the 
time of incubation (moles nucleotides 
X seconds per liter). The effectiveness 
of the C t in inducing reassociation 
depends on the genome size, the pres- 
ence of repeated sequences, the frag- 
ment size, the cation concentration, 
and the temperature of incubation. See 
cation concentration. 

criterion — The standard of precision of 
pairing set by the cation concentration 
in combination with the temperature 
during incubation. Now expressed in 
terms of these two parameters be- 
cause the fraction of unpaired nucleo- 
tides implied is not known. See im- 
precise; precise. 

CsCl density — In an ultracentrifugally 
established equilibrium density gradi- 
ent, the density of the cesium chloride 
solution at the position of a band of 
nucleic acid. Aside from minor effects 
of temperature and pressure it is con- 
trolled by the GC content of DNA and 
by its secondary structure. Single- and 
double-stranded DNA are separable. 

denature — This word should be re- 
stricted to the disruption of secondary 
structure whether of paired or single- 
stranded nucleic acids. Local or tem- 
porary denaturation may occur, and 
the process may be reversed. See sec- 
ondary structure; dissociate. 

discrimination — The capability of a 
measurement of pair formation to dis- 
tinguish between nucleic acids from 
two species. This index of their re- 
latedness is dependent on the criterion. 
See criterion. 

dissociate — To separate completely the 
two strands of a pair so that one may 
diffuse away from the other. See de- 

divergence — 1. The increasing degree 
of mismatch among a set of nucleic 
acid sequences that occurs on an evo- 

lutionary time scale ; example : The 
divergence of the DNA of Salmonella 
and E. coli has lowered the average 
melting temperature of interspecies 
pairs compared with that of intra- 
species pairs. 2. The act of speciation 
which leads to two distinct species 
lines; example: The paleontological 
record indicates that about 300 million 
years have elapsed since divergence of 
the lines leading to the amphibians 
and the reptiles. 

DNA-agar — A system in which large- 
fragment-sized single-stranded DNA 
is mixed with melted agar and rapidly 
cooled. The DNA is immobilized. At 
reasonable criterion sheared fragments 
will diffuse through the agar and re- 
associate with the immobilized DNA. 
See criterion. 

double stranded — Indicates that two 
strands are held together by means of 
complementary base pairs. May refer 
to local or imperfect regions. 

duplex — A base-paired structure be- 
tween two complementary or partially 
complementary nucleic acid strands. A 
nucleic acid strand pair. 

ERRF (urf) — An acronym used origi- 
nally to describe extremely rapidly re- 
associating fractions of higher-orga- 
nism DNA. Now it is used to describe 
the total fraction of DNA that behaves 
as repeated sequences at a given cri- 
terion. Here we risk this convenient 
"word" for the first time in print. In 
the jargon, one says "slow, fast, pre- 
cise, or imprecise ERRF" and there 
are other such unsavory but accurate 
usages. See slow fraction. 

family of repeated sequences — The set 
of related sequences in the genome of 
a given higher organism which will 
reassociate with one another. The con- 
cept of family is useful. At the mo- 
ment, however, different families with 
equal numbers of members and equal 
degrees of divergence are not resolv- 
able. See higher organism; related. 

filter methods — DNA is immobilized 
by binding to cellulose nitrate mem- 
brane filters. Hybridization with RNA 
may be conveniently assayed in this 
way. In a recent modification further 
adsorption of DNA is blocked by treat- 
ment and assay of reassociation of 



fragments with the immobilized DNA 
may be made. 

fragment — A length of nucleotide se- 
quence broken out of a longer mole- 
cule by backbone strand scission. See 

GC content — The fraction of the base 
pairs in double-stranded DNA which 
is made up of guanine and cytosine 
(GC) as opposed to adenine and thy- 
mine (AT). 

genome size — 1. The haploid DNA con- 
tent ^measured by chemical assay) of 
a cell or virus particle. In rapidly 
growing cells partially completed new 
chromosome sets are not to be counted. 
2. The amount of DNA per typical 
unique sequence in the genome (meas- 
ured by rate of reassociation). The 
comparison of these results indicates 
that chromosome pairs of a diploid cell 
are nearly identical in sequence. It is 
open season for difficulties with these 
definitions, as in sex chromosomes; 
polyploid plant hybrids ; significantly 
different genes in a diploid heterozy- 

higher organism — Here used in a some- 
what vague sense to include all orga- 
nisms with highly repetitive DNA se- 
quences. To date, only viruses, bac- 
teria, and a blue-green alga have been 
found not to contain ERRF. It thus 
appears possible that the line between 
organisms that do and those that do 
not contain ERRF will follow that 
between eukaryotic and prokaryotic 
creatures. See ERRF. 

homologous, heterologous — Homologous 
has been used to represent DNAs of 
the same species, and heterologous, to 
represent DNAs of a different species, 
when they are added to a system con- 
taining immobilized DNAs, as in: 
Sheared heterologous carrier and 
homologous tracer DNA were added 
to DNA-agar. This usage of homolo- 
gous is bad, because of confusion with 

homology — The degree of similarity 
between the nucleic acid sequences of 
different species, as in : The homology 
between two species is measured by 
the capability of their DNA to form 
interspecies pairs at a given criterion 
and by the thermal stability of the re- 

sulting pairs. Numerical specification 
is difficult or impossible because more 
than one parameter enters. 

hybrid — 1. A nucleic acid strand pair 
between RNA and DNA. 2. Pairs made 
from DNA strands from two species. 
The second usage should be avoided. 

hydroxyapatite — Calcium phosphate 
crystals produced in a controlled pre- 
cipitation. It has the extraordinary 
property of adsorbing double-stranded 
but not single-stranded DNA at prop- 
erly controlled salt and temperature 
conditions. Imperfect strand pairs of 
DNA bind well. 

hyperchromicity — The increase in ultra- 
violet absorbancy as the secondary 
structure is disrupted. Best expressed 
as percentage of the maximum ab- 
sorbancy when the process is essen- 
tially complete. 

hypochromicity — The fall in ultraviolet 
absorbancy as secondary structure re- 
turns after its disruption. Best ex- 
pressed as percentage of the maximum 

identical — Proposed substitute for 
homologous; example: A competition 
experiment with DNA-agar requires 
unlabeled sheared fragments of DNA 
from a competing species and labeled 
sheared fragments identical to the 
DNA immobilized in the agar. See 

imprecise — Imprecise pairing between 
nucleic acids implies that a certain 
fraction of the nucleotides is not 
complementary to those on the opposite 
strand, thus interrupting the second- 
ary structure and reducing the thermal 
stability. As a round number, a 1% 
mismatch lowers the melting tempera- 
ture 1°C. See melting temperature; 
mismatch ; secondary structure ; ther- 
mal stability. 

incubation— The maintenance of sam- 
ples of nucleic acid at elevated temper- 
atures and controlled cation concentra- 
tion in order to promote reassociation. 

intrasequence heterogeneity — The vari- 
ation of some parameter such as local 
GC content along the length of a 
nucleic acid sequence. The implication 
is that averages taken over a series of 
short lengths will show an interesting 
distribution, such as intermixed blocks 



of high and low values of the param- 
eter. Of interest are parameters such 
as interspecies homology, purine/ py- 
rimidine ratio, and membership in re- 
peated families. See GC content. 

kinetics — Shorthand for the time 
course of a process, as in : The re- 
association kinetics measures the con- 
centration of complementary sequences. 

log C t plot — A method for presenta- 
tion of reassociation kinetics in which 
the fraction of DNA reassociated is 
plotted against log C t. It has the ad- 
vantage that measurements over a 
wide range of concentrations and times 
can be combined to give a more com- 
plete picture of the reassociation of 
higher-organism DNA. See kinetics. 

melting curve — The result of an assay 
of the degree of dissociation or second- 
ary structure disruption as a function 
of temperature. Assay may be the in- 
crease in ultraviolet absorbancy at one 
or several wavelengths, displacement 
of strands from immobilized DNA or 
displacement from hydroxyapatite. See 

melting temperature — The temperature 
at which half of the total change in a 
melting curve has occurred. 

mismatch — The fraction of nucleotides 
in a strand pair which is not comple- 
mentary to the nucleotides on the op- 
posite strand. Avoid usage that implies 
complete lack of complementary se- 

native — Signifies for DNA that the 
Watson-Crick helix present in vivo 
remains in a preparation. There may 
be strand scissions and the DNA may 
be fragmented but the strands have 
not been separated. May be applied to 
other nucleic acids with increasing in- 
definiteness, as the secondary structure 
is less well known in vivo and less 
stable in preparations. 

network — A multibranched structure of 
reassociated DNA, typically seen with 
DNA from higher organisms. As a 
result of past translocations of re- 
peated sequences, paired regions can 
terminate in 4 single-stranded ends. 
These ends reassociate with those of 
other pairs and form branched struc- 
tures. Branching could also occur with 
a mixture of DNA strands in which 

local regions have diverged to a suffi- 
cient extent to create unpaired regions 
that can reassociate with other strands. 
See concatenation ; divergence. 

nonrepeated sequences — Nucleotide se- 
quences that occur only once in the 
(haploid) DNA complement of an or- 
ganism, as might have been expected 
for most moderate length DNA se- 
quences before repeated sequences be- 
came known. See nucleotide sequence. 

NTP — Abbreviation for nucleotide 
pairs used in stating the length of 
double-stranded DNA. 

nucleotide sequence — A specifically 
ordered row of nucleotides in the DNA 
(or RNA) of an organism. In the 
DNA, the complementary sequence is 
always present (except for single- 
stranded viruses) and the two are 
often considered together, as in: The 
rate of reassociation is a measure of 
the degree of repetition of a nucleo- 
tide sequence. See complementary. 

order of reaction — The reassociation of 
complementary strands is controlled by 
a bimolecular collision and thus ini- 
tially follows second-order kinetics. 
Empirically, the reaction actually fol- 
lows the second-order equation until it 
is nearly completed. Thus the fraction 
of unpaired DNA is 1/(1 + K C t) 
where K is a reaction rate constant 
controlled by the variables listed 
under C t. In some cases one class of 
nucleic acid is present in a minority 
and the major component is immobi- 
lized so that the number of sites avail- 
able for pairing remains constant. 
Then a pseudo first-order reaction is 
observed and the fraction of the 
minor component which remains un- 
paired is exp ( — K'C' t). K' is numer- 
ically equal to K if C' t is the effective 
concentration of the major immobilized 
component. See C t. 

pair — To form a base-paired structure 
between two nucleic acid strands be- 
cause of complementary or partially 
complementary nucleotide sequences. 
To reassociate. Also the product of 
this process. 

precise — Precise pairing between 
strands implies that virtually all of the 
bases are complementary and the ther- 
mal stability will be close to that of 



native DXA if the fragments are long 
enough (>200 nucleotide pairs). 

reanneal — To incubate DXA in order 
to induce reassociation, originally over 
a range of temperatures since the op- 
timum temperature was not known. 
S e incubation; reassociate. 

reassociate — To incubate nucleic acids 
under conditions such that dissociated 
complementary strands may collide 
with each other and form base-paired 
double strands. Also the act of form- 
ing such a double strand. Reassociation 
does not imply that the particular 
strands were paired in vivo. The opti- 
mum temperature is 20° to 30° below 
the melting temperature and the rate 
of the reaction increases with ionic 
strength. Sec C t; dissociate; incuba- 

registration — When a complementary 
pair of DXA strands slide (metaphor- 
ically speaking) one along the other, 
a large number of base pairs are com- 
plementary in only one registration. 
The strands are then in register and 
can pair. Homopolymer pairs have as 
many registrations as nucleotides. 
DXA fragments containing several 
members of one family of repeated 
sequences can have several registra- 
tions that lead to reassociation. See 
family of repeated sequences. 

related — Signifies that nucleic acid 
fragments contain complementary nu- 
cleotide sequences and thus will pair. 
It is presumed that complementary 
sequences of any significant length — 
greater than, say, 20 or 25 pairs — 
cannot arise independently. Thus re- 
latedness in this sense implies com- 
mon origin. The boundary between 
relationship and chance similarity is 
not yet clearly defined experimentally. 

renature — Originally and widely used 
to describe complementary double- 
strand formation from single strands. 
The implication that a nativelike struc- 
ture is re-formed is broadly inappli- 
cable. Pairs commonly form between 
partially complementary strands, as in 
the families of repeated sequences of 
higher organisms, or with DNA from 
different species. 

repeated sequence — A segment of 
nucleotide sequence which occurs many 

times in the DNA of one cell of a 
higher organism. A coined word would 
be preferable here, since precise repe- 
tition appears to be the exception 
within the large, greatly divergent 
families of repeated sequences. 

repetition frequency spectrogram — A 
representation of the quantity of DNA 
as a function of its frequency of nu- 
cleotide sequence repetition. Usually 
graphed against the log of the fre- 
quency, since the range is greater 
than 10 6 and the resolution, as yet, is 
not much better than a factor of 10. 

saltatory replications — The hypotheti- 
cal events by which families of hun- 
dreds of thousands of similar nucleo- 
tide sequences are produced in the 
DNA of an organism. Large, closely 
similar (young), as well as greatly 
divergent (old), families are observed. 
Thus families are produced in a time 
short compared to the time required 
for their loss by divergence (a few 
hundred million years). 

satellite DNA — A minor component re- 
solved from the principal band of DNA 
in a CsCl equilibrium density gradient. 
Firmly engraved on the jargon by a 
coincidence in time with Sputnik. 

secondary structure — The intramolecu- 
lar forces in addition to the phospho- 
diester backbone bonds establish a 
more or less stable conformation of a 
nucleic-acid molecule in solution. The 
resulting set of relationships of the 
parts is bagged together as the sec- 
ondary structure. It is thought to be 
determined principally by hydrogen 
bonds and hydrophobic forces. In the 
case of double-stranded DNA the rela- 
tively simple and stable Watson-Crick 
helix is dominant while for single- 
stranded molecules, less stable, more 
transient (and less understood), rela- 
tionships predominate. 

segment — A length of nucleotide se- 
quence distinct in some way from 
neighboring stretches, e.g. a member 
of a repeated family. Usually not 
physically terminated by phosphodi- 
ester backbone scission. See fragment. 

sequence similarity by chance — A given 
short sequence may recur in the DNA 
of an organism, even though not of 
common origin, simply because there 



is a limited number of permutations 
of a sequence. Any sequence n long 
will be expected to occur once in a 
random sequence 4 TO long. Imperfect 
repetitions must be considered and it 
is not known that DNA sequences even 
approach randomness. However, short 
sequences have reduced thermal stabil- 
ity, and experimental test indicates 
that with a reasonable criterion (50°, 
k = 0.18) such sequence similarity is 
undetectable in higher organism or 
bacterial DNA. See criterion. 

shearing — The process of fragmenta- 
tion of DNA by creating shear forces 
in a suspension. Controlled fragment 
sizes have been produced from 10 5 
NTP by stirring at a few hundred 
rpm down to 5 X 10 2 NTP by passing 
through an orifice at 50,000 psi. 

similar — Broadly used to indicate that 
two sequences are partially or imper- 
fectly complementary. 

slow fraction — A phrase of convenience 
to describe the fraction of the DNA 
which, at a given criterion, reassoci- 
ates at the slow rate expected for 
unique sequences, and which there- 
fore does not exhibit repeated se- 
quences. See unique. 

species specific — Strictly, this phrase 
refers to elements particular to a sin- 
gle species, such as the mouse satel- 
lite DNA. A phrase like species de- 
pendent applies better to typical DNA 
sequences since some homology is ex- 
hibited between the DNA of creatures 
even from different orders. See ho- 

strand — Conveniently used to represent 
fragments of single-stranded DNA or 

strand scission — The breaking of, or a 
break in, the phosphodiester backbone 
of a nucleic acid molecule. 

thermal chromatogram — The set of data 
resulting from the assay of thermally 
eluted fractions. See thermal elution. 

thermal elution — The process of dis- 
sociating strand pairs with steps of 
increasing temperature and eluting the 
product from hydroxyapatite, DNA- 
agar, or cellulose nitrate filters. 

thermal stability — A convenient phrase 
to indicate that strand pairs are dis- 
sociated at a particular temperature 

(with a given cation concentration;, 
as in : The thermal stability is influ- 
enced by the GC content and the 
precision of matching of a nucleotide 
sequence pair. 

translocation — The displacement of a 
segment of nucleotide sequence to 
another part of the chromosomal DNA 
through a series of genetic events. 

unique— Designates that a particular 
sequence occurs only once in a genome ; 
for example, the concentration of a 
unique sequence is known from the 
genome size and the total DNA con- 
centration. The typical situation for 
viral and bacterial DNA. See genome 
size ; slow fraction. 



R. J. Britten and D. Kohne 

Our knowledge of the repeated se- 
quences in DNA has expanded con- 
siderably since their discovery 3 
years ago (Year Book 6U) . We do not 
yet know the function or origin of 
these sequences, but feel certain that 
they must play a major role in evolu- 
tion as the generality of their occur- 
rence in large quantities in the 
genomes of higher organisms has be- 
come clear. 

In the past year the list of species 
containing repeated DNA has grown 
and no exceptions have been found 
among eukaryotic creatures. Individ- 
ual organisms (onion, salmon, am- 
phiuma) were examined which have 
a majority of their DNA in this form. 
Also progress was made toward an- 
swering the question whether or not 
saltatory replications have been the 
principal source of the DNA of 
higher organisms. 

Further studies were made of the 
degree of relationship among the 
families of repeated sequences within 
a given individual. It is now possible 
to estimate the average rate of di- 
vergence among the members of re- 
peated families. 

Although repeated DNA indeed ap- 



pears to occur universally among 
higher organisms, there is great va- 
riety in the quantity, the number of 
members in individual families, and 
the degree of divergence among the 
members. A survey of the repeated 
DXA of many species now appears 
necessary to learn the significance of 
this variety in evolutionary history. 

Considerable effort has been in- 
vested in increasing the certainty and 
reproducibility of our methods of 
measuring the rate of reassociation 
and the properties of the reassociated 
DXA molecules. 

Mouse satellite DNA. A 10% frac- 
tion of the DNA of the mouse — called 
the mouse satellite — has played an 
important role in the development of 
our study of repeated DNA se- 
quences. In the early summer of 1964 
the success of the DNA-agar system 
for the study of animal DNA sug- 
gested to Britten the existence of 
repeated DNA. As a result Britten 
and Waring (Year Book 6U) started 
studying mouse DNA and discovered 
in the fall of 1964 that the satellite 
reassociated very rapidly. It was in- 
tuitively recognized that this was an 
example of repetition. Subsequently, 
measurement of the actual rate of 
reassociation and its variation with 
DNA concentration showed that this 
was true, and that a short nucleotide 
sequence was repeated about a mil- 
lion times in the mouse DNA. Having 
this extreme example in our pocket, 
so to speak, made the more modest 
degrees of repetition (tens and hun- 
dreds of thousands) more believable 
and gave momentum to the studies 
which led to the recognition of their 
general occurrence. We did not, how- 
ever, choose mouse DNA because it 
contained the satellite. In fact, we 
were unaware of the existence of the 
satellite until it appeared in the ultra- 
centrifuge, although it had previously 
been observed and reported. 

We have not further studied the 
mouse satellite but some recent work 
must be mentioned because it has im- 
portant implications for repeated se- 
quences in general. Maio and Schild- 
kraut 42 have shown that the location 
of the mouse satellite DNA within 
the cell is not only nuclear but domi- 
nantly, if not entirely, chromosomal. 
This fact draws closer the relation- 
ship between this satellite and other 
repetitive DNA which must be chro- 
mosomal because its great quantity 
admits no other possibility. Peter 
Walker and Ann McLaren 43 in the 
winter of 1964-1965 had prepared 
with hydroxyapatite a fraction of 
mouse DNA which was subsequently 
shown to contain the satellite and 
probably other repetitive DNA. Re- 
cently Flamm, McCallum, and Wal- 
ker 44 showed that the satellite may be 
separated into two fractions by cen- 
trifugation in alkaline CsCl. These 
fractions do not reassociate by them- 
selves but do when mixed and are 
therefore complementary strands. 
Their composition is striking. Al- 
though both contain 34% guanine 
plus cytosine, one strand contains % 
of the pyrimidines and the other, % 
of the purines. 

Occurrence of repetitious DNA, 
Below is a list of those life forms 
which have been examined for 
repetitious DNA: (a) Rate of re- 
association measured directly by 
hydroxyapatite fractionation and/or 
measurement of optical hypochromic- 
ity as a function of time, (b) Labeled 
sheared fragments bind to DNA from 
the same species imbedded in agar 
at a C t so low that repetition must 
be present, (c) Sheared nonradio- 
active fragments of DNA from the 
listed organism compete with the 
DNA-agar reaction (b) of a related 
species, reducing the amount of la- 
beled DNA which binds to the im- 
bedded DNA. 




dinoflagellate (Gyrodinium Cohnii) (a) 

Euglena gracilis (a) 

sponge (Microciona) (a) 

sea anemone (Metridium) (tentacles) 

sea urchin (Strongylocentrotus) 
sperm (a) (b) (c) 

sea urchin (Arbacia) (sperm) (a) (b) 

starfish (Asterias) (gonads) (a) 

sand dollar (E chinarachnis) (c) 

crab (Cancer borealis) (gonads) (a) 

horseshoe crab (Limulus) (hepatopan- 
creas) (a) 

squid (Loligo pealii) (sperm) (a) 

dogfish shark (liver) (a) 

salmon (sperm) (a) (b) (c) 

lungfish (a) (b) 

amphiuma (liver, red blood cells, mus- 
cle) (a) 

frog (Rana pipiens) (b) 

frog (Rana sylvatica) (c) 

toad (Xenopus laevis) (heart, liver, 
red blood cell) 

axolotl (Amby stoma tigrinum) (c) 

salamander (Trituris viridescens) (c) 

chicken (liver, blood) (a) (b) (c) 

tree shrew (c) 

armadillo (c) 

hedgehog (c) 

guinea pig (c) 

rabbit '(c) 

rat (liver) (a) (b) (c) 

mouse (liver, brain, thymus, spleen, 
kidney) (a) (b) (c) 

hamster (c) 

calf (thymus, liver, kidney) (a) (b) 

tarsier (c) 

slow loris (c) 

potto (c) 

capuchin (c) 

galago (c) 

vervet (c) 

owl monkey (c) 
green monkey (c) 
gibbon (c) 
rhesus (b) (c) 
baboon (c) 
chimpanzee (a) (c) 
human (a) (b) (c) 

rye (Secale) (c) 

tobacco (Nicotiana glauca) (c) 

bean (Phaseolus vulgaris) (c) 

vetch (Vicia villosa) (c) 

barley (Hordeum vulgare) (a) (b) 

pea (Pisum sativum, var. Alaska) (a) 

wheat (Triticum aestivum) (a) (c) 
onion (Allium sp.) (a) 

This expanded list supports our 
conclusion (Year Book 65, p. 94) 
that repetitious DNA occurs in most 
organisms that are more complex 
than bacteria. DNA from viral, bac- 
terial (E. coli, Clostridium perfrin- 
gens), and blue-green algae genomes 
examined by our methods have re- 
vealed no repetitious DNA. No fungi 
or higher algae have been examined. 
These forms are intermediate be- 
tween bacteria and higher organisms 
and are therefore of great interest. 

Thus far the line between orga- 
nisms which do and those which do 
not contain repetitious DNA follows 
that between eukaryotic and prokar- 
yotic forms. Eukaryotes contain repe- 
titious DNA; prokaryotes do not. 
Future work will expand the list of 
organisms examined and should clar- 
ify the boundary between those life 
forms which do and those which do 
not possess repetitious DNA. 

The nonrepeated sequences of am- 
phiuma. Amphiuma, sometimes called 
the Congo eel, is a curious amphibian 
related to the mud puppy. It has been 
used as a laboratory specimen be- 
cause of its large cells. It contains 
the largest amount of DNA per hap- 
loid cell of any known creature. A 
haploid cell contains 8.6 X 10 10 nu- 
cleotide pairs, or about 30 times that 
of human sperm. 



We have made an attempt, using 
hydroxyapatite, to measure the rate 
of reassociation of the slow fraction 
of amphiuma DNA. The DNA was 
sheared, dissociated, incubated (C t 
= 1000, 50°, k = 0.21), and frac- 
tionated to remove ERRF. The slow 
fraction was incubated (60°, k = 
0.36) at a DXA concentration of 6 
nig ml, and samples taken for a 2- 
week period and assayed on hydroxy- 
apatite. A slow reassociation was ob- 
served in the fraction which eluted 
at high temperature. Only the initial 
part of the reassociation occurred in 
the 2-week period and the rate con- 
stant had the expected value for 
unique sequences in a genome of this 
size. Thus it appears that, while most 
of amphiuma DNA consists of re- 
peated sequences, a small fraction of 
even this enormous genome appears 
to consist of a set of unique se- 

Time Course of DNA Reassociation 

More complete curves tvith the use 
of hydroxyapatite. When the DNA of 
higher organisms reassociates, the 
first part of the reaction is due to 
collision of sequences that are present 
in perhaps a million more or less 
similar copies. Nonrepeated se- 
quences collide with their comple- 
ments much less often, and only very 
much later — a factor of perhaps one 
million in time — does the reassocia- 
tion reaction go to completion. Pre- 
viously (Year Book 65, p. 91) we 
presented the data separately for the 
fast and slow parts of this reaction. 
Different techniques were used be- 
cause of the widely separated reac- 
tion rates. When hydroxyapatite is 
used to assay the degree of reassoci- 
ation, it is possible to measure the 
complete time course under uniform 
conditions by carrying out the re- 
action at several widely different con- 

Reassociation of calf DNA. Figure 
43 shows the results of such a series 
of measurements for calf thymus 
DNA sheared at 50,000 psi. 

The measurements at the different 
concentrations are concordant. This 
establishes beyond doubt that the re- 
association process results from a bi- 
molecular collision. In turn, the early 
part of the reaction is rapid only be- 
cause of the high concentration (due 
to sequence repetition) of the react- 
ing species. 

In the case of calf DNA there is 
a clear separation between the ERRF 
and the slow fraction. At this cri- 
terion (60°C, K = 0.18) the rapid 
fraction is a somewhat heterogeneous 
collection with an average C t for 
half reaction of about 10~ 2 . Therefore 
the average frequency of repetition 
is about 100,000. The slow fraction 
is — again, at this criterion — a rela- 
tively homogeneous population with 
C t for half reaction about 700 times 
greater than that for E. coli DNA. 
To measure this ratio accurately, la- 
beled E. coli DNA (also sheared at 
50,000 psi) was present at low con- 
centration in the experiment done at 
the highest concentration of calf 
DNA. There will be further comment 
on this experiment in the section 
dealing with the similarity of the 
DNA of diploid chromosome pairs 
(P. 84). 

Figure 44 shows a less extensive 
set of measurements for salmon 
sperm DNA. The striking differences 
between these data and those for calf 
are in the salmon's high quantity of 
ERRF (about 80% of the DNA) and 
the broad range of its repetition fre- 
quency. The criterion for the salmon 
measurements is more relaxed at 50° 
than at 60°. Some measurements 
done at the lower criterion with calf 
show only 55% ERRF. Thus there 
is a real difference. 

The average repetition frequency 
of the families is much less in the 



C t (M x sec/liter) 

Fig. 43. The kinetics of reassociation of calf thymus DNA measured with hydroxyapatite. 
The DNA was sheared at 50,000 psi and incubated at 60° in 0.12 M PB. At various times 
samples were diluted, if necessary (in 0.12 M PB at 60°), and passed over a hydroxyapatite 
column at 60°. The DNA concentrations during the reaction were: triangle, 2; closed circle, 10; 
open circle, 600; closed triangle, 8600 micrograms per milliliter; plus sign, radioactively la- 
beled E. coli DNA at 43 y/ml present in the reaction containing calf thymus DNA at 8600 y/ml. 


0.1 1 10 

Cjjt (M x sec/liter) 



Fig. 44. The kinetics of reassociation of salmon sperm DNA measured with hydroxyapatite. 
The DNA was sheared at 50,000 psi and incubated at 50°C in 0.14 M PB. The samples were 
diluted into 0.14 M PB at 50°C, if necessary, and passed over hydroxyapatite at 50°C. The 
DNA concentrations during the incubation were: closed circle, 8; open circle, 1600 ju,g/ml. 



salmon than in the calf. Some se- 
quences are present only in a few 
thousand copies (C,t = 1). but a 
minority is extremely repetitive, re- 
acting at a Cot of les's than 0.01. The 
few plants that have been examined 
exhibit a pattern more closely re- 
sembling; the pattern of salmon DNA 
than that of calf DNA. 

The contrast between the calf and 
salmon ERRF suggests that repeated 
sequences result from rare events — 
perhaps immensely important to evo- 
lution but nevertheless stochastic. If 
the repeated sequences had a function 
in chromosome structure, for ex- 
ample, or in basic cell biochemistry 
(genes for a protein class needed in 
excess), one would expect great simi- 
larity in the pattern from species to 
species, at least in the vertebrates. 
Here it appears that a carefully 
planned survey is needed, and it 
might produce surprises. 

Precision of Sequence Matching 
and Fractionation 

A wide range of precision of 
matching occurs among reassociated 

pairs formed among the repeated se- 
quences of DNA from a given species. 
The divergence among the members 
of a family of repeated sequences re- 
sults from nucleotide sequence 
changes that occur during the evolu- 
tion of the species. The degree of 
divergence can be taken to be a meas- 
ure of the age of the family and is 
of interest in relation to the history 
of the evolution of species, and evolu- 
tionary processes in general. There- 
fore, we have explored hydroxyapa- 
tite fractionation of repeated se- 
quences on the basis of their thermal 

Hydroxy apatite fractionation. 
When the temperature of a hydroxy- 
apatite column is raised, adsorbed 
double-stranded DNA dissociates and 
is eluted in the single-stranded form. 
For precisely matched sequences the 
temperature of dissociation measured 
in this way is quite close to that 
measured by optical hypochromicity 

(Year Book 65) . Figure 45 shows the 
results of fractionation of reassoci- 
ated salmon sperm DNA. Sheared 

(50,000 psi) DNA was dissociated 









' 1 ' 1 

1 ' 1 - 


/ \ /H 

/ \ 

/ \ / 

\ / \ 

/ \ ' 

V v 

/ \ 1 


1 \ / 

^ t 

/ \ / 

/ \ + 

a /q 

/ \ \ 

+— \ 

/ \ v 
/ \ \ 

/ \ 

/ \ \ 

/ \ 

\/ \ \ 


X A \ < 

/ ^ 

r \ St ^s^ x \ + 

/ , "•"^ 

l , 1 T^-l 

f 1 t 1 







Elution temperature, °C 

Fig. 45. Hydroxyapatite thermal fractionation of reassociated salmon sperm DNA. DNA sheared 
at 50,000 psi v/as incubated at 50° in 0.14 M PB [C Q t = 370) and passed over hydroxyapatite 
at 50' in 0.14 M PB. The adsorbed DNA was eluted by exhaustive washing (0.14 M PB) at 
each temperature (plus signs). Four fractions [65° , 70° , 85°, 90°) were separately dena- 
tured [100°, 5 min) and reincubated (50°, 0.14 M PB, C t about 10) and readsorbed on 
hydroxyapatite under the same conditions. Two of these were again thermally eluted from a 
column: open circles, 65° fraction; closed circles, 85° fraction. The other two were eluted with 
0.4 M PB and melted in the spectrophotometer as shown in Fig. 46. The fraction bound and the 
details of the shape of the elution curves depend on the degree of incubation. 



and incubated (criterion 50° C, k = 
0.21) for a C t of 270, and adsorbed 
on hydroxyapatite. The resulting 
thermal chromatogram (Fig. 45, 
dashed line) shows the expected 
broad range of thermal stability. To 
establish the specificity of the frac- 
tionation, samples eluted at 65° and 
85 °C were reincubated. They were 
then readsorbed and reanalyzed in 
the same way. Specificity of the frac- 
tionation is shown although the peaks 
are quite broad. 

The strand pairs formed during the 
second incubation are not necessarily 
the ones that were originally eluted. 
Instead they are almost certainly new 
duplexes formed by randomly as- 
sorted pairings among this selected 
set of strands. In each case, however, 
the same average degree of precision 
of relationship results. The 65° cut 
peaks again at 65°,. and the 85° cut, 
at 85°. The degrees of divergence are 
thus characteristic of these popula- 
tions. The members of the families 
presumably have been randomly mu- 
tated and all differ from each other 
to about the same extent. Similar 
studies have been done with calf 
thymus DNA with entirely compa- 
rable results. In addition, experi- 
ments with labeled calf DNA frac- 
tions indicate that very little sequence 
homology exists between precisely 
and imprecisely reassociating sets of 
repetitive DNA. Thus these are ap- 
parently quite distinct young and old 
sets of families. It seems reasonable 
to think of families being formed and 
then diverging. A model (Plate 6) 
schematizing this view appears at the 
end of this report. 

Length of repeated sequences. Are 
reassociated pairs complementary 
only in short regions or are they 
complementary over most of their 
length? The thermal stability of a 
pair does not by itself answer this 
question, since it appears that com- 
plementary sequences that are 100 

nucleotide pairs long will have a 
thermal stability approaching that of 
very long complementary sequences. 
However, the ultraviolet hyperchro- 
micity is a measure of the extent of 
sequence matching. Therefore, the 
hyperchromicity of a fraction known 
to contain only strand pairs is a 
measure of the average length of the 
complementary regions present. Re- 
sults for two such preparations are 
shown in Fig. 46. Native, completely 
complementary salmon DNA has a 
hyperchromicity of about 0.25 (cal- 
culated as 1.00 — 0.75 on the scale 
of the figure) . Single-stranded DNA 
has a hyperchromicity of 0.06 and 
melts mostly at lower temperatures, 
as shown by the upper curve. 

The 70° and 90° fractions each 
have a little better than half the hy- 
perchromicity of native DNA. From 
this we may conclude in each case 
that the potentially complementary 
regions are as long as half the frag- 
ments and perhaps nearly as long as 
the whole fragments. 



70 80 

Temperature, °C 

Fig. 46. Optical melting curves in 0.14 M 
PB of fractions of salmon sperm DNA. Frac- 
tions prepared as described in Fig. 45: closed 
circles, fraction eluted at 90°; open circles, 
fraction eluted at 70°. The upper curve (tri- 
angles) is for DNA which did not bind to hy- 
droxyapatite (50°, 0.14 M PB) in the first incu- 
bation and is therefore purely single stranded. 



Several complicating factors inter- 
fere with drawing a more firm con- 
clusion. Reassociated sheared frag- 
ments are bound to have single- 
stranded regions, since the two 
strands do not ordinarily terminate 
at the same points in the sequence. 
All degrees of overlap will occur, and 
for first collisions (no concatenation) 
the expected hyperchromicity for 
perfect pairing is perhaps only % 
that for native DNA. We do not know 
the extent of concatenation in these 
preparations. Finally, of course, the 
sequences in these families have di- 
verged from each other, and the un- 
matched nucleotides occurring within 
the paired sequences reduce the hy- 

These measurements are corrobo- 
rated by the hyperchromicity of re- 
associated ERRF from many animals, 
which is commonly between 0.16 and 
0.20. A few CsCl density measure- 
ments show a marked decrease in 
density upon reassociation, which 
also implies a good extent of comple- 
mentary pairing. 

It appears then that on the average 
repeated sequences are not extremely 
short (not less than 300 nucleotides) 
and may be much longer than our 
fragments, which average perhaps 
500 nucleotides. One imagines that a 
wide distribution of lengths will be 
present after a family diverges and 
that translocations as well as nucleo- 
tide substitutions have occurred. 

Rate of Divergence of Nucleotide 
Sequences during Evolution 

Some studies on the divergence of 
nucleotide sequences during verte- 
brate evolution were reported in 
Year Book 68 (p. 394). This year's 
investigations on repeated nucleotide 
sequences throw additional light on 
these studies. 

General ehar act eristics of repeated 
DXA from individual organisms. Re- 
peated nucleotide sequences comprise 

a large fraction of the DNA of all 
higher organisms thus far examined 
(see list, p. 73) . The repetitious frac- 
tion of the DNA is made up of many 
families of repeated nucleotide se- 
quences. A family of repeated nucleo- 
tide sequences can only arise from 
the manyfold duplication of an exist- 
ing nucleotide sequence. Current evi- 
dence suggests that the production 
of a family is a saltatory event and 
occurs relatively rapidly. (These sal- 
tatory events will be discussed in a 
later section.) At the time of a salta- 
tion all of the members of a family 
are identical. With the passage of time 
and the occurrence of mutations the 
members of a family diverge from 
one another to produce a family of 
similar but not identical sequences. 
The degree of similarity of the mem- 
bers of a family can be used as an 
index of the age of that family : the 
less similar the family members, the 
older the family. 

The size of the fraction of DNA 
which behaves as repetitious DNA 
varies with the criterion set by the 
conditions of the experiments for de- 
tecting such DNA. At a constant salt 
concentration, higher temperatures 
of incubation require a greater preci- 
sion of base-pair matching between 
the interacting sequences in order 
that a stable duplex be formed. Only 
families with relatively similar mem- 
ber sequences can form the more pre- 
cisely paired duplexes needed for 
stability at the higher temperatures. 
At lower incubation temperatures 
duplexes with more base-pair mis- 
matching are stable and therefore 
family members that are less similar 
can form duplexes. Data of Martin 
and Hoyer 45 obtained by the DNA- 
agar technique illustrate this point 
well (Fig. 47). As the temperature 
of incubation is lowered (at a con- 
stant salt concentration) there is a 
rise in the quantity of mouse DNA 
fragments which reacts with the 



40 50 60 70 

Temperature of incubation, °C 

Fig. 47. These data, from Martin and 
Hoyer, 45 were obtained with the DNA-agar 
technique by reacting chicken, rhesus, and 
mouse DNA fragments with mouse DNA-agar 
in 2 X SSC at different incubation tempera- 
tures. Triangles, mouse DNA fragments on 
mouse DNA-agar; squares, rhesus monkey DNA 
fragments on mouse DNA-agar; circles, chicken 
DNA fragments on mouse DNA-agar. 

mouse DNA-agar.* Reassociation 
rate data clearly demonstrate that the 
standard DNA-agar method meas- 
ures only the repeated sequences of 
animal DNAs (Year Book 65, p. 94). 
Since only repeated DNA was meas- 
ured, it is clear that the increased 

* The significance of the quantity of bind- 
ing at 40 °C is not known. Random inter- 
actions of nucleotide sequences may occur at 
this temperature and influence the extent of 

binding at lower temperatures was 
due to an increase in the number of 
families whose members were able 
to interact to form stable duplexes. 
Lowering the incubation temperature 
to examine families with members 
that are less and less similar is anal- 
ogous to looking back in time. These 
families represent a record of histor- 
ical events that have contributed to 
the composition of the present-day 
animal genome. 

Related nucleotide sequences in dif- 
ferent animals. Figure 47 also pre- 
sents data from Martin and Hoyer 45 
concerning the interaction of chicken 
and rhesus monkey DNA fragments 
with mouse DNA-agar. The extent of 
observed DNA interaction between 
two related animal DNAs varies with 
the temperature of incubation. The 
higher the temperature of incubation, 
the lower the measured relatedness 
between the two animal DNAs. Since 
the standard DNA-agar technique 
measures only the repeated DNA se- 
quences, the variation in the meas- 
ured relatedness with temperature 
can only be due to changes in the 
number of families held in common 
which can form stable duplexes. Fur- 
thermore, these studies indicate that 
a large fraction of the present-day 
chicken and rhesus DNA existed as 
repeated DNA sequences at the time 
the respective species lines diverged 
hundreds of millions of years ago. 

The lower average thermal stabil- 
ity of the families held in common 
between two species (Fig. 47), as 
compared to the thermal stability of 
families within a species, suggests 
that family formation has occurred 
many times since the divergence of 
these species. 

Loss of DNA relatedness during 
vertebrate evolution. Measurements 
reported previously (Year Book 63, 
p. 394) and the more recent measure- 
ments of Martin and Hoyer 45 share 
a surprising feature (Fig. 48) . There 







Chicken \ 




100 200 300 

Time since divergence (millions of years) 

Fig. 48. Reaction of different vertebrate DNA 
fragments with mouse DNA-agar under 
standard DNA-agar conditions. Each related- 
ness percentage is the ratio of the extent of 
interspecies binding to the extent of intra- 
species binding (mouse on mouse) X 100. 
Data were taken from several sources to con- 
struct this figure. Values for bovine, man, and 
hamster are from Year Book 63, p. 369; chick 
and rhesus, from Martin and Hoyer 45 ; the rat 
from McLaren and Walker. 46 Values for time 
since divergence for the various animals were 
obtained from paleontological evidence. 

is no detectable "threshold" period 
for accumulation of differences be- 
tween the DNA of different species. 
A threshold was expected since the 
measurements were carried out at 
60°C (about 30°C below the native 
DNA T m ) and very many individual 
nucleotide changes per fragment are 
required at this criterion to prevent 
pairing. The explanation is that the 
relatedness was measured only be- 
tween repeated families of DNA se- 
quences. The new families produced 

in one species will not reassociate 
with the DNA of the other species, 
since it is highly improbable that the 
same sequence would undergo salta- 
tory replication in both species. The 
new families increase the total quan- 
tity of repeated DNA and effectively 
dilute the related sequences, thereby 
reducing the measured relatedness 
between the two species. The phe- 
nomenon of new family production is 
known to occur and is probably the 
dominant mechanism for the initial 
loss of relatedness between species. 
It is also possible that many members 
of repeated families are actually lost 
by a process comparable to genetic 
deletion. This would also result in 
loss of relatedness without the ex- 
pected "threshold. " It is not known, 
however, whether deletions of this 
type occur. 

Figure 49 illustrates an inherent 
aspect of studies of DNA relatedness. 
The curves represent the loss of DNA 
relatedness with time measured at 
different temperatures of incubation 
(at a constant salt concentration). 
The lower temperature gives a larger 
value for the time necessary to lose 
50% relatedness. This reflects the 
greater number of nucleotide se- 
quence changes (longer time after 
species line divergence) necessary to 
render a pair unstable at lower tem- 
peratures. The apparent half-time of 
loss of DNA relatedness depends 
upon the incubation conditions em- 
ployed and must be interpreted with 

Rates of nucleotide sequence 
change. To calculate the rate of nu- 
cleotide sequence change it is neces- 
sary to determine the number of 
nucleotide changes that have oc- 
curred during a known time interval. 
The slow fraction of the DNA (see 
glossary) is useful for this purpose, 
since at the time of divergence these 
sequences can be presumed to have 
been identical and would have exhib- 




100 200 300 

Time since divergence, millions of years 

Fig. 49. The data are from Martin and 
Hoyer. 45 The relatedness values were obtained 
with the DNA-agar technique by reacting 
chicken (open circles) and rhesus (closed cir- 
cles) DNA fragments with mouse DNA-agar at 
different temperatures of incubation, 50°C and 
60°C in 2 X SSC. Each relatedness percentage 
is the ratio of the extent of interspecies binding 
to intraspecies binding X 1 00. 

ited perfect base-pair matching in 
the reassociated pairs. During diver- 
gence base-pair mismatches appear 
in the interspecies DNA duplexes and 
the thermal stability of the duplexes 
decreases. The temperature at which 
only 50% of the sequences of the slow 
fractions of the two animals still pair 
is a useful point for determining the 
average number of base changes 
since the time of divergence. The 
measured relatedness between mouse 
and rat slow fraction DNA is about 
50% at 60°C and K = 0.18. Thus 
60 °C is the estimated T m of the pop- 
ulation of reassociated related se- 
quences. Precisely matched sequences 

would melt at about 85° at k = 0.18. 
From the paleontological record it 
may be estimated that the time since 
divergence of rat and mouse is be- 
tween 10 and 25 million years. From 
this one may estimate that the rate 
of divergence reduces the average T m 
between species pairs 1° to 2°C per 
million years. Only approximate in- 
formation exists that relates the 
number of base-pair mismatches to 
the change in thermal stability of the 
reassociated complex. The best esti- 
mate is that somewhat less than 1% 
base-pair mismatch reduces the T m 
about 1°C (Year Book 64, p. 322). 
The rate of divergence between the 
rat and mouse is therefore between 
1% and 2% base-pair changes per 
1 million years. The rate of change 
of individual sequences is, however, 
one half of this, since the sum of the 
changes in the rat and the mouse has 
been measured. 

The data of Fig. 47 can be used 
to make a rough estimate for the rate 
of change of repeated sequences with 
time. The curve for mouse reassocia- 
tion extrapolates to a point on the 
temperature scale 17° above that to 
which the rhesus-mouse curve ex- 
trapolates. This estimate (weighted 
heavily by the more slowly changing 
sequences) yields about one sixth of 
a degree change per million years. 
It seems that the average rate of 
change of the repeated sequences is 
less than that of the nonrepeated se- 
quences. While the rates calculated 
here are probably relatively inaccu- 
rate, they are nevertheless the best 
available and add another dimension 
to our ideas about the evolution of 
nucleotide sequences and evolution in 

Saltatory Events of Replication 

Two major classes of evidence 
show that repetitive DNA very likely 
originated in rather sudden events of 



excessive replication of particular se- 
quences. These classes are: the rela- 
tive quantity of different degrees of 
repetition, and the relationship of 
repetition frequency to the age of 

Relationship between quantity of 
DXA and repetition frequency. A 
moderate fraction of calf or mouse 
DXA has a repetition frequency be- 
tween 10,000 and a million. There is 
little or no DXA with a repetition 
frequency between 10 and 1000. 
There is a large slow fraction (see 
glossary) which shows no repetition. 
Such a grossly separated bimodal dis- 
tribution could not result from the 
production of a small number of ex- 
tra copies of randomly selected se- 
quences. It could result from the 
recurrent copying of particular 
"marked" sequences, or from events 
in which very many copies are made 
in a short time interval (saltatory 
replication). Evidence is now avail- 
able which clearly indicates saltatory 

Relationship between age of fami- 
lies and repetition frequency. For 
those cases that have been examined, 
young families of repeated sequences 
which have not diverged very much 
have as many or more than the old 
families. The reassociation kinetics of 
the two fractions of salmon DNA 
shown on Fig. 46 were measured at 
50° C. Both fractions showed heter- 
ogeneity, but the 90° fraction started 
to reassociate at one fourth of the 
C t required for the 70° fraction. 
Thus there were present in some 
young families even more members 
than in any of the older families. 
This observation is substantiated for 
both calf and mouse DNA. Subse- 
quent and more precise observations 
of the correlation between number of 
members and age of families will be 
very informative regarding the his- 
torical pattern of saltatory events. 

Suppose family formation were a 

continuous process in which members 
were added in small numbers. Since 
divergence occurs we presume that 
ancient members of any family have 
diverged from each other. Newer 
members must be copied from pre- 
existing members and must reflect 
their divergence. In a continuously 
produced family, then, recent mem- 
bers are each closely related to only 
a small fraction of the whole. Thus 
sequences selected (as, for example, 
the 90° fraction in Fig. 45) for close 
relationship should show a lower fre- 
quency of repetition, but as men- 
tioned above the opposite seems to 
be the case. With saltatory replica- 
tions a family would be produced in 
toto in a time short compared to the 
period required for the divergence of 
its members from each other. Thus 
young families could have just as 
many or more members as old ones. 
A schematic diagram of family 
history. Since families of many de- 
grees of divergence are observed, we 
must assume that a long series of 
such events has occurred in each of 
the species lines leading to present 
higher organisms. Plate 6 shows a 
three-dimensional schematic model 
of the history of families produced 
by saltatory replications. The scale 
on the left represents the period of 
time since each of the families of 
repeated sequences was produced. 
The lower scale is a measure of 
the divergence within each family. 
The cutouts represent a prediction of 
the differential melting curves which 
would be observed for the families if 
the DNA were reassociated under 
conditions such that only repeated 
sequences paired. Thus all the mem- 
bers of a young family are closely 
related and all pairs melt at a tem- 
perature near that of native DNA. 
In ancient families a great deal of 
divergence has occurred and the av- 
erage melting temperature is far be- 
low that of native DNA. By chance 


some members will have suffered a favorably selected phenotype or be 
more changes than others and the associated with a favorable genetic 
melting curve thus broadens with element. (4) Sufficient time must 
time. Melting curves are somewhat pass for its dissemination throughout 
broadened due to initial spread in the population. (5) If the set of 
GC content. The range of quantities events can be described as a saltatory 
of DNA in the various families re- replication, the growth of the family 
fleets the wide range of repetition must be terminated within a reason- 
frequencies observed. ably short time. 

To bring this model to concrete No known processes combine all 

form, a number of arbitrary deci- five of these events. Virus infection, 

sions were made which are not cen- however, does involve some of them, 

tral to the underlying ideas. For ex- In virulent infection very many 

ample, the average rate of divergence copies are made of the virus genome, 

is not well known, and the degree of In a lysogenic state a chromosomally 

spreading is hardly known at all. It integrated segment is transmitted to 

is not known how often the saltatory progeny. Further, it is known that 

events occur; there is some suspicion segments of DNA that are capable of 

that they are quite rare. The model genetic activity in the host may, in 

suggests 10 7 -year intervals but as far some living systems, be carried in the 

as definite evidence goes, they could virus's life cycle. Cases are not known 

occur 10 or 100 times more often. If where large numbers of viral DNA 

about 10 families are observable copies are integrated into the host ge- 

( criterion 60° C) each would contain nome; nor are any cases clearly es- 

(e.g., in calf) about 5% of the DNA tablished in higher organisms where 

and would contain, on the average, new host-type genes are carried in 

10 5 copies of a sequence several thou- a virus's life cycle, 

sand nucleotide pairs long. We are not proposing that salta- 

Divergence betiveen species. It is tory replications are the result of 

possible to visualize, on this model, virus infection, but the relationship 

the divergence of the sequences may suggest new experimental ap- 

shared between species. Suppose, for proaches. There are several instances 

example, that the species lines to in which a quantity of DNA results 

modern horse and cattle diverged from the production of many copies 

about 70 million years ago. If this of particular segments of DNA: the 

model represents the DNA of cattle, tremendous quantity of DNA (mainly 

horse DNA would contain all the mitochondrial) in certain eggs; the 

families older than 70 million years, mitochondrial and chloroplast DNA 

For the younger families horse DNA in more typical cells. In none of these 
would have a quite distinct set of its . cases is it known that the copies may 

own. become integrated in such a way that 

What is a saltatory replication? their descendants are transmitted to 

The appearance in the genome of a the progeny. Such an event is neces- 

f amily of repeated sequences involves sary in order that the members of 

a number of events: (1) Many copies a repeated family may diverge from 

must be made of a segment of DNA. each other. Only the small fraction 

(2) A number of these copies must of the repetitive DNA which does not 

be integrated into the genome in such show divergence (having high ther- 

a way that they are duplicated and mal stability after reassociation) can 

transmitted to progeny. (3) The re- be DNA which has been recently 

suiting family must either determine replicated from a master copy. 


Chromosome Pairs also shows the reassociation of P 32 - 
The nucleotide sequences of the labeled E. coli DNA which was pres- 
DXA of the two members of a ho- ent at y 200 of the concentration of 
mologous pair of chromosomes are the calf thymus DNA. This DNA 
not ordinarily completely identical, served as an internal standard, since 
Genetic diversity derived from the both rates of reassociation are af- 
two parents is normally present in fected to the same extent by the 
a diploid organism. However, a close salt concentration, temperature, vis- 
similarity is expected, since there is cosity (sheared calf DNA at 7 mg/ 
only a limited degree of individual ml) or possible nonspecific molecular 
variation within a species. The meas- interactions. The result is that the 
urements reported here show similar- half period for the reassociation of 
ity on the nucleotide sequence level, the calf slow fraction is 690 times 
In fact, no differences are detectable greater than that for the E. coli 
with a limit of 1 % or 2 % nucleotide DNA. The ratio of the length of 
substitution in one member of the DNA in bull sperm (3.2 X 10 9 NTP) 
pair as compared with the other. Our to the length of the E. coli chromo- 
present estimate, reported in a pre- some (4.5 X 10 6 NTP) is just 710. 
vious section, of the rate of diver- The maximum error resulting from 
gence of nucleotide sequences be- combining all of these data is ap- 
tween the vertebrate species is about parently much less than a factor 
1 % nucleotide substitution in a mil- of 2. We may conclude that the num- 
lion years. Improved measurements ber of different nucleotide sequences 
in the future may somewhat change in the diploid calf thymus cell is 
this number but its exact magnitude about the same as that in the haploid 
does not enter into the argument, sperm cell. 

Thus the nucleotide sequences of the The divergence between homolo- 

DNA of vertebrate species in general gous chromosomes. Therefore, upon 

are slowly changing with time. A reassociation of the slow fraction of 

mechanism must be present which calf thymus DNA, about half of the 

preserves the similarity of the DNA duplexes will be formed with DNA 

sequences of the homologous chromo- derived from the two different mem- 

somal pairs while they slowly change bers of homologous pairs of chromo- 

together. somes. The thermal stability of the 

Hoiv much different DNA in the reassociated DNA is thus a measure 

diploid cell? Figure 43 shows the ki- of the average degree of divergence 

netics of reassociation of calf thymus between the sequences of homologous 

DNA. Our interest here is the right- chromosome pairs. No reduction in 

hand part of the curve, which results thermal stability due to this source 

entirely from the reassociation of is detectable (Year Book 65, Fig. 50, 

what we have termed nonrepeated p. 99) and the limit of accuracy is 

DNA. The rate of reassociation of at present about 2°C, which implies 

such DNA was reported in Year that no more than 2% of the nucleo- 

Book 65 to be about that expected for tides in the average sequences of one 

DNA occurring as single copies in member of a chromosomal pair differ 

the haploid genome. There does not from those of the other member, 

appear to be more than one compo- A direct comparison can be made 

nent in the kinetic curve, and the between this limit (lowering of the 

resulting reassociated DNA appears melting temperature of reassociated 

to have a thermal stability very close strands from the two homologous 

to that of native DNA. Figure 43 chromosomes) and the average low- 


ering of the melting temperature of this conservative mechanism clearly 

reassociated strands from two differ- does not operate to such an extent 

ent species. The melting temperature that all variation is eliminated from 

of the nonrepeated fraction of labeled the population. 

mouse DNA when reassociated with There are several levels at which 

rat DNA was reported (Year Book steps (1) and (2) could occur: A 

65, Fig. 51, p. 101, and Fig. 53, p. haploid set of chromosomes could be 

102) to be at least 20° and possibly duplicated to produce a diploid cell 

as much as 30° below that for the line as may occur in parthenogenesis, 

same tracer on mouse DNA. A failure could occasionally occur in 

It must be pointed out that the the process which supplies nonidenti- 

commercial cattle from which this cal pairs of chromosomes to daughter 

DNA was obtained are relatively in- cells. Crossing over could occur dur- 

bred. Similar experiments are obvi- ing mitotic division, which would 

ously necessary with DNA extracted create homozygous regions in the 

from both haploid and diploid cells chromosomes of daughter cells. If 

of wild species lines. such a mechanism operated on any of 

Implications of the similarity. A these levels it would act to eliminate 

high degree of homozygosity is main- genetic variation within the popula- 

tained within any species while the tion and be quantitatively important 

various species drift apart. The ge- in population genetics, 
netic evidence indicates that this is 

true on the level of expressible genes, A Source of Divergence? 

whereas the evidence described here An extraordinary variety of forms 

indicates that similarity is main- has appeared in the course of evolu- 

tained in the actual nucleotide se- tion, reflecting a restless source of 

quences of the DNA. Some definite originality. The nature of the forces 

statements can be made about the behind the acts of speciation and di- 

mechanisms that must operate to vergence remains a central problem in 

maintain the similarity. The normal the study of evolution. We may look 

events of mating and cell division do forward to surprises that are likely 

not supply a mechanism for the main- to turn up before a full understand- 

tenance of a close degree of similarity ing of these forces and processes is 

between chromosome pairs. One must achieved. In these paragraphs we 

look to relatively rare events to pre- propose a general idea suggested by 

serve similarity between the members the presence of repetitious DNA, and 

of a pair while the nucleotide se- indicate a role that the families of 

quences of both change on an evolu- repeated sequences might play. We 

tionary time scale. take as a starting point the view that 

The necessary elements of these momentous genetic events capable of 

rare events are as follows: (1) the inducing grand deviations in the 

duplication of nucleotide sequences ; course of evolution must occur. Such 

(2) the introduction of homozygous issues are raised with extraordinary 

regions into the chromosomes of the clarity in the following quotation 

germ line; (3) selection in the popu- from G. C. Simpson 47 : 

lation that favors homozygosity. ml , . . » ,.» . , ., ,, 

m, , , jsi. i The history of life is decidedly non- 

These events must occur of ten enough mndom TMs fa evident in many fea . 

and the selection for homozygosity tures of the record> induding such 

must be sufficiently effective that a points already discussed as the phe- 

great similarity is maintained be- nomena of relays and of major replace- 

tween chromosome pairs. However, ments at defined times. It is, however, 



still more striking in two other phe- 
nomena copiously documented by fos- 
sils. Both have to do with evolutionary 
trends : first, that the direction of 
morphological whence also functional 
and behavioral) change in a given 
lineage often continues without sig- 
nificant deviation for long periods of 
time and. second, that similar or paral- 
lel trends often appear either simul- 
taneously or successively in numerous 
different, usually related, lineages. 
These phenomena are far from uni- 
versal : they are not "laws" of evolu- 
tion ; but they are so common and so 
thoroughly established by concrete 
evidence that they demand a definite, 
effective directional force among the 
evolutionary processes. They rule out 
any theory of purely random evolution, 
such as the rather naive mutationism 
that had considerable support earlier 
in the twentieth century. What direc- 
tional forces the data do demand, or 
permit, is one of the most important 
questions to be asked of the fossil 

Suppose that an event occurs in 
which a large number of unexpressed 
genes, or better, a class of genetic 
potentiality, is produced in the ge- 
nome but only a small part appears 
phenotypically. The small, expressed 
part, "the top of the iceberg," if 
favorably selected will introduce the 
as yet unexpressed class into the pop- 
ulation. In effect a hidden package 
of potential genetic effects would 
have been selected which after 
spreading through the population 
could lead to radically new features 
that could not result from a detailed 
balance of mutation and selection. 

A saltatory replication producing 
100,000 copies of the right sort of 
gene is a candidate for a genetic event 
with immense potentiality. Not all 
the copies would be initially ex- 
pressed. Perhaps not many ever 
would be. But mutation, transloca- 
tion, and recombination with other 
genes would yield a whole range of 
potential genetic activity which 

would perhaps turn up long after the 
appearance of the first effects that 
resulted from the saltation. If the 
early effects were selectively advan- 
tageous the whole set of products of 
the saltatory replication would be 
introduced into the population. The 
dynamics of selection would be fun- 
damentally altered. The species might 
then appear to take a surprising and 
even an apparently purposeful course 
of evolutionary change. Owing to the 
great multiplicity of copies, their se- 
lective elimination might be impos- 
sible short of eradicating the species. 
Potentialities important in the longer 
term would then have an opportu- 
nity to survive and reach fruitful ex- 

Events of this general type would 
be capable of causing a striking di- 
vergence of two genetically isolated 
populations even though the environ- 
ments in which the two populations 
existed were identical. 

The wide occurrence of families of 
repeated sequences and the implica- 
tion of saltatory events in their pro- 
duction as well as in evolutionary 
processes indicates that sudden 
events may be far more important to 
evolution than we have heretofore 




S. Falkow and D. B. Cowie 

The DNA of the bacteriophage X 
can be differentiated into three dis- 
similar regions in composition by 
density gradient centrifugation 
(Year Book 65, p. 559). In contrast 
to these findings the virulent bac- 
teriophages of the T series do not 
exhibit a marked intramolecular het- 
erogeneity. In addition, the distribu- 
tion of guanine + cytosine (G + C) 
pairs in the DNA molecules isolated 
from a large variety of bacteria is 
unimodal and approximately gaus- 



sian — even after shearing of the 
DNA molecule. 

The DNA isolated from several sex 
factors may also be separated into 
compositionally distinct regions (Fal- 
kow et al, 1964, 48 1966 49 ). Sex 
factors, A, and virtually all other 
temperate phages fit the definition of 
episomic element; therefore each is 
an independent genetic element of an 
accessory nature, which is infectious 
and which may be established in two 
distinct intracellular, usually mutu- 
ally exclusive states: the extrachro- 
mosomal and the chromosomally at- 
tached state. On the basis of these 
similarities we considered that the 
clustering of G + C pairs might be 
a general feature of episomic ele- 
ments. We have tested this possibility 
with the DNA of several temperate 
phages (genetically all episomic ele- 
ments) by an optical method. Table 6 
lists the phages employed and the 
overall composition of their DNA. 

The usual method for studying the 
thermal denaturation of DNA has 
been to follow the relative increase 
in absorbance at 260 m/x as a function 
of temperature. The midpoint of the 
total increase in relative absorbance 
(T m ) is used to obtain the overall 
base composition while the hetero- 
geneity of the DNA is calculated as 
2o-, assuming that the compositional 

TABLE 6. DNA Base Composition of DNA 

Extracted from Temperate Phages and 

Their Bacterial Hosts 


Mole Fraction 

G + C% 
from Optical 

Mole Fraction 
G + C% from 
Buoyant Den- 
sity in CsCI 







Ab 2 









E. co/i K-12 



S. typhimurium 



distribution of DNA molecules is 
normal. The heterogeneity of distri- 
bution is more strikingly displayed 
by treating the absorbance data at 
any wavelength as a differential plot 
in terms of the increment of absorb- 
ance per unit of temperature (A 
O.D./degrees C) as a function of 
temperature or of mole fraction G + 
C where 1°C is equivalent to 2.1% 
G + C. Figure 50 shows that, in fact, 
the melting data for a variety of bac- 
terial DNA and the DNA of the 
phage T 3 do approximate a normal 
distribution. On the other hand, as 
shown in Figs. 51 and 52 the DNA 
isolated from the temperate phages A, 
$80, and P 22 all exhibit a marked 

Collecting these data requires con- 
siderable care and a large number of 
points must be measured during the 
denaturation process of the highly 
purified DNA. The points obtained 
very early and very late in the de- 
naturation were, of course, the most 
subject to error, since these were 
based on relatively small increases in 
O.D. Small "peaks" were often ob- 
served in the melting of all DNA 
preparations. We have considered 
these as real only if they could be 
reproduced in the majority of prepa- 
rations, as for example, in the case 
of A and <£80 DNA in Figs. 51 and 
52. The reproducibility of the method 
is shown in Fig. 53 in which the 
data from four separate experiments 
with two different </>80 DNA prepa- 
rations are plotted. 

Examination of these melting pro- 
files provides a simple means of de- 
termining the composition of partic- 
ular regions of a phage genome. In 
Fig. 52 the melting profiles of A and 
two A deletion variants, Ab 2 and Adg, 
are shown. A detailed comparison of 
the profiles of A and Ab 2 indicates that 
the DNA region missing in the Ab 2 
deletion mutant is of high A + T 
composition. The Adg variant is 









8 0.03 





P. mira bills 

P. aeruginosa 

40 50 60 

Mole fraction (G + C) 


Fig. 50. Differential plot of the increment of absorbance at 260 m/x per unit of temperature 
(triangle O.D./°C) as a function of mole fraction (G + C) for four bacterial DNAs and for 

phage T3 DNA. 

known to arise by a chromosomal re- 
arrangement consisting of a deletion 
of high G + C (56%) phage DNA 
and a concomitant substitution of the 
bacterial galactose genes. Although 
the base composition of the galactose 
region is unknown, the overall base 

composition of the bacterial chromo- 
some is 50% G + C. A comparison 
of the melting profiles of A and Adg 
DNA is consistent with the genetic 

Phage P 22 was the only generalized 
transducing phage studied. The DNA 



— 0.04 

— 0.03 


0.02 5. 

— 0.01 


44 54 

Mole fraction (G + C) 

Fig. 51. Thermal denaturation of the DNAs of phage <j>B0 or P 2 2 and for Salmonella typhi- 
murium. Data plotted as in Fig. 50. 



34 39 44 49 54 59 
Mole fraction (G + 

Fig. 52. Thermal denaturation of the DNAs of A, Adg, and Ab2 phages. Data plotted 
as in Fig. 50. 



44 54 

Mole fraction (G + C) 


Fig. 53. Thermal denaturation of the DNA of phage <£80. Data plotted as in Fig. 50. 



of this phage possesses an overall 
base composition of 46.5 c c G + C. 
This is significantly lower than that 
of its host S. typhimurium, which 
possesses 50.9% G + C (Fig. 51). 
In the past there has generally been 
a correlation of near identity of base 
composition between a temperate 
phage and its host, although it is rec- 
ognized that identity is not necessar- 
ily a prerequisite for transduction. 
Since phage P :2 is a collection of cir- 
cularly permuted gene sequences, the 
spectral technique provides informa- 
tion that could not be obtained by 
hydrodynamic shear experiments 
coupled with density gradient cen- 
trifugation as employed by Hershey 
{Year Book 63, p. 581) for A DNA. 

It is possible to obtain a better 
quantitative estimate of the hetero- 
geneity of the melting at 260 m/x by 
plotting the denaturation data on 
probability graph paper which con- 
verts a normal distribution into a 
straight line. The points do approxi- 
mate a straight line for the DNA of 
bacteria and virulent bacteriophages, 
whereas a line with three different 
slopes was generated for the DNA of 
the temperate phages. Figure 54 
shows the data for E. coli and X as an 
example of the method. Table 7 sum- 
marizes the overall composition and 
the fraction of the total DNA as de- 
termined from the separate lines for 
several of the temperate phage DNA 
preparations we examined. 

The implications of the differentia- 
tion of the DNA of temperate phages 
into regions of unique composition 
are not clear. Skalka ( Year Book 65, 
p. 565) has shown that the temporal 
differentiation into early and late 
genes appears to mimic the compo- 
sitional differentiation of the X ge- 
nome. These data are not available 
for any other phage. As noted by 
Hershey et al. (Year Book 63, p. 
589), the molecular differentiation of 
X could be physiologic or genetic. The 


1 1 

1 1 M 


1 U 


L — 



Ecoli 11 






'£ 60 

1 5 ° 

1 40 


1 30 




J? 20 


jt . i 



— y 








1 1 



1 1 1 

34 40 46 52 58 

Mole fraction (G + C) 


Fig. 54. Percentage of total denaturation 
plotted on normal probability graph paper. 
The percentages correspond to one hundred 
times the area under the normal distribution 

genetic elements that we have cited 
exhibit some sort of interaction with 
the genome of their host. As we shall 
point out in the following sections, 
they all share nucleotide sequence 
homology with the DNA of their host 
and with each other in more than one 
section along their molecular length. 
The accumulation of bacteria-like 
segments into the genome of a phage 
does not imply that bacterial viruses 
are simply a potpourri of bacterial 
genes, but it does seem likely that 
phages carry more of the bacterial 
chromosome than simply a single rec- 
ognition site. It is unreasonable to 
think that these bacteria-like seg- 
ments are present with no function; 



TABLE 7. Heterogeneity of Bacteriophage DNA from Data at a Single Wavelength 

Overall Mole 

Source Mole Fraction Fraction 

of DNA (G + C) % Total (G + C) 

Mole Mole 

Fraction Fraction 

% Total (G + C) % Total (G + C) 

























P 2 2 








selection undoubtedly demands that 
any gene preserved must play some 
role in the survival and maintenance 
of the phage genome. We report be- 
low that phages usually genetically 
isolated, such as A and P 22 , nonethe- 
less share a significant degree of ho- 
mology. Does this mean that there 
had been previous interaction be- 
tween them or that certain nucleotide 
sequences are conserved from a com- 
mon ancestor? To what extent exist- 
ing phages are really recombinational 
assemblages remains to be seen. At 
present we are simply trying to find 
some evolutionary thread preserved 
in their polynucleotide chains. It may 
be useful to think of the possibility 
that the molecular differentiation of 
the temperate phages represents a 
type of "historical differentiation" 
and reflects their evolution. 




D. B. Cowie and S. Falkow 

The attachment sites for most of 
the temperate bacteriophages of E. 
coli are located at specific regions on 
the bacterial chromosome. In general, 
each region for prophage attachment 
is characteristic for each of the dif- 
ferent viruses. It has been postulated 
that attachment specificity is a con- 
sequence of homology between the 

prophage and bacterial DNAs, and a 
number of temperate phages have 
been shown to be genetically related 
(i.e., their genomes share nucleotide 
sequences) to their bacterial hosts 
(Year Book 65, p. 106). 

The demonstration that DNA-DNA 
homologies play a direct role in at- 
tachment specificity has not been 
made. The problem is complicated, 
since more than one region of DNA 
homology is found between A DNA 
and the DNA of E. coli, and these 
regions are distributed throughout 
the A genome (Year Book 63, p. 385) . 
Furthermore, A also contains several 
regions of DNA homologous to P 22 
DNA and to the DNAs of other tem- 
perate phages (Year Book 63, p. 381) . 

The temperate phages A and ^80 
have different attachment sites and 
transduce different gene segments — 
gafr and try + , respectively. Genetic re- 
combinants between these two phages 
have been obtained by Signer 50 al- 
though these phages are serologically 
unrelated. An investigation of DNA- 
DNA reactions among A, Adg, <£80, 
and E. coli has been carried out to 
examine, in part, the nature of the 
homologies existing among them. The 
thermal chromatographic method of 
investigating DNA-DNA reactions 
was used. The resulting elution pro- 
files obtained are presented and dis- 
cussed in terms of some of the unique 



characteristics of the DNA of these 
temperate phages. 

DNA-DNA Reactions Involving 
Identical DNAs 

Thermal elution profiles, each char- 
acteristic of specific DNA-DNA re- 
actions, have been obtained from 
studies involving labeled DNA frag- 
ments reacting with identical or het- 
erologous DNA-agar preparations. 
Figure do presents the results of re- 
acting labeled E. coli DNA fragments 
with E. coli DNA trapped in agar 
(solid curve). The broken curve 
shows the data obtained by reacting 
680 DNA fragments with $80 DNA- 
agar. The elution profile for the E. 
coli DNA /DNA-agar reaction has a 
maximum at 72 °C. Two maxima 
are observed in the #80 DNA/DNA- 
agar reaction, one at 69 °C, the other 
at 73°C. 

If it is assumed that the release of 
labeled fragments from the DNA in 
the agar represents the melting out 
of fragments bound to the agar-em- 
bedded DNA by the matching of nu- 
merous base pairs, then the temper- 
ature of elution may be attributed to 
some function of the guanine plus 
cytosine (G + C) content of the re- 
acting DNAs. The elution profiles 
shown in Fig. 55 would therefore in- 
dicate that the overall G + C content 
of these two DNAs is somewhat sim- 
ilar, although that of the $80 DNA 
is a little higher than that of the E. 
coli DNA. 

The usual method for studying the 
thermal denaturation of DNA is to 
follow the relative increase in ab- 
sorbance at 260 m/x as a function of 
temperature. Figure 56 represents 
the differential plot of this change in 
optical density observed with the 
heating of native DNAs from $80 
phage and E. coli in SSC/30. A com- 
parison of these data with those ob- 
tained in SSC/30 by the thermal 

chromatographic method (Fig. 55) 
indicates that the two procedures 
provide similar results. The thermal 
chromatographic data for the $80 
DNA show some intrasequence heter- 
ogeneity not resolved by the optical 

In the preceding section (Falkow 
and Cowie) another analytic method 
was employed which demonstrated 
that large-scale intramolecular heter- 
ogeneities not observed with bacterial 
or virulent phage DNAs are found 
among the temperate phage DNAs. 
Results obtained with the thermal 
chromatographic method are quite 
comparable to those obtained with 
the modified optical method and show 
the same molecular heterogeneity 
among the temperate phage DNAs. 
For example, Fig. 57 presents the 
elution profiles resulting from studies 
of the reaction of P 22 DNA fragments 
with P 22 DNA-agar (upper left-hand 
curve) . The elution profile is complex, 
showing two elution maxima. When 
the data for P 22 DNA from the optical 
studies (Falkow and Cowie) are nor- 
malized to the conditions used for the 
DNA-agar thermal elutions and 
plotted (Fig. 57, lower left-hand 
curve), the results prove to be quite 
comparable. A similar comparison is 
shown for T 3 DNA in Fig. 57, where 
there is no large-scale heterogeneity 
of distribution and the profiles are 

Heterologous DNA-DNA 

These studies of identical DNA re- 
actions serve as reference material 
essential for studies of the reactions 
involving heterologous DNAs. 

A portion of $80 phage DNA is 
homologous to the DNA of its bac- 
terial host. Figure 58 (broken curve) 
shows that $80 DNA fragments react 
with E. coli DNA-agar. An elution 
maximum is seen at a temperature 




















</> 80 Phage DNA fragments / \ 
<t> 80 DNA - agar 

E. COli DNA fragments 
E. COli DNA - agar 

Temperature, °C 

Fig. 55. Elution profiles obtained from a study of the reaction of E. co// DNA fragments 
with E. co// DNA-agar (solid circle), and <j>S0 DNA fragments with </>80 DNA-agar (open circle). 

of 76°-77°C and another at 58°- coli DNA fragments with <j>S0 DNA- 

59°C. agar (solid curve). 

An elution profile was also obtained The temperature required to elute 

from the results of a reciprocal ex- some of the heterologous DNA frag- 

periment to study the reaction of E. ments from the DNA-agar prepara- 



tions (76°— 77°C) is considerably 
higher than that required for the 
thermal elution of either </>80 or E. 
coli fragments from their identical 
DNA-agar preparations (Fig. 55). 
This finding indicates that authentic 
matching- of numerous base pairs 
must have occurred between the het- 
erologous DNAs. Furthermore, these 
matching segments will be contained 
in regions in each of the reacting 
genomes where the G + C content is 

higher than their individual overall 
G + C content. 

The elution profiles shown in Fig. 
58 have another interesting feature. 
About 13% of the </>80 DNA frag- 
ments that were bound to the E. coli 
DNA are eluted at temperatures 
between 57° and 60°C. An elution 
maximum at these temperatures is 
observed when: (1) E. coli DNA 
fragments react with ^80 DNA-agar, 
(2) </>80 DNA fragments react with 



>? 20 

4> 80 DNA 




E. coli DNA 


12 ^ 

70 80 

Temperature, °C 




Fig. 56. Differential plot of the change in optical density observed with heating native DNAs 
in SSC/30; </>80 DNA (open circle); E. co///DNA (solid circle). 



(j>S0 DNA-agar, (3) A DNA fragments 
react with A. DNA-agar, (4) A DNA 
fragments react with ^80 DNA-agar. 

This region is of interest because 
of its extremely low elution tempera- 
ture and, as will be seen, it is in- 
volved in reactions between the DNA 
of phages A and </>80. 

Figure 59 demonstrates that <£80 
is genetically related to phage A. The 
solid curve represents the elution pro- 
file obtained from a study of the re- 
action of A fragments with <£80 DNA- 

agar. The elution data show two 
prominent maxima, one at 58 °C, the 
other at 74 °C. Twenty-two per cent 
of the A DNA reacts with the </,80 
DNA. The major elution peak (74°C) 
is above the 72 °C peak observed for 
the reaction of E. coli fragments with 
E. coli DNA-agar (50% G + C) . The 
higher temperature required for elu- 
tion of the A fragments from the E. 
coli DNA indicates that this binding 
is probably mostly concerned with the 
left end of A DNA (56% G 4- C). 

60 65 

Temperature, °C 
70 75 65 70 




£ 0.01- 



j f^ 2 DNA fragments 
I P 22 DNA - agar 

I T 3 DNA fragments 
























40 45 50 45 50 

Mole fraction G + C 


Fig. 57. Elution profiles obtained from studies of the reaction of P22 DNA-agar fragments with 
P22 DNA-agar (upper curves, solid circles) compared with optical data obtained by the thermal 
denaturation of native P22 and T3 DNAs (lower curves, open circles). 














1 1 

£■<?// DNA fragments 



- ? 

80 phage DNA - agar * 


Cp 80 phage DNA fragments 

E. coli DNA-agar 

sc r» 


t i 






i i 




\ \ 

















- l 









60 70 80 

Temperature, °C 

Fig. 58. Elution profiles resulting from a study of the reaction of £. co/i DNA fragments with 
080 DNA-agar (solid circles) and 080 fragments with E. coli DNA-agar (open circles). 


These data agree with the results of agar-bound $80 DNA. When the dif- 

Ingraham, Ehring, and Hershey ferences between the two elution dia- 

(Year Book 65, p. 563), who used A grams (shown in Fig. 65) are plotted, 

DNA fractions (41% G + C, 46% however, two prominent maxima are 

G + C, and 56% G + C) to react observed at 58°C and 72°C, as shown 

with $80 DNA-agar. Ingraham et at. in Fig. 60. These peaks represent A 

found that most of the binding oc- DNA fragments having homology 

curred with the 56 % G + C fraction with E. coli DNA as well as with $80 

and was twice that observed for the DNA. Other A fragments not affected 

46% fraction and almost eight times by the presence of the E. coli com- 

the binding observed with the 41% petitor fragments are homologous to 

G + C fraction. </>80 DNA but not to E. coli DNA. 

The material eluted at 58 °C in The homologous region eluting at 
SSC/30 represents a much lower 58° C is believed to be contained in 
G + C content, and is possibly con- the low G + C segments of the A 
tained in the 41 % G + C fraction of DNA — that is, in the central portion 
the A genome. On the other hand, or the right side of the X genome. 
imperfect base sequence homology be- A portion of the Adg DNA consists 
tween the reacting portions of these of E. coli DNA substituted for the A 
heterologous DNAs would also cause DNA contained in the 56% G + C 
a reduction in the temperature re- fraction (left side of the A genome) 
quired to elute the bound fragments as shown by Kaiser and Hogness. 51 
from the DNA-agar, compared to a When $80 DNA fragments are sub- 
situation where perfect homology jected to reaction with Adg (Fig. 61) 
exists. The fact, however, that this no significant differences are ob- 
region is observed with reactions in- served in the elution profile from that 
volving identical DNA reactions (A obtained when A DNA fragments 
DNA fragments with A DNA-agar or were reacted with $80 DNA-agar 
<£80 DNA fragments with $80 DNA- (solid curve, Fig. 59). This result 
agar) probably means that almost suggests that the deleted portion of 
perfect matching of many nucleotide the Adg molecule does not contain the 
sequences in this fraction had also A region homologous to $80 and elut- 
occurred when the heterologous ing at 58°C. 
DNAs were incubated together. The results presented above may 

A portion of the A DNA which re- be summarized as follows: A sizable 

acts with the $80 DNA is homologous portion of the DNA of lysogenic (or 

to E. coli DNA. Figure 59 (broken semilysogenic) viruses has been 

curve) shows the effect of adding a shown to be homologous to the DNA 

large excess of cold E. coli DNA frag- of their bacterial hosts. Nucleotide 

ments to the radioactive A DNA frag- sequence relationships among four 

ments and then incubating both at viral-host systems (A phage/E 1 . coli), 

60°C with the $80 DNA-agar. A 15 (TAU)~ phage/E. coli, P 22 phage/ 

43% reduction in the binding of the Salmonella, and T 3 phage /E. coli 

X fragments to the $80 DNA-agar have been studied by means of an 

was observed. A reduction in the extension of the DNA-agar technique 

binding of the A fragments over the of Bolton and McCarthy. 52 

entire elution diagram (shown in the The results presented here thus 

figure) is to be expected, since the provide another example demonstrat- 

competitor, cold E. coli DNA frag- ing genetic relationships between the 

ments, reduces the quantity of A frag- lysogenic virus $80 and its bacterial 

ments left available to react with the host E. coli. Furthermore, a portion 




XDNA fragments 
080 DNA - agar 

A + £". coli fragments 
^80 DNA- agar 

Temperature, °C 

Fig. 59. Elufion profiles obtained from a study of A DNA fragments (.075 fig DNA) re- 
acting with cS80 DNA-agar (37 fig DNA, open circles) and a similar experiment (closed circles) 
with unlabeled competitor E. coli DNA fragments (460 /xg DNA, closed circles). 

of the <£80 DNA which is homologous 
to E. coli DNA is also homologous to 
X DNA; other </>80 DNA segments are 
capable of reacting with A. DNA but 
not with E. coli DNA. DNA homolo- 
gies have been shown to exist among 
all of the temperate phages investi- 
gated (with the exception of the de- 
fective phage lysogenic to strain 15 
E. coli); see Year Book 65 (p. 107). 

The viral-viral or viral-host homol- 
ogies are contained in numerous seg- 
ments throughout the viral DNAs. 
In the <£80-A system two DNA frac- 
tions eluting at temperature differ- 
ences of 16°C (58°C and 74°C, Fig. 
59) are found to be homologous. This 
extreme temperature difference prob- 
ably represents an extreme variation 
in G + C content of the reacting 



segments and therefore separate re- 
gions of the A genome, and of the </^80 
DNA as well. 

One component of the A DNA, ho- 
mologous to both E. coli and $80, is 
contained in a nucleotide cluster elut- 
ing at a temperature of 58 °C. If the 
specificity of prophage attachment to 
host chromosome is dependent upon 
unique DNA homologies with host 
cell DNA — and since the attachment 
sites for A and $80 are presumably 


200 — 










60 70 80 

Temperature, °C 

Fig. 60. Plot of the difference between the 
two elution profiles shown in Fig. 59. 

significantly different — the nucleotide 
sequences contained in this one com- 
ponent probably do not enter into the 
attachment process. Another fraction 
of the A genome, that eluting around 
72°C (Fig. 60), might be excluded 
for the same reason. It must be em- 
phasized, however, that the results 
presented are only qualitative, and 
the broad elution maximum at 72 °C 

4>80 Phage DNA fragments 

X dg DNA - agar 



60 70 

Temperature, C C 


Fig. 61. Elution profile obtained from a study 
of the reaction of (£80 DNA fragments with 
Adg DNA-agar. 











c 6 


DNA fragments 
£". Ctf// DNA - agar 


80 DNA fragments 
CO// DNA - agar 




Temperature, °C 


Fig. 62. Elution profiles obtained from a study of the reaction of A DNA fragments and E. coli 
DNA-agar (solid circles) and <£80 DNA fragments with E. coli DNA-agar (open circles). 



might mask a small fraction of re- 
acting DNA that is involved in the 
attachment process. 

Studies of the reactions of A or $80 
DNA fragments with E. coli DNA- 
agar, however, provided significantly 
different elution profiles, which are 
compared in Fig. 62. It is apparent 
that the G + C content of the X-E. 
coli homologies are in general lower 
than that observed for the <j>80-E. coli 

The differences between these two 
elution profiles can be graphically 
presented, as shown in Fig. 63. In 

60 70 

Temperature, °C 


Fig. 63. Plot of the difference between the 
two elution profiles shown in Fig. 62. 

the figure the percentages of the 
bound $80 fragments are subtracted 
from the corresponding values of the 
A fragments, which are bound to the 
E. coli DNA-agar. These difference 
data show two major peaks, one with 
a maximum at 66 °C, the other at 
78°C. A third peak at 58°C is also 
seen, but as noted, this material is 
not believed to be involved in the at- 
tachment process. 

The 66°C region represents the 
percentage of excess binding of the 
A DNA fragments over that observed 
with </>80 DNA. The 78°C peak cor- 
responds to the percentage of excess 
binding of the $80 fragments to E. 
coli over that observed with the A 

If the requirement for attachment 
of viral DNA to host chromosome for 
A and (£80 is dependent upon differ- 
ences in nucleotide composition in a 
viral-host DNA homology, these two 
regions represent the areas having 
the maximal nucleotide differences. 

The results described serve as 
guidelines for future investigations. 
Lambda mutants having altered or 
deleted attachment sites might be ex- 
pected to show alterations in the elu- 
tion profiles in the 66° C region, and 
$80 mutants, in the 78 °C region. The 
thermal chromatographic method 
provides a means of isolating and ex- 
amining specific DNA segments in 
terms of their genetic interrelation- 
ships and functions, and it is hoped 
that further elucidation regarding 
prophage attachment to host DNA 
may result. 

In the preceding section (pages 
94-97) it was demonstrated that 
large-scale intramolecular heteroge- 
neity exists among the temperate 
phage DNAs. The thermal chromat- 
ographic results presented above 
confirm this conclusion. The elution 
profiles obtained from studies of iden- 
tical DNA-DNA reactions involving 
$80 DNA (Fig. 55), P 22 DNA (Fig. 



57), and A DXA {Year Book 65, p. 
114) are all complex. Furthermore, 
studies involving DNA-DNA reac- 
tions among- these temperate phages 
show a large number of different 
elution maxima having a spread in 
elution temperatures ranging from 
5S C to 78 T. The specificity of these 
reactions and the wide differences in 
their temperature characteristics im- 
ply that large-scale heterogeneities 
exist among these DNAs. 


D. J. Brenner and D. B. Cowie 

The percentage of DNA fragments 
bound to DNA-agar yields little in- 
formation concerning the precision of 
complementarity of the bound nucleo- 
tide sequences. Martin and Hoyer, 53 
using animal DNAs, and Brenner, 
Martin, and Hoyer, 54 using bacterial 
DXAs, have shown that the amount 
of interspecies DNA binding in agar 
decreases as the incubation tempera- 
ture is raised. A logical explanation 
for the increasing discrimination at 
higher temperatures between intra- 
species and interspecies duplexes is 
that in the complementary sequences 
held in common between species (or 
at least a significant proportion of 
them) are less precisely paired than 
the intraspecies duplexes. If this is 
true the thermal stability of a paired 
sequence should decrease with an in- 
crease in unpaired bases within the 
sequence. When unique animal DNA 
sequences are reassociated in free so- 
lution the interspecies pairs do dis- 
sociate at lower temperatures than 
the intraspecies duplexes (Britten 
and Kohne, Yeo/r Book 65, p. 78). 

Animal DNAs contain large num- 
bers of highly repetitive nucleotide 
sequences that fall into families with 
imprecise sequence complementarity. 
Since the rate of reassociation is de- 

pendent upon the concentration of 
related sequences and the animal ge- 
nome is so large, only the sequences 
contained in the DNA families re- 
associate within a 16-hour incubation 

Bacterial and phage systems were 
chosen to investigate further the pre- 
cision of interspecies pairs. There is 
no evidence for repetitious DNA se- 
quences in bacteria at incubation 
temperatures above 45°, and no mul- 
tiple DNA sequences have been dem- 
onstrated in bacteriophage, except 
for redundancy in the terminal ends 
of several coliphage DNAs. 

The use of the term "homology" 
has been purposely avoided in the 
preceding paragraphs. Homology was 
used previously to depict intraspecies 
and interspecies binding regardless 
of incubation conditions. In this con- 
text homology has been misinter- 
preted to imply the existence of 
identical stretches of DNA between 
species; for example, if 35% ho- 
mology were reported between E. 
coli and Salmonella typhimurium, 
these organisms might be assumed 
to have 35% identical DNA. It is well 
known that binding percentages in 
bacteria and animals are completely 
dependent upon the stringency of 
incubation conditions. For example, 
Martin and Hoyer found 90% com- 
plementarity between the DNAs of 
mouse and rhesus monkey at 40°, and 
20% complementarity at 64°. 

Restricting our discussion to phage 
and bacterial systems, where DNA 
sequence repetitions do not occur, 
complementary pairs will be termed 
"precise" if they exhibit the thermal 
stability of reassociated intraspecies 
pairs, and "imprecise" if their ther- 
mal stability is less than that seen in 
the intraspecies pairs. 

E. coli and S. typhimurium were 
selected as the organisms with which 
to assess the precision of pairing in 
interspecies bacterial DNA duplexes. 


In agar these organisms show 36% droxyapatite column in E. coli*/E. coli 

nucleotide sequence complementarity and E. coli*/S. typhimurium reac- 

under standard incubation conditions tions (where the asterisk denotes 

(60° in 2 X SSC) and 28% comple- radioactive DNA fragments) incu- 

mentarity when the incubation tern- bated 4 hours at 60° and 66° is the 

perature is raised to 66°. The pre- result of duplexes formed between 

cision of DNA sequences in these or- labeled and unlabeled fragments, 

ganisms was determined by thermally Eighty to 90 per cent of the unlabeled 

eluting reassociated duplexes from hy- fragments reassociate at the concen- 

droxyapatite {Year Book 65, p. 78). tration and period of incubation em- 

At 0.14 M PB hy droxyapatite retains ployed (Britten and Kohne, Year 

double-stranded, but not single- Book 65, p. 78) . 

stranded DNA. Double-stranded DNA The E. coli*/E. coli and E. coli*/ 
is eluted at a 0.4 M PB. S. typhimurium samples were loaded 
Denatured P 32 -labeled E. coli frag- on hydroxyapatite columns held at 
ments were incubated alone (as a the 60° or 66° incubation tempera- 
control experiment) at 60° and 66°. ture. The bound (double-stranded) 
The data in Table 8 show that at both DNA was then denatured by raising 
incubation temperatures little or no the temperature of successive 0.14 M 
reassociation of the labeled frag- PB washes by 5° increments to 100° 
ments with one another, in excess of and finally washing the column with 
that seen in a zero time control sam- 0.4 M PB to remove any remaining 
pie, occurred after 4 hours. One may double-stranded material. The eluents 
assume, therefore, that virtually all were assayed for total DNA and 
the radioactivity bound to the hy- radioactivity to obtain the differen- 

TABLE 8. Elution of P 32 -Labeled E. coli Fragments from Hydroxyapatite 

Zero Time F 32 - 

Labeled E. coli 


P 32 -Labeled E. coli 


60°, 4 hours 

P32-Labeled E. coli 


66° , 4 hours 

Total C f 




Cpm unbound 




Cpm bound 




% bound 




* The total counts are lower in this sample because it was assayed two days after the other 

Three 0.4-/i,g samples of P 32 -labeled E. coli DNA fragments, each contained in 1 ml of 0.14 
M PB, were thermally denatured and quickly cooled. The zero time sample was immediately 
passed through a 5.5-ml hydroxyapatite column at 60°. The two other samples were incubated 
at 60° or 66° for 4 hours and then passed through hydroxyapatite at the temperature of 
incubation. The column was then washed with 0.14 M PB at 60° or 66° until the eluted counts 
were down to a background level of 7 cpm. The column was then consecutively washed 
with 0.14 M PB at 100° and 0.4 M PB to elute all counts initially bound to the column. The 
samples were assayed for radioactivity in the eluting solution by Cerenkov counting. The P 32 - 
DNA fragments contained little acid-soluble material. 



tial elation profiles. Elution profiles 
of the unlabeled material (Fig. 64) 
show the melting characteristics of 
E. coli and 5. typhi 'murium DNA af- 
ter incubation at 60° and 66°. The 
slightly higher G + C content of S. 
typhi murium DNA is reflected in its 
higher T . 

Thermal elution profiles obtained 
from radioactivity assays are pre- 
sented in Fig. 65. E. coli* /E. coli 
elution profiles from 60° and 66° in- 
cubations are quite similar to the 
elution profiles obtained from un- 

labeled fragments (Fig. 64), although 
their T m s are 3° lower. The E. coli*/ 
S. typhi murium DNA profile from 
the 60° incubation is displaced 9.5° 
below that of an E. coli* JE.coli pro- 
file from a 60° incubation and the 
E. coli* /S. typhimurium profile from 
a 66° incubation is displaced 7.5° be- 
low the corresponding E. coli* /E. coli 
profile. Therefore the complementary 
nucleotide sequences in E. coli*/S. ty- 
phimurium duplexes are imprecise. 
While one cannot say how much, if 
any, of the interspecies nucleotide se- 






•Z 5 



E. coli 60 


E.coli 66 

S. typhimurium 66 

S. typhimurium 60 

100 ? U4M PB 

Temperature, °C 

Fig. 64. Thermal elution profiles of unlabeled reassociated bacterial DNA from hydroxy- 
apatite. Fragments of E. coli or S. typhimurium DNA were incubated at 60° or 66° and then 
passed through a hydroxyapatite column. The bound fragments were removed from the column 

by a series of thermal elutions. 



quences are precisely paired, this per- 
centage is necessarily small. 

The T m of the E. coli*/S. typhi- 
murium duplex from a 66° incubation 
is 2° higher than that of the duplex 
formed at 60°. This difference is to 
be expected because there should be 
a class of imprecisely paired se- 
quences that can reassociate at 60° 
but are unable to reassociate at the 
more stringent 66° incubation tem- 
perature. Normalized binding per- 
centages obtained from the 60° and 
66° reactions confirm the existence 

of this class of imprecise relatives 
(Table 9). When the binding per- 
centage obtained with E. coli*/E. coli 
on hydroxyapatite is arbitrarily des- 
ignated 100% E. coli* and S. typhi- 
murium exhibit 42% complementarity 
at 60° and 35% complementarity at 
66°. Under these conditions 7% of 
the paired nucleotide sequences 
formed at 60° are absent at 66°. An 
8% difference occurs in the binding 
of E. coli* fragments to S. typhimu- 
rium DNA-agar in 60° and 66° incu- 
bations. The good agreement between 


















E coli*/ S. typhi murium 
66°(T m = 75.5°) 

E. coli*/ S. typhimurium 
-60°(T m = 73.5°) 

E. coli VE. coli 60 c 



8 — 

80 90 

Temperature, °C 

Fig. 65. Thermal elution profiles of E. co//*/E. coli and E. co//*/S. typhimurium DNA reac- 
tions. Samples of P 32 labeled E. coli DNA fragments were incubated with unlabeled E. coli or 
S. typhimurium DNA fragments at 60° or 66°. The samples were then passed through hydroxy- 
apatite and the bound fragments were removed by a series of thermal elutions. 


TABLE 9. Extent of Duplex Formation between E. coli and S. typhimurium at 60° and 66° 

Hydroxyapatite, % 
(Total C f, 43) DNA-Agar,' % 

60° 66° 60° 66° 

E. coli* E. coli binding 78 76 23-30 18-23 

E. coli* S. typhimurium binding 







Normalized E. coli* S. typhimurium binding 42 35 34 26 

[E. coli/E. coli = 100 per cent) 

1 Taken from the results of Brenner, Martin, and Hoyer. 54 

Four 0.4-ag samples of P 32 -labeled £. coli DNA fragments were denatured and quickly cooled. 
Two samples were mixed with denatured, unlabeled E. coli fragments in 1 ml of 0.14 M PB and 
two samples were mixed with denatured, unlabeled S. typhimurium fragments in 1 ml of 0.14 
M PB. The unlabeled fragments were at approximate concentrations of 1 mg. One member of 
each pair of samples was incubated at 60° for 4 hours and the other member of each pair was 
incubated at 66° for 4 hours. Subsequent treatment of the samples was identical to that de- 
scribed in Table 8. 

the agar and hydroxyapatite data, tions will occupy a part of our efforts 
despite the differences in absolute in the coming year, 
binding percentages is of general sig- Phage-host and phage-phage sig- 
nificance because data obtained by terns. Specific profiles are obtained 
either of these methods may now be following thermal elution of phage 
compared with greater confidence. and bacterial DNA-agar duplexes 
Having established that the major- (Year Book 65, p. 106). Further- 
ity of nucleotide sequences held in more, the elution peaks and T m values 
common between E. coli and S. typhi- observed in thermal elution profiles 
murium are imprecisely paired, we from DNA-agar duplexes closely 
may now consider several questions mimic the hyperchromicity and T m 
concerning relatedness among bacte- values obtained when DNAs in free 
ria: solution are thermally denatured and 

1. Can one determine the percent- assayed optically (see Cowie and 
age (if any) of precisely paired E. Falkow, preceding sections) . 
coli/S. typhimurium nucleotide se- Cowie and Szafranski (Year Book 
quences? 65, p. 106), and Cowie and Falkow 

2. Will precision of sequence com- (preceding sections) have excellent 
plementarity be higher in DNA du- evidence for precise or very nearly 
plexes formed between more closely precise binding between the DNA se- 
related bacteria, e.g., Shigella and E. quences of phage and their bacterial 
colli whose DNAs show 70% binding, hosts and also between the DNAs of 
or among some Proteus species whose different temperate phages. In fact 
DNAs bind as high as 90% ? the DNA-agar elution profiles ob- 

3. Conversely, will less precise tained from the supposedly distant 
DNA duplexes be formed between bacteriophage pairs A and </>80 or A 
bacteria that are less closely related and P 22 showed about 20% precise 
than E. coli and S. typhimurium? binding and an apparent absence of 

Answering these and related ques- imprecise binding. Only precise pair- 


ing was observed in the elution pro- relationships between the DNAs of 

files from A or </>80 with E. coli DNA- these phages and their E. coli host 

agar. These experiments with A, ^80 were investigated. Lambda, 434, and 

and P 22 employed incubation temper- 434 hy were selected for test because 

atures of 60° in 2 X SSC. The agar they are closely related on the bases 

was washed in SSC/30 at 50°, which of recombination and susceptibility to 

is equivalent to approximately 70° in anti-A antiserum. 
2 X SSC. Therefore the DNA frag- The extension of Bolton and Mc- 

ments which bound in 2 X SSC and Carthy's DNA-agar technique that 

remained stable between 60° and 70° employs thermal elution has been pre- 

were not included in the elution pro- viously described (Year Book 65, p. 

files. This washing procedure is quite 106) ; however, sufficient modifica- 

stringent, thus giving increased ere- tions warrant a restatement of the 

dence to the precise reassociation techniques. 1 

that occurs between phage, and be- Phage-bacterial reactions. Elution 

tween phage and bacterial DNAs. profiles from 4M*/E. coli and 434 

On the other hand, distantly related hy*/E. coli DNA reactions are shown 

duplexes with poor complementarity in Figs. 66 and 67. The similarity of 

would not be detected with this wash- these profiles indicates that the nu- 

ing procedure. In the present experi- cleotide sequences in 434 and 434 hy 

ments the SSC/30 washes were capable of pairing with E. coli DNA 

carried out at 30° in order to allow are either very precise or at least 

resolution at the low temperature end have the same average G + C dis- 

of the elution profile. tribution. The precise complementar- 

Quantitative and qualitative DNA ity indicated by elution profiles ob- 

relationships in more closely related tained from reactions between these 

temperate coliphages as well as the phage DNAs and the DNA of their 

1 Preparation of phage DNA. Cultures of RNase and pronase, phenol extracted, etha- 
E. coli K12 lysogenic for X, 434, or 434 hy nol precipitated, and finally resuspended in 
were grown to late logarithmic phase in SSC/100. Specific activities ranging from 
nutrient broth, sedimented, and resuspended 5 x 10 4 to 1 X 10 6 cpm//ig DNA have been 
in 0.01 M MgS0 4 , induced to produce phage obtained in various preparations, 
by ultraviolet light, diluted 1 to 2 in fresh Incubation and elution conditions. Two 
nutrient broth, and incubated at 37° in a tenths gram of agar containing 15-80 fig 
water bath shaker until lysis was complete. of DNA was incubated with 0.2 ml of de- 
Labeled phage lysates were obtained by natured P 32 -labeled DNA fragments in 
resuspending the lysogenic bacteria in a 2 x SSC at a ratio of DNA in agar to 
tris-glucose buffer lacking phosphate, incu- DNA fragments ranging from 500/1 to 
bating 20 minutes at 37°, then adding P 32 100,000/1 for 15-16 hours at 60°. The DNA- 
and allowing lysis to proceed. Chloroform agar was then transferred to a tube fitted 
was added to the lysates, which were then with a saran-screen bottom and washed with 
freed of bacterial debris by differential seven 15-ml portions of 2 x SSC at 60°, 
centrifugation. The resulting phage pellets followed by nine 15-ml portions of SSC/30 
were exposed to DNase and RNase, then at 30°. Each wash required 5 minutes and 
extracted with phenol. Additional phenol the DNA-agar was agitated several times 
and ether extractions followed by ethanol during each wash to ensure efficient removal 
precipitation were carried out to purify the of unhybridized DNA fragments. At this 
DNA further. When pure, the DNA, re- point the DNA-agar received one 10-ml 
suspended in SSC/100, was thermally de- SSC/30 wash of 5 minutes duration at 2° 
natured and immobilized in agar at a final increments up to 80°. All wash fluids were 
concentration of 75-400 fig/g agar. The measured for radioactivity. The percentage 
labeled DNA was similarly purified, then of bound DNA fragments was calculated as 
sheared at 12,000 psi in a French pressure total counts/bound counts x 100. 
cell. The fragments were then exposed to 



bacterial host, as compared to the 
intraspecies phage and E. coli*/E. 
coli DXA profiles, is evident, and 
similar to the precision indicated by 
a A* E. coli DXA elution profile. 
An additional feature of these 

phage ^/bacterial DNA elution pro- 
files is of interest. Structure is pres- 
ent in the low-temperature end of 
these profiles that is not seen in re- 
actions involving identical DNAs. 
The low temperature at which these 



* 10 









rS 6 

E.colWE.coh o 

434hy*/434hy / 



50 60 

Temperature, °C 

Fig. 66. Thermal elution profiles characteristic of reactions between E. coli DNA fragments 
and E. coli DNA-agar (open circles, clashed curve); 434 DNA fragments, and 434 DNA-agar 
(closed circles, solid curve); and 434 DNA fragments and E. coli DNA-agar (X, solid curve). 
The data from these and all succeeding reactions have been converted to percentages of bound 
fragments in order to make direct comparisons between experiments. 



nucleotide sequences are eluted is approximation of relatedness be- 
indicative of mismatched sequences, tween the DNAs of phages A, 434, 
This point is being more closely in- and 434 hy. The data in Table 10 in- 
vestigated, dicate that \ is much more closely 
Phage-phage reactions. Binding related to 434 hy and 434 than to </>80 
percentages were obtained as a first or P 22 . Specificity of the method is 












° 6 



50 60 

Temperature, °C 

Fig. 67. Thermal elution profiles characteristic of reactions between E. coli DNA fragments 
and E. coli DNA-agar (open circles, dashed curve); 434 hy DNA fragments, and 434 hy DNA- 
agar (closed circles, solid curve) and 434 hy DNA fragments and E. coli DNA-agar (X, solid 



TABLE 1 0. Percentage of DNA Binding among Temperate Coliphages 

Labeled DNA 

Aga r-l mmobilized DNA 

434 hy 434 080 


Phage 15 
(TAU)- T 4 













' The data for c/>80, P22/ and T 4 are from Cowie, Year Book 65, pp. 106—123. 


A*/A —ft 

50 60 

Temperature, °C 


Fig. 68. Elution profiles obtained from intraspecies temperate coliphage DNA-agar reac- 
tions: A*/A (open circles, dashed curve); 434 hy*/434 hy (closed circles, solid curve); 434*/ 
434 (X, solid curve). 



shown by the fact that neither A nor 
434 DNA can pair with DNA from 
the blue-green alga Plectonema bor- 
yanum. Lambda DNA cannot pair 
with DNA from phage 15(TAU) _ or 
from T 4 . Direct quantitative compar- 
isons of the A, 434, and 434 hy data 
were precluded because of differences 
in ratios of labeled DNA fragments 
to DNA in the agar. It is clear, how- 
ever, that A and 434 hy DNA bind to 
a large extent as would be expected, 

since Kaiser and Jacob 55 isolated 434 
hy after a series of crosses designed 
to exclude all but l%-2% of the 434 
genome. The close relatedness be- 
tween A and 434 is borne out by the 
high incidence of pairing between the 
DNAs of these phages. 

Elution profiles from 434*/434 and 
434 hy*/434 hy DNA reactions are 
seen in Fig. 68. The A*/A DNA pro- 
file shown for comparison was ob- 
tained by using the more restrictive 

Temperature, °C 

Fig. 69. Reproducibility of thermal elution profiles obtained from reactions between 434 
DNA fragments and 434 DNA-agar. The three experiments were carried out with two different 
DNA-agar preparations and two different labeled DNA fragment preparations. 



elution schedule starting: at 50°. The 
434 hy and 434 DNA profiles are 
quite similar; both of these DNAs 
elute at a higher temperature than A 
DXA. There is little or no fragment 
elution at low temperatures, indicat- 
ing that under these incubation con- 
ditions there is no imprecise multiple 
copy DXA present in the genome of 
these phages. There is in fact no in- 
dependent evidence for the existence 
of imprecise multiple copy DNA in 
a phage genome. 

Phage 434 DNA has approximately 
51.5 c 'c G + C as determined from 
Falkow's detailed optical analysis of 
thermal denaturation data. Since 434 
hy DXA has a higher maximum elu- 
tion temperature than that of A, it is 
possible that 434 hy has either picked 
up some high G + C DNA from its 
434 parent or has had some low G + 
C A DNA deleted (assuming its ge- 
nome size is comparable to the 3 X 
10 : Dalton genome size of A DNA) . 

Sensitivity and reproducibility of 

elution profiles from DNA-agar du- 
plexes are dependent, to a large de- 
gree, upon the thoroughness of wash- 
ing at each temperature increment 
and upon the accuracy with which 
temperature is controlled. The repro- 
ducibility observed in these experi- 
ments is shown in Fig. 69. The major 
434*/434 DNA peaks are quite re- 
producible; however, the peak at 64° 
is distinct in one curve, is present as 
a shoulder in a second, and has been 
obscured in a third. Reproducibility 
in these experiments appears to com- 
pare favorably with that of detailed 
optical determinations of DNA melt- 
ing curves (Fig. 53). 

Interspecies phage reactions give 
no indication of imprecise DNA pair- 
ing in either the main areas of re- 
action or in DNA that elutes at low 
temperatures. The A*/434 hy DNA 
profile (Fig. 70<x) shows a distinct 
A-like peak at 70° and a 74° peak 
characteristic of 434 hy DNA. A 434 
hy*/434 hy DNA reaction is pre- 


50 60 70 80 40 

50 60 70 
Temperature, °C 

80 40 

50 60 70 

Fig. 70. Thermal elution profiles characteristic of interspecies phage DNA-agar reactions, 
a. 434 hy DNA-agar reacted with DNA fragments (open circles, dashed curve) and with 434 hy 
DNA fragments (closed circles, solid curve), b. 434 DNA-agar reacted with A DNA fragments 
(open circles, dashed curve) and with 434 DNA fragments (closed circles, solid curve), c. 434 
DNA fragments and 434 hy DNA-agar (open circles, dashed curve); 434 hy DNA fragments and 
434 DNA-agar (closed circles, solid curve). 



sented for comparison. Figure 706 
shows a A*/434 DNA profile which 
also contains the characteristic 70° 
A-like peak. It is compared with a 
434 */434 DNA profile. The reciprocal 
434 hy*/434 and 434/434 hy DNA 
profiles that peak at 72° are shown 
in Fig. 70c. The temperatures at 
which major elution peaks were ob- 

tained in the reactions studied thus 
far are summarized in Table 11. 

A more sensitive means of deter- 
mining the precision of nucleotide 
sequence pairing is to reisolate the 
duplexed fragments from an inter- 
species phage reaction and react them 
in the intraspecies system. The exper- 
iment was carried out with labeled 










434 *e luted from 
434 hy agar / 434 




50 60 

Temperature, °C 


Fig. 71. Thermal elution profiles of a typical intraspecies 434 DNA-agar reaction and an 
intraspecies 434 DNA-agar reaction in which the 434 DNA fragments had been previously bound 
to and reisolated from 434 hy DNA-agar. 



fragments eluted from a 434*/434 hy 
DXA reaction. The fragments were 
ethanol precipitated, concentrated, 
thermally denatured, and then incu- 
bated with 434 DNA agar. The re- 
sulting elution profile, shown in Fig. 
71, closely parallels the profile ob- 
tained from elution of phage 434 
DXA duplexes. Therefore the preci- 
sion of the reassociated complemen- 
tary nucleotide sequences between 
434 and 434 hy mimics that of the 
reassociated 434 sequences. 

It should be noted that a small de- 
gree of mismatched base pairs in nu- 
cleotide strands would not be distin- 
guishable from the data presented in 
thermal elution profiles. The high de- 
gree of pairing between X and 434 
DNA should have made this an ideal 
system in which to detect divergence 
among nucleotide sequences. It is 
clear, however, that no great impre- 
cisions exist within the duplexed 
portion of X and 434 DNA. 

The data describing the precise 
complementarity in interspecies du- 
plexes may be summarized: 

1. E. coli*/S. typhimurium DNA 
duplexes contain mainly or com- 
pletely imprecise pairs. 

2. Phage-host DNA reactions ex- 
hibit mainly precisely paired nucleo- 
tide sequences; however, there are 
imprecisely paired sequences, as evi- 
denced by those pairs that dissociate 
at low temperatures. 

3. The interspecies phage DNA re- 
actions examined to date showed no 

evidence of imprecise nucleotide se- 
quence pairing. 


A. V. Rake and R. B. Roberts 

For the past 10 years the Biophys- 
ics Section has maintained close con- 
tact with the studies of Dr. L. B. 
Flexner and Dr. J. B. Flexner, car- 
ried out at the University of Pennsyl- 
vania. During this period they have 
determined rates of protein synthesis 
in newborn and adult mice and more 
recently have uncovered the remark- 
able effect of puromycin in blocking 
the memory of maze training. During 
the past year this continuing collabo- 
ration has been supplemented by the- 
oretical and experimental studies car- 
ried out at this laboratory. 

Theoretical. The arrival of the IBM 
1130 computer evoked the writing of 
a program that might simulate the 
behavior of a network of 100 inter- 
connected neurones. The object of this 
exercise was to begin with very simple 
rules of connection and add elabora- 
tions as required to approach perfor- 
mance simulating adaptive behavior. 

In brief, the computer reads the 
initial conditions from punched cards 
and then proceeds to compute the 
state of the network at one instant 
from the external signals applied and 
the state of the network at the pre- 
vious instant (Fig. 72). If a cell has 
fired on the previous cycle it is con- 

TABLE 1 1 . Thermal Elution Peaks from Phage DNA-Agar Reactions 






434 hy 



P 2 2 

E. coli 


64, 70 

70, 74 




66, 70 

434 hy 


70, 74 


. . . 

. . . 






. . . 



58, 74 

. . . 

. . . 




P 2 2 



















Fig. 72. Initially all the cells of the network are considered inactive as indicated in cycle 0. 
On the first cycle the computer reads a card indicating what input stimulation has been ap- 
plied; the cells receiving inputs are considered firing and indicated by the numeral 2. On the 
second and succeeding cycles the computer reads additional cards indicating further stimulation, 
and calculates which cells fire as a result of signals transmitted to them by cells fired on the 
previous cycle. These are shown by the 1 in cycle 2. After 10 cycles the results are printed out. 

sidered refractory and will not fire 
on the present cycle; if a cell receives 
an external input (as read from a 
card) it fires unless it is refractory; 
if neither of these conditions applies, 
the signal delivered by each of the 
cells that did fire on the previous 
cycle is calculated and summed. If 
this sum exceeds a designated firing 
level the cell is noted as fired and the 
computer proceeds to the next cell. 
After completing the calculations for 
10 cycles the results are printed out. 
The key element of the program 
lies in the rules determining the in- 
terconnections of one cell and an- 
other, i.e., the synapses. With simple 
geometrical rules (Fig. 73) the net 
is highly unstable. An external input 
dies away rapidly if the firing level 
is high or expands to fill the net when 
the firing level is below a critical 
level. This inherent instability was 
controlled by introducing a variable 
firing level (increasing if too many 
cells fire) and by adding inhibitor 
cells. With these additional features 
in the program the propagation of 
diffuse waves through the network 
could be observed. Such a variable 
firing level might occur in the brain 

if excessive firing reduced the avail- 
able energy supply or released inhib- 
itors. Furthermore, the fluctuating 
sensitivity due to oscillations of the 
control system might possibly be re- 
lated to the a waves of the brain. 

I 00000000 

|0| 12222210 


301 12333210 

00 1 123x32101 

u 1 23 3 3 2 1 ' 


=>0 J 1 222 2 2 1 
Z0 | 1 1 1 1 1 1 1 




"~0 1 00000 

o u 


Fig. 73. A simple geometrical rule for inter- 
cell connections. A cell (indicated by X) is 
connected to each of its nearest neighbors by 
a strength of 3, to its next nearest neighbors 
by a strength of 2, etc. The column of cells at 
the left is considered reserved for input signals 
and does not receive signals from other cells of 
the network. The output column does not de- 
liver signals to other cells of the net. This iso- 
tropic network is highly unstable and confusing 
because it allows propagation in all directions. 



These diffuse waves of activity 
might give some primitive organism 
a repertoire of different response pat- 
terns corresponding to different stim- 
uli and might be representative of 
one characteristic feature of the 
brain. Another, possibly more inter- 
esting, property is propagation along 
more clearly defined channels. These 
channels suggest strongly that cer- 
tain sequences of connections are 
much more effective than others. In 
adding such a feature to the program 
a choice must be made whether to 
specify the strength of each inter- 
connection or to specify properties of 
the cells which then define the 
strength of the connections. 

YVe have chosen the second alter- 
native from considerations of inher- 
ited behavior patterns. Certain se- 
quences of synapses appear to be 
effective and these are designated by 
the genes of the organism. As the 
genes' best understood action is the 

determination of the distribution 
(i.e., quantity and types) of proteins 
within a cell, we have postulated that 
synapses are effective when they have 
similar proteins on both faces. Thus 
the DNA would not need to carry 
information concerning each synapse 
but only enough to designate the pro- 
tein distribution. In applying this 
principle to the computer, each cell 
is assigned a complement of proteins 
as shown in Fig. 74, and the strength 
of the synapse between two cells is 
made proportional to the number of 
similar proteins that the two cells 
hold in common. Thus the distribu- 
tion in the figure gives strong chan- 
nels horizontally as these cells have 
80 units in common. 

The same basic mechanism allows 
the introduction of learning into the 
program. A modification is made 
whereby the connections are made 
stronger by use. If one cell fires and 
a second fires in response, then on 







4 5 6 7 












































Fig. 74. Each cell is initially assigned a certain complement of proteins. The strength of the 
connection from one cell to another is then measured by the number of protein units held in 
common. The distribution illustrated here provides strong channels for propagation horizontally, 
as all cells lying in one row (for example, cells 2, 1 2, 22) have 80 units in common. Cells in 
the same column (1, 2, 3, 4, etc.) have 20 units in common, giving a weaker channel in the ver- 
tical direction. The distribution changes as a result of firing as described in the text. 



the next cycle a fraction of the pro- 
tein complement of the firing* cell is 
transferred to the responding cell. 
The protein distributions thus become 
more similar and the connection is 
accordingly strengthened. 

With these features in operation 
almost any desired channels of prop- 
agation can be developed in the net- 
work by a short period of "training." 
Reverberating rings are of particular 
interest, as they allow a single stimu- 
lus to initiate a prolonged response. 
In addition, the formation of a con- 
ditioned reflex has been simulated by 
firing first one and then another of 
two adjacent channels. 

The operation of the network to 
date suggests other developments, 
first, in a better simulation of learn- 
ing and memory and, second, in mov- 
ing further toward adaptive behavior. 
Connections between, cells in which 
the strength is determined by the 
number of proteins in common at the 
two faces of the synapse provide a 
satisfying simulation of inherited be- 
havior and possibly of long-term mem- 
ory. The stability of such a connec- 
tion does not depend on the stability 
of any particular molecule but only 
on the stability of the state of differ- 
entiation of the cells. Presumably this 
stability in turn depends on the sta- 
bility of a self-sustaining system. 
Furthermore, a slight modification of 
this concept is adequate to include 
the behavior patterns that are evoked 
by the administration of hormones. 
Another class of proteins can be 
added to the cell with the proviso 
that common proteins of this class 
will cause a strong connection be- 
tween the cells if and only if a hor- 
mone (or its products) is present. 
Thus the DNA could designate a se- 
quence of synapses that were inactive 
until made effective by a hormone. 

The device needed to simulate 
evoked behavior suggests a similar 
mechanism capable of simulating the 

short-term learning (and memory) 
that can occur during inhibition of 
protein and RNA synthesis. Let us 
postulate a third class of proteins 
that requires a factor X if the pro- 
teins are to facilitate a synapse. X is 
assumed to be present at all times 
but enters the synapse only when the 
synapse has transmitted a pulse. The 
short duration of such a memory im- 
plies that X has a short half-life in 
the synapse. Furthermore, the en- 
hanced ability to learn "significant" 
events can be attributed to an in- 
creased concentration of X released 
as a consequence of events inter- 
preted as "significant." Unfortu- 
nately, this type of memory is difficult 
to include in the program, as it re- 
quires the computer to store data 
describing the state of each individual 

A different direction of evolution 
would lead toward adaptive behavior. 
With the present program the net- 
work learns to repeat whatever it has 
done whether the outcome is good or 
bad. The addition of a "pain center" 
could serve to inhibit activity and 
thus prevent any further learning of 
inappropriate responses. 

If the outcome is not evident until 
the action is complete, adaptive be- 
havior is more difficult to achieve. As 
a minimum it seems necessary to 
store the stimulus and the response 
taken until the outcome becomes 
known. At that time favorable stimu- 
lus-response patterns would be trans- 
ferred to a long-term memory and 
the unfavorable ones discarded or in- 
hibited. Such a mechanism suggests 
the need for short- and long-term 
memories in the brain but is difficult 
to incorporate in a small program. 

The principal result of these exer- 
cises with the computer has been a 
sharpening of our concepts of mem- 
ory mechanisms. Farley has sum- 
marized this value: "To use the 
computer is to require precision of 



statement." We concur, and add a 
corollary that writing the program is 
sufficient for most purposes; its ac- 
tual execution on the computer fre- 
quently adds little. 

Experimental work. The objective 
of our experimental work has been 
to supplement the investigations of 
the Flexners. Several years ago the 
assumption was made that the long- 
term stability of a well-established 
memory might be similar to the sta- 
bility of the pattern of protein syn- 
thesis in a differentiated cell. A 
learning experience might trigger the 
synthesis of new proteins and RNA. 
If these products included an inducer 
(or derepressor) then the synthesis 
of the new products would be self- 
sustaining and the cell would assume 
a slightly altered state of differentia- 
tion. Early experiments, in which the 
Flexners observed that injections of 
puromycin (an inhibitor of protein 
synthesis) blocked the expression of 
memory, were consistent with this 
hypothesis. Furthermore, increased 
synthesis of RNA after a learning 
experience was reported by others. 
Accordingly, we hoped to observe the 
properties of this RNA by using the 
hybridizing techniques that had 
proved so useful in other studies. 

As an initial step the RNA of the 
brain was to be compared with RNA 
of other organs. Exploratory experi- 
ments carried out several years ago 
had indicated that the mRNA of the 
brain had little in common with liver 
mRNA. If these experiments were 
promising then RNAs from different 
regions of the brain and RNAs pro- 
duced after a learning experience 
could be characterized. 

Unfortunately, these efforts have 
encountered a number of difficulties. 
RNA precursors (including formate, 
orotic acid, and uridine) when in- 
jected intraperitoneally (i.p.) contrib- 
ute very little to the RNA of the brain. 
The resulting radioactivity is barely 

sufficient as a measure of RNA syn- 
thesis and far below the level needed 
for hybridization experiments. For 
this reason most experimenters have 
used intracranial injections (i.e.) of 
the RNA precursors. With this pro- 
cedure the resulting specific radioac- 
tivities are higher but still marginal 
for use in hybridization experiments. 

An alternative technique (of lim- 
ited application) is to use unlabeled 
RNA from the brain as a competitor 
to labeled RNA from other sources. 
As shown in Fig. 75 competition does 
occur, indicating that brain RNA 
shares common sequences with RNA 
of liver. The fraction of RNA that 
does hybridize in experiments is al- 
ways low and comprises not a ran- 
dom sample of the RNA but a special 
fraction. Therefore newly formed 
RNA made as a result of learning 
would be detectable only if it hap- 
pened to fall in this restricted class. 
In summary, these experiences with 
hybridization give no encouragement 
to the belief that this technique will 
be of much value in studies of the 
formation of memory. 

In the course of making numerous 
i.e. injections of radioactive uridine 
we noted that the region labeled was 
critically dependent on the depth of 
injection (Table 12) . It seemed there- 
fore that measurements of RNA syn- 
thesis during a learning experience 
could be subject to error if the in- 
jection were made i.e. Furthermore, 
it did not seem that the mice which 
had just undergone such a drastic 
experience could be considered "nor- 
mal" in their reaction to the training. 
Other labeling procedures will have 
to be investigated. 

In the year since these experiments 
started several findings have altered 
our views of the molecular events in- 
volved in memory formation. Mice 
can learn and remember (for a short 
time at least) during inhibition of 
protein and RNA synthesis by puro- 



TABLE 12. Incorporation of C 14 Uridine, Injected Intracranial^, into Mouse Brain Subfractions 

2 mm 

4 mm 













































Nerve cord 

Corpus striatum 
Frontal cortex 
Temporal cortex 
Parietal cortex 

In this experiment 1 A of C 14 uridine, 2 /xc, was injected bilaterally at the indicated depth 
beneath the surface of the frontal lobe of the brain. Ninety minutes later the brain was collected, 
the tissue dissected, pooled when bilateral, weighed, and frozen. The fragments were extracted 
by the usual Schmidt-Thannhauser method and the radioactivity in the nucleic acid fraction 
determined. The results are expressed as cpm/mg of wet tissue for two animals. 


300 500 

Competitor RNA 

Fig. 75. Competition of unlabeled normal mouse brain RNA with P 32 pulse-labeled liver RNA. 
80y of 1.5 hours labeled P 32 liver RNA (2000 cts/min/O.D. 26 o) was incubated with 0.25 g of 
agar containing 1 OOy mouse DNA for 18 hours at 60°C in the presence of unlabeled RNA from 
normal mouse brain (two preparations) and with unlabeled liver and kidney RNA for comparison. 
The relative radioactivity bound to the DNA was determined as percentage compared to no com- 
peting unlabeled RNA. The P 32 bound to agar without DNA is also indicated. 



myein and actinomycin D. Acetoxy- 
cycloheximide, a stronger inhibitor 
of protein synthesis than puromycin, 
differs from puromycin, as it does 
not block memory and in fact pre- 
vents blocking by puromycin when 
both are injected together. The action 
of puromycin is not to erase a mem- 
ory as was originally thought, but to 
block its expression ; the memory can 
be restored by a subsequent injection 
of saline. For short-term memory at 
least, synthesis of the macromolecules 
does not seem to be required. Fur- 
thermore, other reports have indi- 
cated drastic effects of hormones on 

the retention of memory. Possibly the 
search for the molecules involved in 
memory formation should be shifted 
toward peptides. 

In summary, the past year has 
given us some first-hand experience 
in a new field and we can only hope 
that the next year will prove more 
fruitful. If new molecules are pro- 
duced by numerous cells of the brain 
as a result of learning, they should 
be detectable; if only a few cells are 
involved or if existing materials are 
simply released, the search will be 
very much more difficult. 


1. Forbush, S. E., and L. Beach, Carnegie 
Inst. Wash. Year Book 65, 28-36, 1966. 

2. Babcock, H. W., Sci. Am., 202, 53-62, 

3. Ness, N. F., and J. M. Wilcox, Phys. 
Rev. Letters, 13, 461-464, 1964. 

4. Parker, E. N., Astrophys. J., 128, 664- 
676, 1958. 

5. Wilcox, J. M., and N. F. Ness, J. 
Geophys. Res., 70, 5793-5805, 1965. 

6. Venkatesan, D., and A. Dattner, Tellus, 
11, 116-129, 1959. 

7. McCracken, K. G., V. R. Rao, B. C. 

Fowler, M. A. Shea, and D. F. Smart, 
IQSY Instruction Manual No. 10, 1965. 

8. Herrin, E. T., and J. Taggart, Bull. 
Seis. Soc. Am., 52, 1037-1046, 1962. 

9. Tatel, H. E., L. H. Adams, and M. A. 
Tuve, Proc. Am. Phil. Soc., 97, 658-669, 

10. Hart, P. J., Variation of velocity near 
the Mohorovicic discontinuity under 
Maryland and northeastern Virginia, 
Ph. D. thesis, Harvard University, 1954. 

11. Press, F., J. Geophys. Res., 65, 1039- 
1051, 1960. 

12. Press, F., and S. Biehler, J. Geophys. 
Res., 69, 2979-2995, 1964. 

13. Smith, T. J., J. S. Steinhart, and L. T. 
Aldrich, ./. Geophys. Res., 71, 1141- 
1181, 1966. 

14. Steinhart, J. S., and R. P. Meyer, 
Carnegie Inst. Wash. Publ. 622, 1961. 

15. Carnegie Inst. Wash. Year Book 62, 
280, 1963. 

16. Borcherdt, R. D., and J. C. Roller, 
Technical Letter No. 43, U. S. Geologi- 
cal Survey, 1966. 

17. Lewis, B. T. R., and R. P. Meyer, Contr. 
167, Geophys. and Polar Res. Center, 
Dept. of Geo!., Univ. of Wise, 1966. 

18. Knopoff, L., Publ. 383, Inst. Geophys. 
and Planetary Phys., Univ. of Calif., 
Los Angeles, 1966. 

19. Meyer, R. P., J. S. Steinhart, W. E. 
Bonini, B. F. Howell, Jr., D. E. Willis, 
and F. Novacheck, in Abstracts of 
Papers, vol. Ill, International Associa- 
tion of Seismology and Physics of the 
Earth's Interior (27), International 
Union of Geodesy and Geophysics, 13th 
Gen. Assembly, Berkeley, Calif., 1963. 

20. O'Brien, P. N. S., Boll. Geofisica T. ed 
A., 9, 22-65, 1967. 

21. Smith, T. J., J. S. Steinhart, and L. T. 
Aldrich, J. Geophys. Res., 71, 1141- 
1172, 1966. 

22. Berry, M. J., and G. F. West, Bull Seis. 
Soc. Am., 56, 141-171, 1966. 

23. Bancroft, A. M., Seismic spectra and 
detection probabilities from explosions 
in Lake Superior, in The Earth Beneath 
the Continents, J. S. Steinhart and T. J. 



Smith, eds. Amer. Geophys. Union 
monograph 10, 234-240, 1966. 

24. O'Brien, P. N. S., J. Geophys. Res., 72, 
2569-2575, 1967. 

25. Kanamori, H., J. Geophys. Res., 72, 
2181-2186, 1967. 

26. Sumner, R. D., Bull. Sets. Soc. Am., 57, 
173-190, 1967. 

27. Anderson, D. L., A. Ben-Menahem, and 
C. B. Archambeau, J. Geophys. Res., 70, 
1441-1448, 1965. 

28. McCamy, K., R. P. Meyer, and T. J. 
Smith, Bull. Seis. Soc. Am., 52, 923-955, 

29. Stueber, A. M., and V. R. Murthy, 
Science, 153, 740-741, 1966. 

30. Hart, S. R., Program annual meetings, 
Geol. Soc. Am., 86-87, 1964. 

31. Stockwell, C. H., Geol. Surv. Canada, 
Paper 64-17, Part II, 1964. 

32. Birch, F., Am. J. Sci., 252, 1-25, 1954. 

33. Roy, R. F., Heat Flow Measurements 
in the United States, Ph.D. Thesis, 
Harvard Univ., 1963. 

34. Lachenbruch, A. H., U. S. Geol. Surv. 
Bull., 1083-A, 36 pp., 1959 

35. Von Herzen, R., and A. E. Maxwell, J. 
Geophys. Res., 64, 1557-1563, 1959. 

36. Ratcliffe, E. H., J. Geophys. Res., 65, 
1535-1541, 1960. 

37. Burke, B. F., K. C. Turner, and M. A. 
Tuve, A high-resolution study of M31, 
in The Galaxy and the Magellanic 
Clouds, IAU-URSI Symp. 20, 99-102, 

38. Rubin, V. C, S. Moore, and F. C. 
Bertiau, S.J., Astron. J., 72, 59-64, 

39. Keyworth, G. A., G. C. Kyker, Jr., E. G. 
Bilpuch, and H. W. Newson, Nucl. 
Phys., 89, 590, 1966. 

40. Christian, R. S., and J. L. Gammel, 
Phys. Rev., 91, 100, 1953. 

41. Extermann, P., Nucl. Phys., A95, 615, 

42. Maio, J. J., and C. L. Schildkraut, J. 
Mol. Biol, 24, 29, 1967. 

43. Walker, P.M.B., and A. McLaren, 
Nature, 206, 1175, 1965. 

44. Flamm, W. G., M. McCallum, and 
P. M. B. Walker, Proc. Natl. Acad. Sci., 
57, 1729-1734, 1967. 

45. Martin, M., and B. H. Hoyer, Biochem., 
5, 2706, 1966. 

46. McClaren, A., and P. M. B. Walker, 
Genet. Res. Camb., 6, 1965. 

47. Simpson, G. C, in This View of Life, 
Harcourt, Brace and World Inc., N. Y., 
pp. 164-165, 1964. 

48. Falkow, S., J. A. Wholhieter, R. V. 
Citarella, and L. S. Baron, J. Bacteriol, 
87, 209, 1964. 

49. Falkow, S., R. V. Citarella, A. J. Whol- 
hieter, and T. Watanabe, J. Mol. Biol., 
17, 102, 1966. 

50. Signer, E. R., Virology, 22, 650, 1964. 

51. Kaiser, A. D., and D. S. Hogness, J. 
Mol. Biol., 2, 392, 1960. 

52. Bolton, E. T., and B. J. McCarthy, Proc. 
Natl. Acad. Sci., U.S., 48, 1390, 1962. 

53. Martin, M. A., and B. H. Hoyer, Bio- 
chem., 5, 2706, 1966. 

54. Brenner, D. J., M. A. Martin, and B. H. 
Hoyer, J. Bacteriol, 94, 486-487, 1967. 

55. Kaiser, A. D., and F. Jacob, Virology, 
4, 509, 1957. 


Adams, J., see Brown, L. 

Aldrich, L. T., see Asada, T., Hart, S. R., 
and Smith, T. J. 

Asada T., and L. T. Aldrich, Seismic obser- 
vations of explosions in Montana, in The 
Earth beneath the Continents, Geophysical 
Monograph 10, 382-390, J. S. Steinhart 
and T. J. Smith, eds., Wash., D. C, Amer- 
ican Geophysical Union, 1966. 

Asada, T., see also Hales, A. L. 

Baum, W. A., J. S. Hall, L. L. Marton, and 
M. A. Tuve, Committee on image tubes 
for telescopes, Carnegie Inst. Wash. Year 
Book 65, 191-194, 1966. 

Bendich, A. J., and E. T. Bolton, Related- 
ness among plants as measured by the 
DNA-agar technique, Plant Physiol., 42, 
959-967, 1967. 

Bertiau, F. C, S.J., see Rubin, Vera C. 



Bolton, E. T., Nucleic acid interactions: a 
molecular approach to the study of genes 
and their products, Cancer Res., 26, 1964- 
1970, 1966. 

Bolton, E. T., see also Bendich, A. J. 

Boyce, P. B., see Ford, W. K., Jr. 

Brenner, D. J., S. Falkow, and D. B. Cowie, 
DXA homologies among temperate coli- 
phages (abstract), Science, 156, 535-536, 

Britten, R. J., and D. E. Kohne, Repeated 
nucleotide sequences of DNA (abstract), 
Science, 156, 536, 1967. 

Britten, R. J., sec also Waring, M. J. 

Brown, L., W. Haeberli, and W. Trachslin, 
Phase shifts and polarization in proton- 
alpha scattering from 940 to 3200 keV, in 
Proceedings of the 2nd International Sym- 
posium on Polarization Phenomena of 
Nuclcons, Karlsruhe, Germany, Sept. 6- 
10, 1965, pp. 349-351, P. Huber and H. 
Schopper, eds., Basel and Stuttgart, Birk- 
hauser, Verlag, 1966. 

Brown, L., W. Haeberli and W. Trachslin, 
Phase shifts and polarization in proton- 
alpha scattering from 0.9 to 3.2 MeV, 
Nucl Plujs., A90, 339-352, 1967. 

Brown, L., W. Trachslin, D. Robson, and 
J. Adams, Polarization excitation of the 
1.85-MeV resonances in 40 Ar (p,p) 40 Ar 
(abstract), Bull. Am. Phys. Soc., 11, 840, 

Brown, L., and W. Trachslin, Polarization 
excitation of the 1.87 MeV resonances in 
40 Ar(p,p) *oAr, Nucl. Phys., A96, 238-240, 


Brown, L., and W. Trachslin, Scattering in 
polarized protons by helium from 0.9 to 
3.2 MeV, Nucl. Phys., A90, 334-338, 1967. 

Brown, L., see also Trachslin, W. 

Cowie, D. B., Genetic relationships among 
viruses and bacteria (abstract), Science, 
156, 537, 1967. 

Cowie, D. B., see also Brenner, D. J. 

Duggal, S. P., M. A. Pomerantz, and S. E. 
Forbush, Long-term variation in the mag- 
nitude of the diurnal anisotropy of cosmic 
rays, Nature, 211*, 154-155, 1967. 

Falkow, S., see Brenner, D. J. 

Flexner, J. B., see Flexner, L. B. 

Flexner, L. B., J. B. Flexner, and R. B. 
Roberts, Memory in mice analyzed with 
antibiotics, Science, 155, 1377-1383, 1967. 

Flexner, L. B., J. B. Flexner, and R. B. 
Roberts, Stages of memory in mice 
treated with acetoxycycloheximide before 
or immediately after learning, Proc. Natl. 
Acad. Sci. U. S., 56, 730-735, 1966. 

Forbush, S. E., Solar influences on the 
cosmic-ray diurnal variation (abstract), 
Science, 156, 537, 1967. 

Forbush, S. E., Variations from 1937 to 
1965 in the annual mean diurnal variation 
of cosmic-ray intensity from ionization 
chambers at three stations (abstract), 
Trans. Am. Geophys. Union, 48, 171, 1967. 

Forbush, S. E., see also Duggal, S. P. 

Ford, W. K., Jr., Astronomical uses of cas- 
cade intensifiers, in Advances in Elec- 
tronics and Electron Physics, Photo- 
Electronic Image Devices, vol. 22B, 697- 
704, J. D. McGee, D. McMullen, and E. 
Kalian, eds., New York, Academic Press, 

Ford, W. K., Jr., and P. B. Boyce, Infrared 
spectroscopy with a cascaded image in- 
tensifier (abstract), Astron. J., 71, 385, 

Ford, W. K., Jr., and Vera C. Rubin, Quasi- 
stellar objects with small redshifts 
1217+02, 3C 249.1, and 3C 263, Astro- 
phys. J., U5, 357-360, 1966. 

Ford, W. K., Jr., see also Rubin, Vera C. 

Haeberli, W., see Brown, L. 

Hales, A. L., and T. Asada, Crustal struc- 
ture in coastal Alaska, in The Earth be- 
neath the Continents, Geophysical Mono- 
graph 10, 420-432, J. S. Steinhart and 
T. J. Smith, eds., Wash., D. C, American 
Geophysical Union, 1966. 

Hall, J. S., see Baum, W. A. 

Hart, S. R., Radiometric ages in Uruguay 
and Argentina and their implications 
concerning continental drift (abstract) , 
Geol. Soc. Am. Program, 79th Ann. Meet- 
ing, San Francisco, Calif., p. 86, Nov. 
14-16, 1966. 

Hart, S. R., and L. T. Aldrich, Fractiona- 
tion of potassium/rubidium by amphi- 
boles: implications regarding mantle com- 
position, Science, 155, 325-327, 1967. 

Hart, S. R., J. S. Steinhart, and T. J. 
Smith, A regional heat flow survey in 
Lake Superior (abstract), Trans. Am. 
Geophys. Union, 48, 211, 1967. 

Hart, S. R., and G. R. Tilton, The isotope 
geochemistry of strontium and lead in 
Lake Superior sediments and water, in 
The Earth beneath the Continents, Geo- 



physical Monograph 10, 127-138, J. S. 
Steinhart and T. J. Smith, eds., Wash., 
D. C, American Geophysical Union, 1966. 

Hart, S. R., see also Ohmoto, H. 

Holland, H. D., see Ohmoto, H. 

James, D. E., Crustal structure of the Mid- 
dle Atlantic States (abstract), Earth- 
quake Notes, Eastern Section Seis. Soc. 
Am., 27, 16, 1966. 

James, D. E., T. J. Smith, and J. S. Stein- 
hart, Crustal structure of the Middle 
Atlantic States (abstract), Trans. Am. 
Geophys. Union, 48, 198, 1967. 

James, D. E., and J. S. Steinhart, Structure 
beneath continents: a critical review of 
explosion studies 1960-1965, in The Earth 
beneath the Continents, Geophysical 
Monograph 10, 293-333, J. S. Steinhart 
and T. J. Smith, eds., Wash., D. C, 
American Geophysical Union, 1966. 

Kohne, D. E., see Britten, R. J. 

Kouvo, O., and G. R. Tilton, Mineral ages 

from the Finnish Precambrian, J. Geol., 

74, 421-442, 1966. 

Marton, L. L., see Baum, W. A. 

Moore, Sandra, see Rubin, Vera C. 

O'Brien, P. N. S., Analysis of a small num- 
ber of seismic records along an E-W 
Alpine profile, Boll. Geofis. Teorica ed 
Appl, IX, 22-65, 1967. 

O'Brien, P. N. S., Quantitative discussion on 
seismic amplitudes produced by explosions 
in Lake Superior, J. Geophys. Res., 72, 
2569-2575, 1967. 

Ohmoto, H., S. R. Hart, and H. D. Holland, 
K-Ar and Rb-Sr ages of intrusive rocks 
and hydrothermal minerals in the Provi- 
dencia area, Mexico (abstract) , Geol. Soc. 
Am. Program, 79th Ann. Meeting, San 
Francisco, Calif., pp. 152-153, Nov. 14-16, 

Otsuka, M., Analysis of P-wave travel-time 
residuals in Peru (abstract), Earthquake 
Notes, Eastern Section Seis. Soc. Am., 37, 
24, 1966. 

Pomerantz, M.A., see Duggal, S. P. 

Roberts, R. B., see Flexner, L. B. 

Robson, D., see Brown, L. 

Rubin, Vera C, and W. K. Ford, Jr., Image 
tube spectra of quasi-stellar objects (ab- 
stract), Astron. J., 71, 396, 1966. 

Bertiau, S. J., Faint blue objects in the 
Virgo cluster region, Astron. J., 72, 59- 
123, 1967. 

Rubin, Vera C, see also Ford, W. K., Jr. 

Sacks, I. S., A broad-band large dynamic 
range seismograph, in The Earth beneath 
the Continents, Geophysical Monograph 
10, 543-553, J. S. Steinhart and T. J. 
Smith, eds., Wash., D. C, American Geo- 
physical Union, 1966. 

Sacks, I. S., Diffracted P-wave studies of 
the earth's core, 2, Lower mantle velocity, 
core size, lower mantle structure, J. Geo- 
phys. Res., 72, 2589-2594, 1967. 

Smith, T. J., Translation from Russian, 
Problems in Deep Seismic Sounding, 166 
pp., S. M. Zverev, ed., N. Y., Consultants 
Bureau, 1967. 

Smith, T. J., J. S. Steinhart, and L. T. Aid- 
rich, Crustal structure under Lake Supe- 
rior, in The Earth beneath the Conti- 
nents, Geophysical Monograph 10, 181- 
197, J. S. Steinhart and T. J. Smith, eds., 
Wash., D. C, American Geophysical 
Union, 1966. 

Smith, T. J., see also Hart, S. R., James, 
D. E., and Steinhart, J. S. 

Steinhart, J. S., and T. J. Smith, Foreword, 
in Problems in Deep Seismic Sounding , 
p. v, S. M. Zverev, ed., N. Y., Consultants 
Bureau, 1967. 

Steinhart, J. S., and T. J. Smith, Time 
terms and structure in western Lake 
Superior, in The Earth beneath the Con- 
tinents, Geophysical Monograph 10, 198- 
204, J. S. Steinhart and T. J. Smith, eds., 
Wash., D. C, American Geophysical 
Union, 1966. 

Steinhart, J. S., see also Hart, S. R., and 
James, D. E. 

Sumner, R. D., Attenuation of earthquake 
generated P waves along the western 
flank of the Andes, Bull. Seis. Soc. Am., 
57, 173-190, 1967. 

Suyehiro, S., A search for small, deep earth- 
quakes using quadripartite stations in the 
Andes, Bull. Seis. Soc. Am., 57, 447-461, 

Suyehiro, S., Quadripartite observations in 
the Andes (abstract), Earthquake Notes, 
Eastern Section Seis. Soc. Am., 37, 29, 

Rubin, Vera C, Sandra Moore, and F. C. Tilton, G. R., see Hart, S. R., and Kouvo, O. 



Trachslin, W., and L. Brown, Polarization 
in p-" ii; 0-scattering from 2.5 to 3.0 MeV 
(abstract), Bull. Am. Phys. Soc, 12, 86, 

Trachslin. W., and L. Brown, Scattering of 
polarized protons by deuterium from 1 to 
3 MeV, Nuclear Phys., A90, 593-596, 

Trachslin, W., see also Brown, L. 

Tuve. M. A.. Physics and the humanities — 
the verification of complementarity, The 
Third Cosmos Club Award, pp. 5-20, 
Cosmos Club, Washington, D. C, May 9, 
1966. Reprinted in The Search for Under- 

standing, pp. 41-55, C. P. Haskins, ed., 
Wash., D. C, Carnegie Institution of 
Washington, 1967. 

Tuve, M. A., Radio echoes (the origin of 
radar), in The Search for Understanding, 
pp. 73-76, C. P. Haskins, ed., Wash., 
D. C, Carnegie Institution of Washing- 
ton, 1967. 

Tuve, M. A., see also Baum, W. A. 

Waring, M. J., and R. J. Britten, Nucleotide 
sequence repetition: a rapidly reassociat- 
ing fraction of mouse DNA, Science, 154, 
791-794, 1966. 




Ellis T. Bolton 

Associate Director 
L. Thomas Aldrich 

Roy J. Britten 
Dean B. Cowie 
Scott E. Forbush 
W. Kent Ford, Jr. 
Stanley R. Hart 1 

Staff Members 

Richard B. Roberts 
I. Selwyn Sacks 
T. Jefferson Smith 
John S. Steinhart 

Louis Brown 
David E. Kohne 

Staff Associates 

Vera C. Rubin 
Kenneth C. Turner 

Section Chairmen 

Astrophysics: W. Kent Ford, Jr. Biophysics: Dean B. Cowie 

Geophysics : L. Thomas Aldrich 


Don J. Brenner, Fellow of the U. S. Pub- 
lic Health Service 

David E. James, Fellow of National 
Science Foundation and Carnegie In- 
stitution 2 

Akira Kamitsuki, Kansai University, 
Osaka, Japan 3 

Thomas E. Krogh, Massachusetts Insti- 
tute of Technology 4 

Fernando Munizaga, University of Chile, 
Santiago, Chile 5 

Peter N. S. O'Brien, British Petroleum 
Company, Ltd., Sunbury-on-Thames, 
Middlesex, England 6 

Michio Otsuka, Kumamoto University, 

Kumamoto, Japan 7 

Claude Petitjean, University of Basel, 
Basel, Switzerland 8 

Adrian V. Rake, University of British 
Columbia, Vancouver, B. C. 

German Saa, S. J., Universidad del Norte, 
Antofagasta, Chile 

Alan M. Stueber, Washington University 

Shigeji Suyehiro, Meteorological Re- 
search Institute, Tokyo, Japan 9 

Walter Trachslin, University of Basel, 
Basel, Switzerland 

Carlos Varsavsky, Instituto Argentino de 
Radioastronomia, Villa Elisa, Argen- 

Junior Felloiv 
Rodolfo Anzoleaga, Universidad Mayor de San Andres, La Paz, Bolivia 10 

1 On leave of absence from September 1, 

2 National Science Foundation to Novem- 
ber 15, 1966. Carnegie Institution from 
November 16, 1966. 

3 Deceased August 21, 1966. 

4 Through July 31, 1966. 

5 Through January 31, 1967. 

6 From September 1, 1966. 

7 Through October 31, 1966. 

8 From April 8, 1967. 

9 Through December 31, 1966. 
10 Through December 31, 1966. 



Collaborators and Visiting Investigators 

J. L. Adams, Florida State University 

Pablo Aparicio, Washington University 

David Axelrod, National Institutes of 

Arnold Bendich, University of Wash- 

Bernard F. Burke, Massachusetts Insti- 
tute of Technology 

Ramon Cabre, S.J., Observatorio San 
Calixto. La Paz, Bolivia 

Bayne Carew, University of Wyoming 

Mateo Casaverde, Instituto Geofisico del 
Peru, Lima, Peru 

Salvador del Pozo, Instituto Geofisico 
Boliviano, La Paz, Bolivia 

Luis Fernandez, S.J., Observatorio San 
Calixto, La Paz, Bolivia 

Louis B. Flexner, University of Penn- 

J. Flexner, University of Pennsylvania 

Enrique Gajardo, University of Chile, 
Santiago, Chile 

Albert Gelderman, National Institutes of 

Alberto A. Giesecke, Instituto Geofisico 

del Peru, Lima, Peru 
Willy Haeberli, University of Wisconsin 
Bill Hoyer, National Institutes of Health 
Brian Lewis, University of Wisconsin 
Malcolm Martin, National Institutes of 

Martin F. McCarthy, S.J., Observatorio 

Astronomico, Specola Vaticana, Italy 
Leonard Miller, U. S. Naval Medical Re- 
search Center 
Alois Th. Purgathofer, Universitat 

Sternwarte, Vienna, Austria 
D. Robson, Florida State University 
Reynaldo Salgueiro, Instituto Geofisico 

Boliviano, La Paz, Bolivia 
Richard G. Seyler, Ohio State University 
Fernando Volponi, Universidad Nacional 

de Cuyo, San Juan, Argentina 

Engineer and Research Assistant 
Everett T. Ecklund 

John B. Doak 
Paul A. Johnson 

Research Assistants 

Charles A. Little 

Liselotte Beach 
H. Lowell Belin 1X 
Kenneth D. Burrhus 12 
Stephen J. Buynitzky 

Laboratory Assistants 

Margaret E. Chamberlin 
Louise E. Magruder 13 
Glenn R. Poe 
Neltje W. van de Velde 


Chief, Fiscal Section: Helen E. Russell 
Office Manager: William N. Dove 
Librarian : Lelah J. Prothro (part time) 
Secretary: Claudine C. Ator 

Stenographers: Dorothy B. Dillin, E. 

Kathleen Hill 
Typist: Mary T. Sheahan (part time) 
Accounting Assistant: Glenda J. Johns- 


Chief Instrument Maker and Shop Man- 
ager : John G. Lorz 

Instrument Makers : Robert Hoffmaster, 
Michael Seemann 

Machinist-Instrument Makers 

Mossor, 14 Carl M. Rinehart 
Machinist: Francis J. Caherty 

Doy E. 

11 From January 16, 1967. 

12 From February 1, 1967. 

is Through February 28, 1967. 
14 Through October 31, 1966. 



Buildings and Grounds 

Carpenter and Maintenance Foreman : Caretaker : Stanley Gawrys 

Leo J. Haber Assistant Caretakers : Raymond L. Bat- 
Assistant Maintenance Foreman: Elliott tie, 15 Bennie Harris, 10 Willis Kilgore, 
M. Quade Jr. 17 

Carol Anfinsen 
Robert Cadmus 
Dorothy Canter 
Allen Forsbacka 
John Harvey 
Samuel Humphrey 
Sandra Moore 

Part-Time and Temporary Employees 

Milan Pavich 
John Roddy 
Martin Roddy 
Paul Roddy 
Robert Singleton 
Doris Titus 
Anne Unger 

15 July 19, 1966-February 28, 1967. 

16 From March 1, 1967. 

17 From December 19, 1966. 

Plate 1 

Department of Terr en trial M aynfJAxm 


" PMG20 Aug 1965 

h= 129Km,A = 134.5% m = 6.0, 19.03 s, 69.07 w *68, T 09 h 42 m 48.5% 
PP 10 h 04 m 27 s (T-JB = -3), SKP 10 h 05 m 10 8 (T-JB = -3), sSKP 10 h 06 m 13 s 
(T - JB = 0) 

Plate 1. SKP and sSKP arrivals at PMG standard station. The upper trace is the long- 
period vertical component; the lower trace, the short-period vertical. Note (upper trace) that 
sSKP is a mirror image of SKP. The bars on the lower trace show the position and duration of 
the longest of the time windows used. The center bar indicates the section of the record used 
to determine the noise. Both record sections show minute marks. 

Plate : 

Department of Terrestrial Magnetism 

Plate 3 

Department of Terrestrial Magnetism 

Deep focus of Santiago del Estero 

Zonda station 

20th December 1966 


P 12 h 28 m 46 s GMT (corrected) 




of energy 

: ?3 



P (400) 

Recording velocity 44.5 mm/min 
Mechanical seismograph T<> = 3 sec 

Plate 3. A clear example of the S-P converted phase. 

Plat, - A 

Department of Terrestrial Magnetism 

^tt|^ V >^; N tz z 


*>, ,. k .„^s 


^ o 

Plate 4. Seismograms obtained at Zanda (Z), Pie de Palo [PP) f and Hualilan (H). March 6 
and 7, above; February 21 and 23, opposite. 

Plate UB 

Department of Terrell/rial, MofjueUyrn. 

»*,/■■„.. X""**S 


'■■""C'fi" ■ 

fy-ff'/fy* ■ #" -%;■*& 

l^.*-*--^_|t^— ^~^ --"»■' -wf4/>^"" --■-^'-|4*"- iw *' ,i ^ ! *4 



W^°'iSfM Jfcy*^ 





Plate 5 

Department of Terrestrial Magnetism 

S E 


54 01 













































































































































































































Department of Terrestrial Magnetism 








lift*. % ^.. • » . ■ . 

f"r*r >3j, ,, 

' ■:■■:..> ■ 


. ' '■ 

iifii3iii : |iBi||^sSS| 

,v- ; : 


• ^~ . - ' *- 

He . ' -,: - 

' s 

•■■■ ■ ■ : ' ' ■ ■ ' 

'v. ■ : -■ - 


60 70 80 

Melting temperature 

Plate 6. Schematic diagram suggesting the history of families of repeated DNA sequences. 
The left-hand scale indicates the period of time since the saltatory replication that produced 
each family. The lower scale — as a measure of divergence — represents the thermal stability 
of the reassociated pairs of DNA strands formed by the members of a family. The height of 
the peaks indicates the amount of DNA of a given age and thermal stability. 

Committee on Image Tubes 
for Telescopes 

Cooperative Project of Mount Wilson and Palomar Observatories 

Department of Terrestrial Magnetism, Lowell Observatory 

National Bureau of Standards, and United States Naval Observatory 

W. A. Baum 
Lowell Observatory 

John S. Hall (Chairman) 

Director, Lowell Observatory 

Flagstaff, Arizona 

L. L. Marton 

National Bureau of Standards 

M. A. Tuve 

Department of Terrestrial Magnetism 


Review of Activities 135 

Allocation of Intensifier Systems 135 

Spectrograph Cameras for Image Tubes 136 

Radial Velocities from Image Tube Spectra 137 

Acknowledgments 140 

Carnegie Institution Year Book GO, 1966-1967 

Image Tubes for Telescopes 



















P^O cp 
r-cr>cn, <r> 

tO^OiD to 




o *• 

ro ro 

CO U> 






A night-sky spectrogram obtained with an RCA cascaded image tube with the Bowen cassegrain 
schmidt camera on the DTM spectrograph. For accurate radial velocity work it has been 
necessary to allow for the distortions introduced by the magnetically focused tube. These dis- 
tortions are shown in an exaggerated form in the schematic diagram. 

Schematic diagram showing distortions greatly 
exaggerated in image tube spectrum. 


The Carnegie Image Tube Commit- several groups have constructed fast, 

tee has continued its successful dis- high-resolution spectrograph cameras 

tribution program of image tube sys- designed specifically for image tube 

terns to astronomical observatories, use. One such camera, designed by 

using special funds granted by the Dr. I. S. Bowen, has been used by 

National Science Foundation for this Baum, Ford, and Rubin as part of the 

purpose. These systems are built Committee's program of tests and 

around the cascade image intensi- demonstrations of image tube tech- 

fiers developed for the Committee niques. This camera, a cassegrain 

over a period of years by the Electron schmidt system, provides much better 

Tube Division of the Radio Corpora- image tube spectra than those ob- 

tion of America. In addition to the tained previously with ordinary 

intensifying tube, the system uses a lenses. 

permanent magnet for focusing, a It is becoming increasingly impor- 
mechanical mounting with sufficient tant, as more astronomical observa- 
insulation for the required 20 kV, a tions are being made with the aid of 
voltage divider, and a transfer lens image tubes, to know in detail the 
and focusing device. Hence all of the photometric properties and the di- 
basic parts needed to operate the in- mensional characteristics of the in- 
tensifying system are furnished to tensifier systems. Low-dispersion 
each observatory. With the intensify- spectrograms obtained by Ford and 
ing system, astronomical observations Rubin with the DTM image tube spec- 
can be made that are beyond the ca- trograph have been measured, and 
pability of ordinary photographic Dr. Rubin has worked out a radial 
techniques used on the same telescope, velocity reduction program that cor- 

At present the most important as- rects for the small distortions intro- 

tronomical application of image tubes duced by the image tube system. The 

is in spectroscopic observations. Un- results indicate that accurate radial 

fortunately, image tube systems are velocities can be obtained from these 

too bulky to be substituted directly plates if a calibration plate is used 

for photographic plate holders in to determine the distortions. It has 

spectroscopic cameras. Many of the been demonstrated that, for exposures 

cameras in question are fast schmidt of several hours, in most applications 

systems with the focus located inter- no magnetic shielding is required 

nally and hence inaccessible for bulky around the RCA cascaded tube, 
image tube equipment. Therefore, 


With the support of the National are manufactured to the rather high 

Science Foundation, the Carnegie performance specifications set by the 

Committee has been able to purchase Committee, are individually tested 

a number of cascaded image intensi- and each one is fitted into a complete 

fiers (RCA type C33011) and to have intensifier system at DTM by Dr. 

the basic hardware required for op- Ford. During the report year eight 

erating these tubes fabricated in va- such systems were distributed and in- 

rious small shops. The tubes, which stalled for the Committee by Dr. 




Ford and Dr. Purgathofer. Four 

other tubes, with some auxiliary 
parts, were made available on tempo- 
rary loan for special testing programs 
for the Committee. 

Allocation of the Carnegie-NSF 
systems is handled by a joint NSF- 
Carnegie Allocations Committee un- 
der the chairmanship of Dr. John 
Hall. Lowell Observatory. This year 
sets were lent to Dr. Thornton Page, 
Wesleyan University, for use at the 
Astronomical Observatory of the 
Universidad Nacional de Cordoba, 
Cordoba, Argentina; to Dr. F. Ber- 
tola, Asiago Astrophysical Observa- 
tory of the University of Padova, 
Padova, Italy; to The Astronomer 
Royal, Sir Richard Woolley, Royal 
Greenwich Observatory, Herstmon- 
ceux Castle, Hailsham, Sussex, Eng- 
land ; to Dr. K. Osawa, Tokyo Astro- 
nomical Observatory, Okayama, 
Japan; to Dr. A. Purgathofer, Uni- 
versitats-Sternwarte, Vienna, Aus- 
tria ; to Dr. C. R. O'Dell, Yerkes Ob- 
servatory; to Dr. Harlan J. Smith, 
University of Texas (SI system) ; 
and to Dr. Arne Sletteback and Dr. 
Paul L. Byard, Perkins Observatory, 
Ohio State and Ohio Wesleyan Uni- 
versities. Previous allocations have 

been to Yerkes Observatory, Kitt 
Peak National Observatory, Lick Ob- 
servatory, Lowell Observatory, Mount 
Wilson Observatory, Flagstaff Sta- 
tion of the U.S. Naval Observatory, 
McDonald Observatory, Mount 
Stromlo Observatory, Steward Ob- 
servatory, Mount Wilson and Palo- 
mar Observatories, High Altitude 
Observatory, and the Leander McCor- 
mick Observatory. The sets sent out 
this year bring the number of com- 
plete cascaded tube systems that have 
been distributed to 21. 

Astronomers at these observatories 
are using image tubes for observa- 
tions in a wide variety of spectro- 
scopic programs at both high and 
low dispersions. Problems under 
study include high-dispersion radial 
velocities of interstellar lines and 
stars, planetary spectroscopy, classi- 
fication and line identification in late- 
type stars, moderate- and low-disper- 
sion observations of barred spiral 
galaxies, peculiar galaxies, radiogal- 
axies, and quasi-stellar objects. One 
laboratory application of astrophysi- 
cal interest is a quantitative study of 
high-dispersion emission spectra in a 
luminous shock tube of atoms. 


The performance of an image- 
intensifying system depends on the 
quality of the input image. It has 
proved quite feasible with coude 
spectrographs to intercept the con- 
verging beam from the camera with 
a diagonal mirror and thus to bring a 
high-quality image out to an acces- 
sible location for the image tube 
system. However, because of the rela- 
tively small size (38 mm in diameter) 
of the photocathode in the RCA cas- 
caded tube, the intensifier competes 
unfavorably with photography for 
many problems in which very long 

photographic spectra can be recorded 
at high dispersion. On the other hand, 
because of the extended red sensitiv- 
ity of the S20 photocathode, low-dis- 
persion image tube spectra tend to 
be longer and hence give more infor- 
mation than ordinary photographic 
spectra. However, the best of these 
low-dispersion systems uses fast 
schmidt cameras that are poorly 
suited for image tube work because 
the focus is internal and therefore 

There have been several successful 
approaches to this problem. One of 


these, an //l system involving two a schmidt-type corrector of 100-mm 
concentric reflecting spheres, has aperture ; and a simple plano-concave 
been designed by Bowen for the Palo- field-flattening lens. The fused quartz 
mar cassegrain spectrograph. Suffi- block is 12.5 cm in diameter and 7 cm 
cient working distance is provided at thick. The dispersed light from the 
the focus so that a fairly conventional spectrograph grating passes through 
permanent magnet can be used for the corrector plate and enters the 
focusing the image tube. Another sys- block through a flat, annular surface, 
tern, that of Meinel and Hiltner, is a After traversing the block it strikes a 
folded schmidt camera that, because spherical reflecting surface, is re- 
of the short working distance, re- fleeted to the smaller secondary re- 
quires a solenoid with magnetic field- fleeting surface located in the center 
shaping shims. An //1.5 cassegrain of the front of the block, and is then 
schmidt, designed by Lynds at Kitt reflected out through the center of 
Peak, is sufficiently small in diameter the second surface. Since the two 
(4.2 inches) to fit inside a conven- spherical surfaces are concentric, the 
tional focusing solenoid or permanent system has no off -axis aberrations, 
magnet so that working distance is and the placing of the schmidt cor- 
not a problem. Finally, an //2.25 cas- rector plate at the virtual center of 
segrain schmidt, designed by Bowen, curvature of the spheres serves to 
has a 6-cm working distance, and eliminate spherical aberration. Resid- 
therefore can be operated with the ual chromatism due to refraction at 
simple permanent magnet that is in- the last surface causes the spectrum 
eluded in the standard cascaded tube to be tilted a few degrees with re- 
system. The optical parts of this spect to a plane perpendicular to the 
Bowen camera have been fabricated optical axis. The slight curvature of 
for the Committee by Davidson Op- the focal plane is removed by the thin 
tronics and mounted with an intensi- concave lens mounted just in front of 
fier system on the DTM spectrograph, the photocathode. Well-corrected and 
This cassegrain schmidt consists well-flattened images are obtained 
of three elements : a solid quartz block over a 20-mm diameter. This system 
having an annular flat surface and has tremendously improved the qual- 
two spherical surfaces forming a pri- ity of the spectra obtained with the 
mary and secondary mirror system; DTM spectrograph. 


The DTM image tube spectrograph The plates are measured by Dr. 

has been used by Ford and Rubin to Rubin on a two-coordinate Mann 

obtain plates from which accurate measuring machine; both x and y are 

radial velocities are being deter- recorded for each setting. For each 

mined. With the schmidt camera and line on the plate, settings are made at 

a Bausch and Lomb 300-line/mm intervals of 0.03 mm along the line, 

grating blazed at 7200 A, the disper- Measurements are made with the 

sion is 132 A/mm. The slit width is crosswires set at 45° to the #-direc- 

150 fx, which corresponds at the 84- tion of travel, so that only the portion 

inch telescope to 50 n on the plate or of the spectrum at the point of in- 

a resolution of 20 line pairs/mm. A tersection of the wires is considered 

slit length of 12.5 mm is generally in each measurement. All plates are 

used, corresponding to 160 arc sec on measured twice, with the plate ro- 

the sky. tated 180° for the second measure- 


ment. From the two-dimensional tral range 4500 A to 7000 A. Over 
measure of several dozen plates, it is this longer interval, the distortions 
concluded that accurate radial veloci- are increased correspondingly. How- 
ties can be obtained from image tube ever, they are still sufficiently small 
plates with no greater difficulty than that the reduction procedure can 
that encountered in determining ve- adequately correct for their effects, 
locities from conventional spectro- We show in the frontispiece a 176- 
scopic plates. As an illustration, for minute exposure on the night sky, 
plates centered near Ha, the following covering the spectral range 4500 to 
results are obtained : 7000 A. It can be seen here that the 

1. For 10 neon comparison lines, distortions are still small. 
5881.895 A to 6929.468 A, the for- For each plate, the radial velocity 
ward and reverse measures agree to reduction program written by Dr. 
within 0.1 A (2/*). Rubin for the IBM 1130 computer 

2. A cubic polynomial to relate corrects for the S-distortion and the 
measured distance on the plate to variation in the tilt of the lines from 
wavelength will reproduce each of the the measurements of the comparison 
10 laboratory wavelengths to within lines on that plate. The curvature cor- 
=0.09 A (4 km/sec). rections are predetermined from 

3. The distortions due to the cam- measurements made on test plates 
era-image tube system are a mini- which are taken with the comparison 
mum for spectra centered at the cen- source illuminating the entire slit. 
ter of the image tube, and can be The night sky [OI] 6300-A line is 
corrected satisfactorily in the reduc- measured at intervals along the line 
tion procedure. Three distortions are as a check on the curvature correction 
present: S-distortion, nonparallel and the reduction procedure. This 
lines, and curvature across lines, line is close to Ha 6563 A on the 
These are illustrated schematically plate and in a typical exposure is of 
enlarged in the frontispiece. For the comparable density with Ha. Table 1 
1000- A region discussed above, the gives results from three such meas- 
variation of the y coordinate of the urements of this line. 

lower edges of the spectral lines, Ay, It may be seen that there is no 

is 12 p. The height of the lines in- systematic trend from one end of the 

creases 60 ft for lines 7 mm high. The line to the other. This is important 

variation in the tilt of the lines, Ax, for studies of velocity variations 

amounts to about 25 fx over this along a line. The mean of the abso- 

range. The curvature across each lute values of the residuals is about 

spectral line depends principally on 0.2 A, or less than 2 ^. The mean 

the height of the line on the image wavelength of the [OI] line differs 

tube. With a long slit (12.5 mm) at from the laboratory wavelengths by 

the Perkins 72-inch //18 telescope, about the same amount. This is sig- 

the spectral lines plus comparison are nificantly less than corresponding 

3 mm high on the plate; the curva- values generally found in galactic 

ture of each line is less than 3 \u For studies at lower dispersion. From 

the same slit length at the Kitt Peak these results we conclude that a 

84-inch //7.6 telescope, the length of single measurement of a point on a 

lines on the plate is over 7 mm, and line is accurate to ±2 ^ or ±12 km/ 

the maximum value of the curvature sec at 132 A/mm. 
is 7 /a. A single plate covers the spec- 
















-« £ 
















I* 2 

["< o 

■5 2 

U ts 

^ c 










r— ■ 
















6 ^ r-' 6 d d 6 6 6 cn co d ^ -' 6 d -V -' r-' <oV 

+I++++ +I+++II++I+++II 

r-"- 1 tC0 , 0i0O"0OK l 0'- | - , 0'O00C>0000TfioCN 


+ I ++++ 

+ I + + + I I 

o o o o o o 

+ + I + + + I I 

C0'OCNOC0CNNCN>0^J-C000 l 0OC0iO00i0«OO^^ 



'O>0 < , 0^'0 , 0^ , , > , > , > , , , > , > > 




n ci cn "6 6 <- 6 cn 6 f- r-' r-' 6 co 

I I ++ I + I + I + I I I + 


^ rr co q i- cn - co o >- f- r- q ^ 

I I ++ I + I + I + I 11 + 



>0 > , , 00'0 , 0>0 V , 000 > 0"0 


r-' >- O CO r^ r-' r-' r-' r-' rJ CN O "O 

I I + + + + Mill! + + 

,— •— CNOKK l 0^ l O < — CNCOCO 


I I + + + + I I I I I I + + 


O 0> ; <— ; >0 CO CO «> ; O O 0> CO r— r- CO 


< 3.-* 


CN K r— 

6 ^ ^ 







< 3.J 



<6 ^ d> 







The Committee wishes to express Recently Dr. Ford has been ably as- 
its thanks to the National Science sisted by Dr. Vera Rubin. The coop- 
Foundation for the support which has eration and interest of many astrono- 
been so vital to the success of this mers have resulted in many helpful 
program. suggestions and technical improve- 

It also wishes to acknowledge the ments in this extensive program. Of 

outstanding technical skill, tact, and particular significance is the design 

good judgment of Dr. \V. Kent Ford, of the schmidt camera system by Dr. 

Jr., in carrying the main burden of I. S. Bowen. 
the entire project for so many years. 

Department of Plant Biology 

Stanford, California 

C. Stacy French 


Introduction 143 

Biochemical Investigations 149 

Role of P700 and cytochrome / in the reaction center of photosystem 1 . . 149 
Transfer of energy between reaction centers of photosystem 1 in algae . . 155 
Light-induced shifts in the absorption spectrum of carotenoids in red, 

brown, and yellow-green algae and in a barley mutant 160 

Quenching by quinones of chlorophyll fluorescence in vivo 165 

Some essential considerations in the measurement and interpretation of 

absorption spectra of heterogeneous samples 171 

A spectrophotometer primarily for light-scattering samples at low tem- 
perature 175 

Absorption spectra of chlorophyll a in algae 177 

Elect rophoretic study of the chlorophyll-lipoprotein complexes of Euglena . 186 

The chlorophylls extracted from plants by organic solvents 189 

Chlorophyll fluorescence in algae and chloroplasts 192 

Absorption and fluorescence spectra of Ochromonas danica 196 

The action spectrum for blue-light-stimulated oxygen uptake in Chlorella . 197 
Intermediates in phytochrome transformation in vivo and in vitro . . . 203 

Experimental Taxonomy Investigations 208 

Growth responses of Mimulus races and F x hybrids at the Stanford, Mather, 

and Timberline transplant stations 208 

Photosynthetic rates of Mimulus races and hybrid derivatives .... 214 
Carboxydismutase activity in sun and shade ecotypes of Solidago . . . 216 
Further studies on the effect of oxygen concentration on photosynthetic 

C0 2 uptake in higher plants 220 

Effect of oxygen concentration on dry matter production in higher plants . 228 

Hybridizations in Solidago 233 

Comparative physiological studies on Solanum dulcamara 233 

Clusters of tree species on both sides of the Pacific 234 

Staff Activities 243 

Bibliography 243 

Speeches 244 

Personnel 245 

Carnegie Institute Year Book 66, 1966-1967 


As an introduction to reports of rate photochemical steps rather than 
the photosynthesis group during the a single light-driven reaction are now 
past year it may be useful to mention recognized. This discovery of Emer- 
first a few of the critical discoveries son and of Blinks provided an en- 
that have had a particularly strong tirely new concept about the mechan- 
influence on the Department's line of ism of the process. Furthermore, this 
interest. During the past twenty-five led to thinking about the efficiency of 
years many people from very differ- the overall process as a resultant of 
ent fields of science have discovered the quantum yields of two separate 
the fascination of studying the pho- photochemical reactions and of the 
tosynthetic process, and their efforts losses in the electron transport sys- 
have greatly changed the basic con- tern through which the light-activated 
cepts of the subject. Warburg's intro- steps are correlated, 
duction of Chlorella as experimental The discovery by Duysens — that it 
material for quantitative investiga- was possible to measure changes in- 
tions had been widely taken up before duced by light in the absorption spec- 
that time. At the beginning of this tra of components of the photosyn- 
period the evolution of oxygen by thetic mechanism — initiated many 
isolated chloroplasts had recently detailed investigations of the nature 
been discovered by Hill; the long and sequence of individual steps in 
series of investigations with isolated the electron transport chain. The 
chloroplasts had begun, and today steps of this chain are now being ar- 
continues with increasing understand- ranged into a sequence that is logi- 
ing and complexity. Some of the other cal both thermodynamically and ki- 
subjects of concern at the start of netically, an enterprise to which Dr. 
this period were the chemistry of the Fork and his collaborators at the De- 
intermediates in the path of carbon, partment are making numerous basic 
the maximum efficiency of light con- contributions. Absorption change 
version by photosynthesis, and the measurements have not only been 
possibility of deriving from rate used for following various steps in 
measurements a comparatively simple photosynthesis but, in the hands cf 
physicochemical mechanism for the Kok, have also led to the discovery 
process. Another problem of that of a new form of chlorophyll into 
time, which is still with us, is the which the energy absorbed by one of 
clarification of the chemical nature the photochemical systems is concen- 
and mode of action of the pigment trated. This pigment called "P700" 
complexes that capture the energy of provides a reasonable interpretation 
sunlight for driving photosynthesis in of the so-called photosynthetic unit, 
plants. Some of the forms in which the existence of which was deduced 
chlorophyll a occurs have been identi- long ago by Gaffron and by Emerson 
fied and their complementary func- and Arnold from studies of photo- 
tions in photosynthesis have become synthesis rates in flashing light. 
evident. This problem remains of Another of the striking discoveries 
major concern to the Department. of this period was the finding of 
Studies on the participation of var- Krasnovsky that, under appropriate 
ious pigments in photosynthesis have conditions, chlorophyll can be reduced 
changed radically because two sepa- by light to a pink substance. Just 




what the functional significance of 
this pink chlorophyll may be in photo- 
synthesis still remains to be clarified. 
Why this simple reaction of a sub- 
stance, so widely and thoroughly 
studied for a hundred years by in- 
numerable chemists, was not found 
long ago is astonishing. From this ex- 
ample we must suspect that there 
may well be equally simple, basic 
phenomena that have not yet been 
stumbled upon. 

Another highlight in the under- 
standing of chlorophyll chemistry 
was Dr. Smith's elucidation of the 
nature of the precursor of chloro- 
phyll, protochlorophyll, and its trans- 
formation into chlorophyll by light, a 
reaction which takes place only when 
the protochlorophyll is attached to 
its particular protein. 

During this twenty-five-year period 
the path of carbon was described in 
considerable detail by Calvin and his 
collaborators. Another significant dis- 
covery was made by Frenkel with 
purple bacteria and by Arnon with 
green plants: the formation of adeno- 
sine triphosphate through photo- 
chemical action. Just how this pro- 
duction of a high-energy storage 
compound is linked with the electron 
transport chain is now one of the 
more vigorously pursued aspects of 
our subject. These and other impor- 
tant findings have built a framework 
within which the new discoveries 
may, at least temporarily, be placed. 

When plants are exposed to light 
a dynamic equilibrium is eventually 
reached between the reduced and the 
oxidized forms of each electron-trans- 
ferring component. For any one com- 
ponent the ratio of the oxidized to the 
reduced form at equilibrium may 
vaiy with the light intensity. This 
variation with the intensity of the 
proportion of the oxidized forms of 
cytochrome / and of the pigment 
P700 has been used this year by Dr. 
Fork and Dr. Amesz to study the 

energy-transfer system within and 
between the submicroscopic func- 
tional packages of pigments and en- 
zymes that constitute the photo- 
synthetic units. 

Within a single unit the light en- 
ergy absorbed by any of the hundreds 
of chlorophyll molecules migrates 
around among them until it is finally 
caught by the reduced form of a par- 
ticular trapping molecule. There has 
been some discussion as to whether 
cytochrome / or P700 acts as this en- 
ergy trap for system 1. These inves- 
tigators found that photosynthesis can 
proceed at its maximum rate when a 
part of the cytochrome is oxidized but 
not when any appreciable fraction of 
the P700 is oxidized. This means 
that P700 but not cytochrome / is the 
active trapping center of system 1. 
They compared the rates of photo- 
synthesis with the equilibrium ratios 
of the oxidized and reduced forms of 
both cytochrome / and of P700. From 
the relation between the oxidation 
ratio of P700 and the rate of photo- 
synthesis they concluded furthermore 
that in certain red algae there was 
transfer of migrating exciton energy 
from one photosynthetic unit to an- 
other. The probability of such inter- 
unit transfer from a unit where P700 
is nonoperational to a unit containing 
an active reduced P700 molecule was 
about 0.5. 

Spectral absorption changes caused 
by light can be used to follow the par- 
ticipation of certain substances in 
photosynthesis only when the sub- 
stances responsible for the changes 
at particular wavelengths are known. 
Not all of these measurable changes 
can be definitely attributed to known 
substances. This year Dr. Fork and 
Dr. Amesz found that increased ab- 
sorption at certain wavelengths and 
corresponding decreases at other 
wavelengths, in a variety of algae, 
and in a leaf lacking chlorophyll b, 
were due to carotenoids. The changes 


show that the spectra of some carote- usable at liquid-nitrogen tempera- 
noids shift to longer wavelengths ture. At this low temperature the ab- 
when the live plants are given light sorption bands of the forms of chloro- 
that is absorbed by chlorophylls. phyll are appreciably more distinct 

Some quinones found in plants are than at room temperature, 
believed to be an integral part of the With the new spectrophotometer 
photosynthetic system. Furthermore, we have investigated the absorption 
"unnatural" quinones added to chloro- spectra of particular mutants of 
plast suspensions can be reduced by Chlorella that, when grown in the 
light acting on the chloroplasts. Var- dark, contain very small amounts of 
ious quinones quench the fluorescence chlorophyll a and are completely free 
of chlorophyll in algae or chloroplasts of chlorophyll b. One of these mutants 
and also in pure solutions. The also lacks carotenoid pigments and 
quenching effect of added quinones, has a preponderance of one form of 
in algae or in chloroplasts, has been chlorophyll a. With this alga we have 
attributed to the resulting oxidation come closer than ever before to de- 
of natural quenchers that are in close termining the absorption spectrum of 
contact with the chlorophyll. Further- a single in vivo form of chlorophyll 
more, quenching of fluorescence has a over the entire visible spectrum, 
been taken as an indication of the in- By curve analysis of spectra for 
fluence of quinones on the photosyn- these mutants, and for a variety of 
thetic electron-transport system. Re- different algae grown under appro- 
cently, however, Dr. Amesz and Dr. priate conditions, we hope to be able 
Fork have found that many quinones to establish the shapes of the absorp- 
added to live algae act directly tion spectra of the natural chlorophyll 
on the chlorophyll rather than a complexes. Such information is par- 
quench the fluorescence indirectly ticularly important because the dif- 
through oxidation of components in ferent forms of chlorophyll a are 
the electron-transport chain. The known to be associated with two dif- 
effect of a given quinone concentra- ferent photochemical reactions in 
tion on chlorophyll fluorescence is photosynthesis. At present we can 
greater in live algae than it would be give only a very rough estimate of 
in solution. This result is attributed the proportions of light of a particu- 
to the accumulation of the quinone lar wavelength that are absorbed by 
in fats surrounding the chlorophyll, the different forms of chlorophyll. 
For detecting the presence of the However, by knowing the shapes of 
different forms of chlorophyll, we the spectra of the different forms, it 
have been using for many years a should be possible to calculate the dis- 
specialized kind of spectrophotometer, tribution between the different chlo- 
which gives the first derivative of rophyll forms of absorbed light at 
absorbance with respect to wave- any wavelength, 
length. The curves it produces are A basic unanswered question is 
excellent for the detection of chloro- whether there are only a few specific 
phyll components but are not easy to forms of chlorophyll a, identical in 
analyze in quantitative terms. We their spectra, occurring in various 
therefore modified the instrument to proportions in different plants, or 
plot also conventional absorbance each of the recognized "forms" is 
spectra. We have modified the sample actually a class of pigments within 
holder, which is now particularly which the individual components may 
suitable for studying highly scatter- have appreciable variation of wave- 
ing material such as intact algae, and length peak position. 



This question is being investigated 
by comparison of absorption curves 
of intact algae and by attempts to 
isolate the different forms of chloro- 
phyll from each other by using dis- 
integrated chloroplast material from 
appropriate algae. In this isolation 
work the experimental difficulty is to 
find effective methods for disintegra- 
tion and separation that do not them- 
selves modify the spectral character- 
istics of the chlorophyll complexes. 

Most of the past attempts to sep- 
arate the forms of chlorophyll from 
each other have depended mainly on 
differential centrifugation. Mr. Jean- 
Marie Michel has been using electric 
fields in polyacrylamide gels to sep- 
arate particles of disintegrated chlo- 
roplasts that differ in their electric 
charges. The particles used for these 
electrophoretic separations, already 
selected by differential centrifuga- 
tion, gave preparations differing in 
their absorption spectra only after 
treatment with a detergent. A par- 
ticularly significant preparation from 
Euglena was free from carotenoids. 
Various methods of further chloro- 
plast disintegration are being tried to 
avoid the need for detergents while 
other separation procedures are being 
studied by Mr. Michel and Dr. Michel- 

Plants grown in the dark, then ex- 
posed to light, form chlorophyll a rap- 
idly. The newly formed chlorophyll 
changes its absorption peak from 684 
m/x to 672 mix in a short time. During 
this transition period chlorophyll ac- 
quires its phytol tail. It was therefore 
thought that the difference between 
the forms of chlorophyll in fully 
greened plants might correspond to 
differences in the chlorophyll extrac- 
table by organic solvents. Such ex- 
tracts had been found by Dr. Sironval 
and Dr. Michel-Wolwertz in Belgium 
to give several modified chlorophylls 
separable by paper chromatography, 
in addition to the common chloro- 

phylls a and b. It was thought that 
these "satellite chlorophylls'" might 
be related to the forms of chlorophyll 
in live plants. This year, however, Dr. 
Michel-Wolwertz found that they 
were formed by oxidation during the 
extraction and chromatographic sepa- 

The detection of certain pigments 
in the presence of others can often be 
done by fluorescence spectroscopy, 
since the wavelength distribution of 
light from a fluorescing pigment is 
as characteristic a property as is its 
absorption spectrum. Dr. Brown has 
investigated the fluorescence spectra 
of various algae to correlate, insofar 
as possible, the absorption and fluo- 
rescence spectra of the forms of chlo- 
rophyll in living algae. 

The relative contribution of each 
pigment to the observed total may be 
very different in absorption and in 
fluorescence spectra. This is because 
the fluorescent efficiency of the indi- 
vidual pigments can be very different 
and, furthermore, the efficiency of 
energy transfer from one pigment to 
another may be greatly influenced by 
the spatial relations between them. 

Many investigations have been 
made of the variations in rate of 
photosynthesis with time when plants 
are illuminated after a dark period. 
The time-course curves from such ex- 
periments show all sorts of complex 
induction and outburst effects that 
have been widely used to deduce the 
nature of various steps in the mech- 
anism of photosynthesis. Not only the 
rates of oxygen and of carbon dioxide 
exchange resulting from illumination, 
but also the converse effect, the re- 
adjustment of the rate of gas ex- 
change in the beginning of a dark 
period following a light exposure, 
have been the subject of many 

Such research has usually shown 
big differences in the results when 
blue light is compared with other 



parts of the spectrum. Blue light also 
has other specific effects, such as con- 
trolling the shape of higher plants, 
and influencing the types of chemical 
compounds formed by photosynthesis. 
For blue light to do these things there 
must be a particular pigment, or sev- 
eral of them, that absorbs blue light 
preferentially. This year Dr. James 
M. Pickett measured the influence of 
short exposures of blue light on the 
oxygen uptake of Chlorella in the suc- 
ceeding dark period. The results show 
that some sort of a flavin type of com- 
pound is the pigment responsible for 
this blue light effect. Simultaneous 
work by Dr. Kowallik in Professor 
Gaffron's laboratory in Florida and 
by Professor Ried of Frankfurt, a 
former Visiting Investigator at the 
Department, gave identical results. 
Because of the different plants and 
procedures used in the three labora- 
tories the agreement greatly strength- 
ens the theory that this pigment may 
be of basic significance and may be 
widely distributed in various plants. 
Phytochrome is another pigment 
that is probably not directly a part 
of the mechanism of photosynthesis, 
but is very important in controlling 
the pathways of biochemical metabo- 
lism. By contrast with the flavin re- 
sponsible for blue light effects, phyto- 
chrome has been isolated in pure form 
and is far better known as a chemical 
entity. When phytochrome is illumi- 
nated it changes from one form to 
another. The direction of the change 
depends on the color of the light to 
which it is exposed. This year Pro- 
fessor Briggs of Stanford and Dr. 
Fork continued the collaborative 
work on phytochrome, discussed in 
last year's report, in which some in- 
termediate substances in the trans- 
formation reaction of the purified 
pigment were described. The recent 
results show that similar intermedi- 
ate products are formed when phyto- 
chrome is transformed by light in 

live plants, as well as when it is in 
pure form. The rate constants for the 
opposing reactions, whose balance 
determines the relative concentration 
of the two forms for an exposure to 
a particular wavelength, were, how- 
ever, found to differ in the plant and 
in solution. They also determined the 
absorption spectra for some of the 
intermediate forms in the photo- 
chemical transformation of phyto- 

In recent years many of the investi- 
gators of photosynthesis have come 
together annually or more often from 
all over the world. Furthermore, sev- 
eral comprehensive books on the sub- 
ject appear each year and a new 
international journal, Photosynthet- 
ica, has been started. Such close com- 
munication between different workers 
is essential in planning experiments 
and in modifying theories so that the 
current work is relevant to present 
thought on the subject. There is, 
however, a danger inherent in too 
much unity of thought in science, as 
well as in politics and religion. When 
most of the people in any field think 
along the same lines, there may be 
little likelihood that they will accept 
new concepts. No doubt there have 
been occasions when the investigators 
of photosynthesis have all been simul- 
taneously following major misconcep- 
tions. However, the volume of special- 
ized journal articles has become so 
great that no one person can keep in 
mind the detailed findings of various 
laboratories even within a narrow 
part of the field of photosynthesis. 
This fact itself may help to diversify 
thinking on the subject, thus increas- 
ing the probability of hitting upon 
important new interpretations. 

In spite of all the intense research 
efforts on different aspects of the 
mechanism of photosynthesis, many 
basic questions are still far from 
clarification and some significant 
areas are practically neglected in the 



rush to make new contributions to 
the detailed understanding of some 
of the more popular aspects of the 

Experimental taxonomy. Much of 
the current year's effort of the Ex- 
perimental Taxonomy group has 
been directed toward continued re- 
searches on the comparative physiol- 
ogy and biochemistry of plants 
originating from contrasting environ- 
ments. These studies are aimed at 
discovering basic physiological mech- 
anisms that operate in natural selec- 
tion and evolution of higher plants. 

The discovery last year that the 
amount of oxygen present in normal 
air depresses the photosynthetic rate 
in most higher plants to about 30% 
below that in oxygen-free air has led 
to experiments to determine whether 
or not growth of such plants is also 
inhibited in normal air. Initial results 
from two unrelated plants, monkey 
flower and beans (Mimulus cardinalis 
and Phaseolus vulgaris), do show that 
dry weight yield may be twice as 
great during a less than 2-week pe- 
riod in an atmosphere containing only 
2.5% or 5% 2 as compared with 
21%. In contrast, a third plant, corn 
(Zea, mays) , whose photosynthetic 
C0 2 uptake is unaffected by 2 con- 
centration in the range 0%— 21%, 
failed to show a significant weight 
increase under the same conditions. 
These results open a new field of in- 
quiry into basic differences that have 
evolved in the photosynthetic mech- 
anism of higher plants. 

Using improved quantitative tech- 
niques, Dr. and Mrs. Bjorkman have 
demonstrated that the light-saturated 
photosynthetic rate ingoldenrod (Sol- 
idago virgaurea) is closely correlated 
with the activity of the enzyme 
carboxydismutase in the same leaves 
of these plants. Furthermore, genet- 
ically determined differences in the 
light-saturated photosynthetic rates 
in sun and shade races of Solidago are 

closely linked with their capacity to 
produce this enzyme. 

A race of Solidago originally from 
a sunny habitat in northern Norway 
attains a high photosynthetic rate 
under saturating light intensities 
when the plants are previously grown 
at high light intensities. The amount 
of the enzyme carboxydismutase of 
the leaves likewise is high. Another 
race of the same species but from a 
shaded habitat in southern Sweden 
is capable of producing only about 
half as much of the enzyme. It also 
lacks the ability to attain a high light- 
saturated photosynthetic rate. 

The marked differences in the ca- 
pacity of lowland and alpine races of 
the Mimulus cardinalis-M. lewisii 
complex to fix C0 2 under light-satu- 
rating conditions reported last year 
by Dr. Hiesey, Dr. Bjorkman, and Dr. 
Nobs have been followed in first- and 
second-generation progeny of crosses 
between alpine and lowland forms. 
The differences in photosynthetic ca- 
pacity under light-saturating condi- 
tions are inherited in much the same 
way as morphological differences such 
as flower color and leaf structure. The 
differences in photosynthetic capacity 
are partially linked through genetic 
coherence with the morphological 
characters. The chlorophyll content 
of leaves also differs markedly in con- 
trasting races, but is inherited inde- 
pendently of photosynthetic capacity, 
and bears no relation to light-satu- 
rated photosynthetic rates. 

This year Dr. Nobs and Dr. Hiesey 
completed an extensive five-year 
study of the performance of first-gen- 
eration hybrids between ecological 
races of Mimulus of various degrees 
of relationship. The study was carried 
out at the Stanford, Mather, and Tim- 
berline transplant stations. There are 
marked differences in the degree to 
which hybrid vigor is expressed in 
these three contrasting environments. 
The expression of these differences is 


highly dependent both upon the ge- tudinal vegetational belts in Japan in 

netic composition of the parental comparison with comparable altitu- 

races used in the crosses and upon dinal transects in western North 

the environment in which they are America, with particular emphasis on 

being observed. An important finding tree species. The belts of tree vegeta- 

is that the survival capacity of first- tion are closely related on both sides 

generation hybrids between distinct of the Pacific. This^ concept is 

ecological races within this species- achieved when one studies clusters of 

complex in widely different climates morphologically closely related spe- 

is inherited in an intricate and largely c ies that circle the earth in various 

unpredictable way. The results are latitudinal belts, 

being incorporated in a monographic Modern tendencies to split species 

study of the Erythranthe section of and a into smaller and smaller 

Mimulus that includes the results of ^ them ag . have 

biosystematic, transplant, and physio- , \ & , . , ,, , 

logical investigations. obscured our perception of the rela- 

Dr. Jens Clausen, on field excur- tionships that exist between gross 

sions in conjunction with attendance morphological characters and broad 

at the Eleventh Pacific Science Con- evolutionary clusters of species occu- 

gress at Tokyo, made a study of alti- pying different continents. 


Role of P700 and Cytochrome / sorption. One light-induced signal 

in the Reaction Center showed a delay upon onset of illumi- 

OF Photosystem 1 nation and a relatively fast (^0.5 

sec) decay upon darkening. The nega- 

Jan Amesz and David C. Fork tive maxima in the difference spec- 
trum of this signal were at 435 and 

In spite of extensive studies rela- 705 m/x, indicating an oxidation of 

tively little is known of the nature of P700 (Kok and Hoch, 1961) in the 

the primary photochemical reactions light. The second signal showed an 

that bring about photosynthesis. Al- immediate response to illumination 

though it may be stated that more is and a slow (^6-10 sec) decay in 

known about photosystem 1 than the dark. The difference spectrum 

about photosystem 2, our knowledge with minima at 553 and 420 m/x and 

even of the photochemistry of system a maximum at about 402 m/x indicated 

1 is limited. For this reason we stud- the oxidation of an /-type cytochrome, 

ied light-induced reactions of P700, probably similar to that isolated by 

the presumed primary reactions of Katoh (1960) from Porphyra tenera. 

photosystem 1 and cytochrome / in the Figure 1 shows some typical kinetics 

red algae Iridaea splendens, Schizy- of light-induced absorbancy changes. 

menla pacifica, and Porphyra per- Except for a lowering of the rate of 

forata. dark decay upon lowering the tem- 

Illumination of Iridaea, which had perature, the absorbancy changes in 

been in the dark for a few seconds, response to light were essentially the 

with red or far-red (>680 m/x) or same at room temperature and 1°C, 

blue light mainly absorbed by system as well as in the presence of DCMU 

1, gave absorbancy changes in the [3- (3, 4-dichlorophenyl)-l, 1-dimeth- 

regions of P700 and cytochrome ab- ylurea]. No evidence was found for 










435rn/j ' 


^-y;*^w;/.*W^V^V%«,*,vr'4*' v *' 


705m/ii \ 

b T 

AA=2xlO" 3 


- Fluorescence 

AA = 5xlCT 3 



Fig. 1 . Kinetics of light-induced absorbance changes AA in Iridaea splendens at 420, 435, and 
705 ma due to oxidation of the f-type cytochrome and P700. For traces a and fo the wavelength 
of the actinic light was 708 m/x (6.1 nanoeinstein cm -2 sec -1 ). For traces c and d the actinic light 
was a band around 414 m/x, 0.7 nanoeinstein cm -2 sec -1 . Trace d shows the absorbance change, 
superimposed on a fluorescence signal, which caused the rapid deflections upon illumination and 
darkening; trace c shows the fluorescence signal alone. The preceding dark time was 6 seconds 
for all traces and the temperature was 20°C. 

reactions of &-type cytochromes under 
the conditions applied. 

The amounts of oxidizable P700 
and cytochrome / relative to chloro- 
phyll were calculated from the max- 
imum light- induced absorbancy 
changes to be 1/4.1/310 in Iridaea. 
Calculations were based on estimated 
specific extinction coefficients of 70 
m^I' 1 cm -1 at 420 m^ for the oxidation 
of cytochrome and of 73.3 mM -1 cm -1 
for bleaching of P700 at 705 m**, 
which is the same as that of chloro- 
phyll a in 80% acetone at the red 

maximum. Comparison of the size of 
the absorbancy changes at 435 and 
705 m/x indicated a specific extinction 
coefficient of 25.5 mMr 1 cm -1 at 435 hi/a 
for P700. The amount of cytochrome 
is about the same as found by Nishi- 
mura (1967) in Porphyra, but high 
compared to that present in blue- 
green algae, where cytochrome and 
P700 appear to be present in about 
equal amounts of one per several hun- 
dred chlorophyll molecules. 

Quantum efficiency for cytochrome 
oxidation. As illustrated in Table 1, 



TABLE 1. Quantum Yields for Cytochrome Oxidation 












(nE cm _2 sec _1 ) 



































































| 1 












Quantum yields were measured from the initial decrease in absorbance at 420 m/x upon illumi- 
nation. Preceding dark time was 6 seconds for Iridaea, 5 seconds for Schizymenia and 3 sec- 
onds for Porphyra. Thallus absorption was about 60%-70% at 692 m/x and 20%-30% at 
708 m/x. 

quantum efficiencies for the oxidation 
of the /-type cytochrome in Iridaea 
and Schizymenia were about 0.5-0.65 
molecules of cytochrome oxidized per 
quantum of red or far-red light ab- 
sorbed. These quantum efficiencies 
were calculated from the initial rate 
of the absorbancy decrease at 420 m^ 
after a preceding dark time of a few 
seconds. Calculation of the efficiency 
for oxidation of P700 from the steep- 
est part of the absorbancy curves 
at 435 m^ gave considerably lower 
yields of about 0.2. In Porphyra per- 
forata the highest yield for cyto- 
chrome oxidation observed was about 
0.15. Similarly low or lower values 
have been reported by Nishimura 
(1967) for other Porphyra species. 

Relation between the rates and oxi- 
dation levels of P700 and cytochrome 
f. An explanation for the delay in 
P700 oxidation upon illumination 
could be that photooxidized P700 
reacts rapidly with reduced cyto- 
chrome / and that accumulation of 
oxidized P700 can occur only when all 
cytochromes in the same reaction 
center have been oxidized. It has been 

shown (Beugeling and Duysens, 1966 ; 
Duysens 1966) that this hypothesis 
can explain satisfactorily the kinetics 
of P890 (the presumed analogue of 
P700) and of a cytochrome in the 
purple bacterium Chromatium. 

Figure 2 illustrates an experiment 
with Iridaea, in which the alga was 
illuminated for 3 seconds with an in- 
tensity of light sufficient to oxidize 
P700 and cytochrome nearly com- 
pletely. After various dark periods a 
second illumination of short duration 
was given and the oxidation of these 
compounds caused by the second illu- 
mination was recorded. At the begin- 
ning of the second illumination P700 
was almost completely reduced in all 
experiments, but the level of oxida- 
tion of cytochrome varied widely for 
different dark times applied. If it is 
assumed that the oxidation of cyto- 
chrome by oxidized P700 proceeds 
very rapidly, then the relative initial 
rate of P700 oxidation in the light 
will be proportional to the fraction of 
reaction centers that contain only 
oxidized cytochrome molecules. When 
it is further assumed that the chance 





On On P700(434 mj j) 

0.5 sec dark -*/ / ~ c r\ 

\ ' \ a ~J O.osec Un 

""l * AA = I0' 3 dark 



1 ♦. 




AA=5xl0" 3 





3 sec dark \J* 

b Off 



,•"" 6 sec dark f 

-fW c 


Time, sec 

Fig. 2. Absorbance changes AA in Iridaea at4°C at 434 m/x (P700) and 420 m//, (cytochrome) 
induced by a 1 75-msec exposure to 708 m//, (4.2 nanoeinstein cm -2 sec~ T ) after preceding illu- 
mination with 707 m/x. Dark intervals of various lengths were applied as indicated. 

of being reduced in a given time in 
the dark is the same for every cyto- 
chrome molecule, then 

X = x n (1) 

where X is the fraction of reaction 
centers in which all cytochromes are 
oxidized, x is the fraction of cyto- 
chromes that is oxidized, and n is the 
number of cytochromes per reaction 
center. Thus 

<fi x — fCX n 


where fa is the quantum efficiency for 
P700 oxidation; k is a constant, equal 
to fa =1 . As Fig. 3 shows, the experi- 
mental values for the initial rate of 
P700 oxidation fit the calculated rela- 
tion for n = 4 rather well, in good 
agreement with the earlier mentioned 
finding that the amount of cytochrome 
in Iridaea is 4.1 times larger than 
that of P700. 

As Fig. 3 also shows, the sum of the 
rates of P700 and cytochrome oxida- 

tion, which might be taken as the rate 
of the primary photochemical reac- 
tion, is independent of the oxidation 
level of cytochrome. Experiments in 
which we measured the sum of these 
rates as a function of the oxidation- 
reduction level of P700 gave the re- 
sults shown in Fig. 4. 

As discussed in this year's report 
(pp. 155-160), if P700 is a primary 
reactant, photochemistry should be- 
come less efficient when part of P700 
is in the oxidized state, because 
part of the reaction centers then are 
not operative and light quanta will 
be wasted, unless they are trans- 
ferred to another reaction center. 
The results of Fig. 4, in agree- 
ment with those obtained on photo- 
synthesis of Cryptopleura violacea 
(pp. 155-160, this volume), suggest 
that P700 is a primary reactant. How- 
ever, the decline in photosynthetic 
efficiency appears to be less than pro- 
portional to the fraction of P700 that 





.2 0.4 






0.2 h 

1 1 1 1 1 I 1 1 1 


n O 

d P700 + Cyt oxid / / 

o P700 oxid / //// 


a°X / /// 

$y / /// 

.ov o/ / / / 
^/ A oy / / / 

/ \<?7 a/ // 

/ y ( yWM 
/ yyyyyy 

/ //yy* 

-^ir^^^^x^-^r^yC-- ^yy^\ 1 1 1 1 

C y^0x/Cyt Total 


Fig. 3. Quanfum yield for the initial rate of P700 oxidation in Iridaea (circles) and for the 
sum of the initial rates of P700 and cytochrome oxidation (squares) as a function of the level of 
oxidation of cytochromes. The oxidation rates were measured from the rate of absorbance 
change at 435 and 420 m/x, in experiments as shown in Fig. 2. The solid lines give calculated 
curves for P700 oxidation for reaction centers containing one P700 molecule and up to 5 cyto- 
chrome molecules (as described in the text). 

is oxidized. This suggests that energy 
transfer between the reaction centers 
occurs. A calculated curve (A. and P. 
Joliot, 1964) for 50% probability of 
energy transfer to a second reaction 
center when the first one is inopera- 
tive seems to give the best fit for the 
experimental data. 

Conclusion. The results are in quan- 
titative agreement with the hypothe- 
sis that a reaction center in Iridaea 
contains one P700 molecule and four 
cytochrome / molecules. The observa- 
tion that the efficiency of the photo- 
chemical reaction is dependent on the 
oxidation level of P700 but not on 

that of cytochrome indicates that 
P700 but not cytochrome / is a pri- 
mary reactant. The kinetics and effi- 
ciencies of cytochrome and P700 oxi- 
dation in the light can be explained 
by the hypothesis that both P700 and 
cytochrome are in the main path of 
photosynthetic electron transport and 
that P700, when oxidized photochem- 
ically, oxidizes cytochrome / in a rapid 
reaction. Accumulation of oxidized 
P700 occurred only when all cyto- 
chromes in a reaction center were 

The results do not prove that cyto- 
chrome / is oxidized via P700, 






E 0.5 




1 1 1 f 

P700 red /P700 TofQ , 


Fig. 4. Sum of the rates of P700 and cytochrome oxidation in Iridaea as a function of the 
level of reduction of P700. The dashed lines give the expected relations for different proba- 
bilities (p) of transfer of excitation energy between reaction centers (see text). Open and solid 
circles represent two different sets of measurements, one done at 0°C and the other at 4°C, with 
different samples and actinic light of 708 mfi. 

although this is the simplest hypoth- 
esis that explains the results quan- 
titatively. Direct information about 
this could possibly be obtained by the 
use of intense flashes of short dura- 
tion to oxidize P700. In this respect 
it is interesting" to note a recent ex- 
periment of Nishimura (1967) who 
reported the oxidation of a relatively 
large amount of cytochrome / in 
Porphyridium cruentum by a single 
laser flash. This could argue against 
the hypothesis that the cytochrome is 
oxidized via P700, which is probably 
present in smaller amounts than the 
amount of cytochrome that was oxi- 


Beugeling, T., and L. N. M. Duysens, in 
Currents in Photosynthesis, edited by 
J. B. Thomas and J. C. Goedheer, Ad. 
Donker, Rotterdam, p. 49, 1966. 

Duysens, L. N. M., in Currents in Photo- 
synthesis, edited by J. B. Thomas and 
J. C. Goedheer, Ad. Donker, Rotter- 
dam, p. 263, 1966. 

Joliot, A., and P. Joliot, Compt. Rend., 
258, 4622, 1964. 

Katoh, S., Plant and Cell Physiol., 1, 91, 

Kok, B., and G. Hoch, in Light and Life, 
edited by W. D. McElroy and B. Glass, 
Johns Hopkins Press, Baltimore, p. 
397, 1961. 


Nishimura, M., in Energy Conversion by transfer between system 1 units in 

the Photosynthetic Apparatus, Brook- algae we therefore compared, in a 

haven Symposia in Biology, No. 19, num ber of algal species, the relative 

Brookhaven National Laboratory, efficiency of Q 2 evolution with the 

Upton, N. Y., 1967. oxidation level of P700 and of cyto- 

Transfer of Energy between chrome / as a function of light in- 

Reaction Centers of Photosystem tensity. ^ 

1 in Algae ^ various intensities of light we 

measured the fraction of P700 and 
David C. Fork and Jan Amesz cytochrome / that was in the reduced 
The concept of the photosynthetic state. Under the same conditions the 
unit stemming from the early work relative rate of 2 evolution was 
of Emerson and Arnold (1932, a, b) measured as a function of light in- 
has been substantiated considerably tensity with a Teflon-covered elec- 
during recent years. A photosyn- trode (Year Book 61, p. 343). We 
thetic unit can be defined as a pri- used adjacent parts of the thalli or, 
mary reaction center with associated with Schizothrix, samples of the same 
pigment molecules (chlorophylls, ca- culture. In all experiments discussed 
rotenoids, biliproteins) having a below, the actinic light was of a 
higher probability of transferring wavelength band such that absorption 
their excitation energy to this reac- by system 2 exceeded that by system 
tion center than to another center. 1 (blue light for the green alga Viva 
Experiments with purple bacteria lobata ; green light for the red algae) . 
( Vredenberg and Duysens, 1963 ; and Therefore, we can assume that photo- 
Clayton, 1966) and with the green chemistry in system 1 occurred at 
alga Chlorella pyrenoidosa (A. and maximum efficiency in the linear part 
P. Joliot, 1964) indicate that the pho- of the light curve of photosynthesis, 
tosynthetic units in purple bacteria, Figure 5 shows the rate of 2 evo- 
as well as the photosynthetic units of lution and the absorption changes 
system 2 (the 2 evolving system) in corresponding to the oxidation level 
algae are not separated, but that ex- of P700 and of cytochrome as a func- 
citation energy can move more or less tion of light intensity in the red alga 
freely from one unit to another. When Cryptopleura violacea. The rate of 2 
one reaction center is inoperative, by evolution was the steady-state rate 
just having trapped an exciton, a after several minutes of illumination, 
second exciton may be transferred to The oxidation levels of P700 and cyto- 
a different reaction center. chrome were measured with the ap- 
The extent of energy transfer from paratus described earlier (Year Book 
one photosynthetic unit to another is 63, p. 435) by the increase of absorp- 
an important factor in determining tion at 706 for P700 and 420 m^ for 
the amount of light energy that even- cytochrome, upon turning off the ac- 
tually reaches a functional reaction tinic light. The exposure was long 
center when some of the reaction cen- enough to give a steady-state rate of 
ters are inoperative. Therefore, it is 2 evolution. To minimize signals 
possible in principle to obtain inf or- caused by chlorophyll fluorescence the 
mation about this transfer by meas- photomultiplier was placed about 30 
uring the efficiency of conversion of cm from the sample cuvette. The 
light energy into photochemical prod- measurements were corrected for flu- 
ucts as a function of the fraction of orescence by subtraction of the signal 
reaction centers that is operational, obtained under the same conditions 
To obtain information about energy but without the measuring beam. 




b 3 


P O) « 

c c 
o o 










] 1 


1 1 J 1 1 1 1 




^-"~ ~^- t o- 


^°*^ 2 evolution 


/ / 
/ / 

^ 5^ ^706*^700) 



L^^T ^420nvj(Cyt.) 



1 'i 
/ // 

^7 #/ 
/ / / 
/ / / 

-/ // 

T (f / 
1 / J 

I s y 3 

/ / _/ 

1 1 1 1 1 1 1 



Light intensity, ergs cm -2 sec" 1 xlO 


Fig. 5. Steady-state level of oxidation of P700, f-type cytochrome, and steady-state rate of O2 
evolution in Cryptopleura v/o/acea as a function of the intensity of green light (540 m/x max), 40 
nut half bandwidth. 

As Fig. 5 shows, P700 was almost 
completely in the reduced state at 
light intensities where photosynthesis 
proceeded with maximum efficiency; 
only above the point where the in- 
tensity curve for photosynthesis 
started to depart from linearity did 
a sizable proportion of P700 become 
oxidized. By contrast, the cytochrome 
was already partly oxidized at lower 
intensities. These results therefore 
indicate that P700, but not cyto- 
chrome, behaves as a true reaction 
center because a maximum rate of 
photosynthesis can be expected only 
when all reaction centers are opera- 
tional (in the reduced state). 

For quantitative evaluation, the 
data of Fig. 5 were replotted as 
shown in Fig. 6, which gives the rela- 
tion between the relative efficiency of 
2 evolution and the fraction of P700 
and of cytochrome that was in the re- 
duced state. The relative efficiency for 

2 evolution was obtained by dividing 
the relative rate by the light inten- 
sity. The relative efficiency corre- 
sponding to the dashed line in Fig. 5 
was normalized at unity. The fraction 
of P700 and of cytochrome in the re- 
duced state was calculated by sub- 
tracting the light-induced signal from 
the maximum signal obtainable with 
high actinic intensity of red or blue 
light (system 1) and with 5 • 10~ 5 M 
DCMU. The shaded areas in Fig. 6 
indicate the estimated uncertainty. At 
a high level of reduced P700 or cyto- 
chrome the uncertainty was mainly 
determined by the measurement of 
the relative photosynthetic efficiency; 
at a low level of reduction, by the 
measurement of the total amount of 
oxidizable P700 or cytochrome. 

Figure 6 shows that even when a 
significant proportion (about 20%) 
of the cytochrome is oxidized, photo- 
synthesis can proceed at maximum 



0.5 1.0 

P700 red /P700j o t a | or Cyi rec |/Cyif f a | 

Fig. 6. Relative efficiency of O2 evolution as a function of the level of reduction of P700 and 
of the f-type cytochrome in Cryptopleura v/o/acea. 

efficiency. By contrast, the oxidation 
level of P700 appears to be closely 
correlated with photosynthetic ef- 
ficiency; a drop in the level of reduced 
P700 is accompanied by a decline in 
relative efficiency of photosynthesis. 

If no energy transfer between pho- 
tosynthetic units occurred, a relation 
between photosynthetic efficiency and 
level of reduction of P700 should exist 
as represented by the diagonal line in 
Fig. 6. However, the curve for P700 
lies above this line, indicating that 
energy transfer does occur. The 
dashed curves have been computed by 
A. and P. Joliot (1964) from a sim- 
plified model of system 2 for different 
probabilities for energy transfer. The 
experimental data agree rather well 
with a curve corresponding to a prob- 
ability of about 50% to 60% for 

transfer of energy to a second reac- 
tion center when the first one is in- 

Figure 7 shows results obtained 
with the blue-green alga Schizothrix 
calcicola. With this alga the results 
were more complicated than with 
Cryptopleura. Here also the left part 
of the band lies above the diagonal, 
suggesting transfer between reaction 
centers of system 1. However, at the 
right side of the figure (correspond- 
ing to low light intensities) the meas- 
urements drop considerably below 
the diagonal line. This effect may be 
caused by a reaction of oxidized P700 
with reduced products of system 1, 
either directly or via intermediate 
substances such as cytochrome /. 
Such a reaction could keep the level 
of reduced P700 high and increase 








O 0.5 




i 1 


1 « i i i A 


/ k 

/ M 


/ r- 



^0^ £ ' 


\ M 



9 \ 



- / 


/, , 


1 i i i i 


P700 red /P700 fofa | 


Fig. 7. Relative efficiency of O2 evolution as a function of the level of reduction of P700 in 

Schizothrix ca/cico/a and Porphyra perforata. 

the rate of electron transport in sys- 
tem 1 to a level higher than com- 
puted from Oo evolution, as discussed 
below. The same explanation could 
apply to the red alga Porphyra perfo- 
rata, also shown in Fig. 7. Results 
with Ulva lobata were similar to 
those obtained for Porphyra. 

Discussion. The results obtained 
with Cryptopleura support the hy- 
pothesis that P700, but not cyto- 
chrome /, acts as a primary reactant 
of light reaction 1 in photosynthesis. 
Accumulation of oxidized P700 was 
accompanied by a decline in photosyn- 
thetic efficiency, but the cytochrome 
was already partly oxidized at light 
intensities where no such decline was 
observed. The results with the other 
algae also agree with the assumption 
that P700 is a primary reactant. In no 

instance did oxidized P700 accumu- 
late when 2 evolution proceeded at 
optimum efficiency. From these ex- 
periments alone it might be argued 
that the oxidation of P700 is a side 
reaction that only occurs when photo- 
synthesis is saturated or inhibited. 
However, this would be in disagree- 
ment with the observation of Kok et 
al. (1963) that the turnover rate of 
P700 in spinach chloroplasts is 
strongly stimulated by the addition 
of NADP and ferredoxin and is in- 
hibited by DCMU. 

The results with Cryptopleura and 
Schizothrix indicate that energy 
transfer occurs between photosyn- 
thetic units of system 1. In Crypto- 
pleura photosynthesis proceeded with 
50% relative efficiency even though 
P700 was about 70% oxidized. A sim- 


ilar phenomenon was seen in Schizo- between photosynthetic units prior to 
thrix at high light intensities. the first visit. A formally different 
No evidence for energy transfer model, applied to bacteria, by Vreden- 
was obtained by the experiments with berg and Duysens (1963) implies 
Porphyra. However, these experi- that the reaction centers are homo- 
ments could prove the absence of geneously dispersed between the light- 
energy transfer in this alga only if harvesting pigment molecules, and 
two conditions were fulfilled : that the chance of finding a second 

1. The rate-limiting step in photo- reaction center depends only on the 
synthesis is not in the Calvin cycle or number of random walks involved, 
in one of the reactions leading to the This model can probably not be ap- 
reoxidation of the hypothetical pri- plied to algae, for reasons discussed 
mary-reduced product of system 1. below. 

If it were, accumulation of primary In Schizothrix no correlation has 
reductant could limit the efficiency of been found (Vredenberg and Duy- 
light utilization in system 1. It is un- sens, 1965) between fluorescence of 
likely that this effect occurred in our system 1 and the oxidation level of 
experiments, since the concentration P700. For this reason it has been as- 
of C0 2 was not rate limiting; more- sumed (Duysens, 1965) that energy 
over, there is other evidence that the from the bulk chlorophyll is trans- 
rate-limiting step is in the electron- ferred to the reaction center via a 
transport chain between the two sys- special, weakly fluorescent chloro- 
tems. phyll that is present in small amounts 

2. The second condition is that the and located near the reaction center, 
rate of 2 evolution is equal to the Back transfer of energy from this 
rate of electron transfer by system 1. chlorophyll to the bulk chlorophyll 
This is not true when a ' 'cyclic' ' reac- would be negligible, so that bleaching 
tion occurs between the reduced and of P700 would not affect the fluores- 
oxidized products of system 1, be- cence yield of the bulk chlorophyll. A 
cause then the rate of photochemistry consequence of this hypothesis is that 
in system 1 would be higher than esti- energy transfer between reaction 
mated from measurements of 2 evo- centers would occur via the special 
lution. There is evidence that cyclic chlorophyll which should form a more 
reactions occur in intact algae and in or less continuous structure contain- 
isolated chloroplasts, especially in the ing the reaction centers. 

presence of redox catalysts. This 

would explain the results with For- References 

phyra and other divergent results as, 

for example, those with Schizotrix at Clayton, R. K., Photochem. Photobiol, 5, 

low light intensity. 807, 1966. 

It must be noted that the probabil- Duysens, L. N. M., Arch. Biol. (Liege), 

ity of energy transfer p of 50% to 76, 251, 1965. 

60% as derived from the data of Emerson, R., and W. Arnold, /. Gen. 

Fig. 6 probably gives only an approxi- Physiol, 15, 391, 1932a. 

mate measure of the extent of energy Emerson, R., and W. Arnold, /. Gen. 

transfer between the photosynthetic Physiol, 16, 191, 19326. 

units in system 1. The number p j liot, A., and P. Joliot, Compt. Rend., 

gives the probability for transfer to 258, 4622, 1964. 

a second reaction center when the Kok> B#) B . Cooper, and L. Yang, in 

first one visited is inoperative, but Studies on Microalgae and Photosyn- 

gives no information about transfer thetic Bacteria, Japanese Soc. of Plant 



Physiologists, The Univ. of Tokyo 
Press, p. 373, 1963. 

Vredenberg, W. J., and L. N. M. 
Duysens, Nature, 197, 355, 1963. 

Vredenberg, W. J., and L. N. M. Duy- 
sens, Biochcm. Biophys. Acta, 94, 355, 

Light-Induced Shifts in the 

Absorption Spectrum of 

Carotenoids in Red, Brown, and 

Yellow-Green Algae and in a 

Barley Mutant 

David C. Fork and Jan Amesz 

It has been known for a number of 
years that certain photosynthetic bac- 
teria exhibit light-induced changes of 
absorption in the region 450 to 540 
niju, which are apparently produced by 
a shift toward longer wavelengths of 
the absorption of a carotenoid pig- 
ment. Characteristic difference spec- 
tra are observed which show maxima 
and minima separated by about 30 
m^t. There is evidence (Amesz and 
Vredenberg, 1966) that these changes 
are not caused by oxidation-reduction 

reactions of carotenoid pigments but 
rather that they may be produced by 
a change in the environment of the 
pigment molecules resulting from 
electron transport. 

We have found in a number of 
variously colored algae and a barley 
mutant light-induced absorbance 
changes which appear to be caused 
by a shift toward longer wavelengths 
of the absorption bands of a carote- 

Kinetics and absorption difference 
spectra. Figure 8 shows absorbance 
changes at 495 m/x seen in the red 
alga Iridaea splendens, which we have 
attributed to a shift of absorption of 
a carotenoid. The absorbance change 
produced by low-intensity orange light 
(622 mp, absorbed mainly by system 
2), were relatively simple (trace d). 
Illumination produced a slow decrease 
of absorbance followed by a steady 
state and a reversal upon darkening. 
Red light (684 nn<,, absorbed mainly 
by system 1) produced more compli- 
cated kinetics. A rapid negative 
transient occurred, followed by a 
slower increase to a steady state 



Fig. 8. Light-induced changes of absorbance AA at 495 m/x in the red alga Iridaea splendens 
induced by actinic light of 684 and 622 m/x. The intensities, I, are expressed in nanoeinstein cm -2 

— i 



above the dark level (traces a, b). 
The response induced by light of 622 
ni/x of high intensity was rather vari- 
able. Here the first decrease was fol- 
lowed by a second, slower one (trace 
c) , but sometimes it was followed by 
a slow and small increase in absorb- 
ance. Kinetics at 515 rmt were simi- 
lar but opposite in sign, suggesting 
that they were at least partly caused 
by the same compound. Kinetics at 
these wavelengths indicated that the 
changes were not caused by P700 or 
cytochrome /; moreover, these com- 
pounds have little absorption in this 
spectral region. 

Light-induced absorption difference 
spectra were measured in the region 
440 to 540 m^ in Iridaea. The differ- 
ence spectrum of the steady-state 
change produced by orange light is 
shown in Fig. 9. A very similar spec- 
trum resulted from plotting the rate 
of change instead of the steady state. 
Figure 10 shows difference spectra 
obtained with red actinic light and the 
same conditions as for trace a of Fig. 
8. The open circles give the deflection 
after 20 msec of illumination, which 

is the time needed to achieve the 
maximum negative deflection. The 
curve with the solid circles gives the 
difference between the maximum de- 
flection and the steady state, corre- 
sponding to the slow reversal of the 
signal at 495 m/x. 

The spectra of Figs. 9 and 10 are 
all similar and show maxima and 
minima at about 465, 480, and 495 
m/x, and a maximum (or shoulder) at 
515 m/x, which suggests that they are 
due to the same compound. The shape 
of the spectrum of Fig. 9, which 
seems uncomplicated by other 
changes in the region above 510 m/x, 
suggests that it is caused by a shift 
toward longer wavelength of a com- 
pound with three maxima at or some- 
what below 445, 473, and 504 m/x. The 
location and distance of the maxima 
(30 m/x) suggest that this compound 
is a carotenoid. 

The difference spectrum of the red 
alga Schizymenia pacifica induced by 
orange light, shown in Fig. 11, was 
very similar to that of Iridaea, al- 
though the changes were much 
smaller, and showed maxima and 





+ 1.0 

- r 

l i i 


i 1 

i l i 






N^ - 





-i n 


1 I i 



l I I 





Fig. 9. Light-minus-dark difference spectrum of Iridaea produced by 1 second of illumination 
with actinic light of 622 m/x (1.1 nanoeinstein cm -2 sec -1 ). 







+ 0.5- 







Fig. 10. 


Wave length, mfj 
Difference spectra of Iridaea produced by light of 684 m^t (17.7 nanoeinstein cm -2 

sec -1 ). Open circles: absorbance change, AA, produced during the first 20 msec of illumination. 
Solid circles: spectrum for the reversal of the absorbance change during illumination (difference 
between maximum deflection produced initially and the steady state; see Fig. 1, trace a). 

minima at about 452, 463, 480, 495, gests a similar shift in carotenoid 

and 511 m/t. The difference spectrum absorption as in the other red algae, 

of Porphyra perforata, also shown in but the spectrum is apparently more 

Fig. 11, had maxima and minima at strongly distorted by other absorption 

470, 483, 497, and 515 m/x, which sug- changes. This was also true of the 

"i r 

i i 1 r 






Fig. 1 1 . Difference spectra of the red algae Schizyrnenia pacifica and Porphyra perforata pro- 
duced by light of 622 m/x (1.5 and 8.7 nanoeinstein cm -2 sec -1 , respectively). 



brown alga Phaeostrophion irregular e 
(Fig. 12). 

The shape of the difference spec- 
trum (Fig. 13) for Botrydiopsis 
alpina (Xanthophyceae) was very 
similar to that of Iridaea and had 
maxima at 450, 482, and 520 and 
minima at 468 and 498 m^. 

We were unable to distinguish a 
carotenoid shift in green algae (Chlo- 
rophyceae), since the presence of 
chlorophyll b — which produces the 
large negative change at 475 and the 
positive change at 515 nut (Year Book 
65, p. 473) — obscured other changes 
in this region. However, we were 
fortunate to have a mutant strain 
(Chlorina 2) of barley (Hordeum 
vulgar e), which lacks chlorophyll b 
(Highkin and Frenkel, 1962) but is 
capable of growing to maturity. This 
mutant, which was obtained by 
Robertson in 1933, was made avail- 
able to the Department through the 
courtesy of Dr. Harry R. Highkin. 
The plant produced a difference spec- 
trum (Fig. 14) which was still com- 
plicated by other absorbance changes 
but nevertheless had maxima and 
minima suggesting a shift in absorp- 
tion of a carotenoid. 

We did not invariably observe the 
carotenoid absorbance changes. In 
Iridaea, at times, we were unable to 
observe them. We have thus far not 
seen them in the blue-green algae 
examined (Synechococcus cedrorum 
and Schizothrix calcicola). In barley 
we previously saw ( Year Book 65, p. 
474) only another change with a peak 
near 525 nut and no negative changes 
below 500 mcifx. This change may dis- 
tort the spectrum of Fig. 14. 

Pigment system responsible for the 
absorption shift. The kinetics of the 
absorption changes in light mainly 
absorbed by system 1 or by system 2 
illustrated in Fig. 8, suggest that the 
shifts in carotenoid absorption were 
driven by both pigment systems; ex- 
citation of system 1 causes a rapid 
change, followed by a slow reversal, 
and system 2 causes only a slow 
change of absorption. Additional 
evidence that the rapid spike (as in 
trace a of Fig. 8) is brought about by 
system 1 was obtained by experiments 
with DCMU which showed no inhibi- 
tion of the spike by this compound. 
For Iridaea the relative efficiency of 
684- or 622-m/x light for bringing 
about the carotenoid absorption shift 






| 1 1 1 


I 1 


1 1 

1 1 

+ 1.0 




p - 





! t i i i ! » i i i 1 



Wavelength, mp 

Fig. 1 2. Difference spectrum of the brown alga Phaeostrophion irregulare produced by light of 
622 m/j, (1.3 nanoeinstein cm -2 sec -1 ). 







1 i ' 



' 520 ' 




+ 0.5 





-i n 

i ! i i 

i i 



i i i 





Fig. 13. Difference spectrum for the steady-state absorbance changes produced in Botrydiopsis 
alpina by actinic light of 672 m/x (8.5 nanoeinstein cm -2 sec -1 ). 

was measured, and the results (Table 
2) indicated that the relative activi- 
ties were the same as for cytochrome 
oxidation, a well-known system 1 re- 

The shifts in carotenoid absorption 
are remarkably similar in properties 
to those seen at 475, 515, and 650 m/x 
in the Chlorophyceae and in other 
green organisms containing chloro- 
phyll b (Year Book 63, p. 441 and 
Year Book 65, p. 473). In both cases 
excitation of system 1 gives rise to a 
rapid transitory change upon illumi- 
nation, followed by a reversal upon 
darkening. The initial, fast carote- 
noid shift, like the 515-nu/, change, is 
dependent upon the preceding dark 
time, and increases up to a certain 
maximum with increasing dark time. 

It has been argued (Fork, Amesz, 
and Anderson, 1967; Witt et al., 
1967) that the 515-m/x change in 
green algae is not produced by an 

oxidation - reduction reaction but 
rather by a change in the environ- 
ment of a compound, probably chlo- 
rophyll b, which gives rise to a rela- 
tively small change in its absorp- 
tion spectrum. The change in environ- 
ment would then be caused in some 
way by electron transport in system 1 
or 2. The same might be true for the 
carotenoid shift. There is evidence 
(Amesz and Vredenberg, 1966) that 
the shift in carotenoid absorption in 
the purple bacterium Rhodopseu- 
domonas spheroides is not caused by a 
chemical reaction, because the ab- 
sorption of one quantum of light by 
this organism resulted in a shift of 
the absorption of about 3 molecules of 
carotenoid. A similar analysis of the 
changes observed here has, so far, 
not been possible because the location 
of the absorption bands of the carote- 
noid is not known. 











Fig. 14. Difference spectrum for the initial 
change of absorbance in attached mutant bar- 
ley leaves (lacking chlorophyll b) produced by 
excitation with a broad band of red light, from 
about 620 to 800 m/x (3.6 X 10 5 ergs cm" 2 
sec -1 ). A schedule of 6 seconds light and 12 
seconds dark was used. 


Amesz, J., and W. J. Vredenberg, in 
Currents in Photosynthesis, edited by 
J. B. Thomas and J. C. Goedheer, Ad. 
Donker, Rotterdam, p. 75, 1966. 

Fork, D. C, J. Amesz, and J. M. Ander- 
son, in Energy Conversion by the 
Photo synthetic Apparatus, Brookhaven 
Symposia in Biology, No. 19, Brook- 
haven National Laboratory, Upton, 
N.Y., p. 81, 1967. 

Highkin, H. R., and A. W. Frenkel, Plant 
Physiol, 37, 814, 1962. 

Robertson, D. W., Genetics, 18, 148, 1933. 

Witt, H. T., G. Doring, P. Schmidt- 
Mende, U. Siggel, and H. H. Stiehl, in 
Energy Conversion by the Photosyn- 
thetic Apparatus, Brookhaven Sympo- 
sia in Biology, No. 19, Brookhaven 
National Laboratory, Upton, N.Y., p. 
161, 1967. 

Quenching by Quinones of 

Chlorophyll Fluorescence in 


Jan Amesz and David C. Fork 

During the past years quinones 
have attracted interest as possible 
intermediates in photosynthesis. 
There is evidence that plastoquinone 
is an intermediate in the photosyn- 
thetic chain, and as early as 1954 
Wessels made the hypothesis that a 
quinone functions as the primary 
photooxidant of photosystem 2. Qui- 
nones have been found to act as Hill 
oxidants, as cofactors of photosyn- 
thetic phosphorylation, and as inhibi- 
tors of photosynthesis. 

Livingston and Ke (1950) reported 
that benzoquinone and substituted 
benzoquinones quench the fluores- 
cence of chlorophyll in organic solu- 
tion, and more recently some observa- 
tions have been published about 
quenching of fluorescence of chloro- 

TABLE 2. Relative Activity of Light of 684 and 622 mp, in Causing the Shift in 
Carotenoid Absorption and Cytochrome Oxidation in Iridaea splendens 


Flash Interval 


Activity Ratio, 
684/622 m/x 

Carotenoid shift 
Cytochrome oxidation 

1 min 
3 sec 
1 min 
3 sec 
6 sec 







The measurements were done at 495 m//, for the carotenoid shift and at 420 mfx for the cy- 
tochrome oxidation. Actinic light was given in 1 0-msec flashes with dark intervals as indicated 
in the second column. The last column gives the activity of absorbed quanta of 684 m^ relative 
to those at 622 m/*. The concentration of DCMU, where used, was 5 X 10 -5 M. 



phyll in isolated chloroplasts by 

We have studied quenching of 
chlorophyll fluorescence in Swiss 
chard chloroplasts and algae. The re- 
sults indicate that the quenching proc- 
ess is caused by direct interaction 
of chlorophyll and quinone molecules 
and not by stimulation of photosyn- 
thetic electron transport, as has been 

The effect of DCMU and quinones 
on fluorescence kinetics. Figure 15 
shows recordings of the time course 
of fluorescence for Swiss chard chlo- 
roplasts in the absence and in the 
presence of 10~ 5 M DCMU. The initial 
level of fluorescence upon illumination 
is about the same under both condi- 
tions but with DCMU the fluorescence 
rises much faster and to a higher level 
than in its absence. According to 
Duysens and Sweers (1963) this in- 
crease is caused by the reduction of 

Q, the primary photooxidant of sys- 
tem 2. The slower rise in the absence 
of DCMU probably reflects the reduc- 
tion of a larger pool, which may give 
rise to the oxygen burst reported 
earlier (Year Book 61, p. 334), and 
which tends to keep Q in the reduced 
state initially. In the following we 
will call the final level of fluorescence 
the "total fluorescence" and the 
difference between the total and the 
initial fluorescence will be called the 
"variable fluorescence." 

Figure 15 also shows the effect of 
2-methyl-l,4-naphthoquinone (mena- 
dione, vitamin K 3 ). Menadione 
strongly quenched chlorophyll fluores- 
cence, especially the variable fluores- 
cence, which was already strongly 
quenched at 1.8 X 10 -5 M, a, concen- 
tration that hardly affected the initial 
fluorescence. The quenching effect as 
a function of concentration is shown 
in Fig. 16. 


i 1 i 1 


i 1 i 




(~ DCMU 
I0" 5 M Menadione 



5xlO" 5 M Menadione 




i i i 1 

I0~ 3 M Menadione 

i i t i 


Time, sec 

Fig. 15. The effect of DCMU (10~ 5 M) and of DCMU with three different menadione concen- 
trations on the kinetics of fluorescence in Swiss chard chloroplasts. The excitation light was a 
band centered at 545 m/x for the traces on the left and at 420 m/i for the traces on the right. 
The intensities and vertical scales for the traces on the left and right, which were made with 
separate samples, are different. A 24-second dark period preceded each illumination. 














Chloroplasts, Swiss chard 

Initial fluorescence 





I0~ 4 
Menadione cone. (M) 

Fig. 16. Quenching of the initial, variable, and total fluorescence of Swiss chard chloroplasts 
as a function of menadione concentration in 1 0~ 5 M DCMU: <£ is the relative fluorescence yield in 
the absence of quencher and <£' in the presence of quencher. The excitation light was green (545 

A number of other quinones were 
tested for quenching activity. Table 3 
summarizes the results. Like mena- 
dione, most quenchers acted much 
more strongly on the variable than on 
the initial fluorescence. Some qui- 

nones such as phthiocol and lawsone 
showed little quenching activity but 
others were equal to or more active 
than menadione. Reduced menadi- 
one and 2,3,5,6-tetramethyl-benzo- 
quinone had little or no quenching 

TABLE 3. Quenching of Initial and Variable Fluorescence in Swiss Chard Chloroplasts 1 


Initial Fluorescence Variable Fluorescence 

Concentration (ftM) for 50% Quenching 



2, 3-dimethy!-p-benzoquinone 


2, 3, 5, 6-tetramethylbenzoquinone 


1 , 4-naphthoquinone 


2-methyl-l, 4-naphthoquinone 



2-hydroxy-l , 4-naphthoquinone 


> 2000 

2-hydroxy-3-methyl-l , 4-naphthoquinone 


> 1000 

5-hydroxy-l , 4-naphthoquinone 



1, 2-naphthoquinone 


1, 2-naphthoquinone-4-sulfonic acid 

> 2000 



1, 2-dihydroxyanthraquinone 










> 2000 








* Fluorescence was excited by green light (545 m/x, 2 to 6 • 1 3 ergs cm -2 sec -1 ), after a 
preceding 24-second dark period. The DCMU concentration was 10~ 5 A4. 



activity. Several quinones were also 
tested for their effect on fluorescence 
of intact algae. Menadione and 1,4- 

naphthoquinone, which were active 
with chloroplasts, were found to be 
active with algae too, although gen- 
erally at a somewhat higher concen- 
tration. Lawsone and phthiocol, in- 
active with chloroplasts, were in- 
active with algae also. Figure 17 
illustrates the effect of 1,4-naphtho- 
quinone on the red alga Porphyra 

Mechanism of the quenching effect. 
There are several indications that the 
quenching, which occurred both with 
DCMU and without it, is not caused 
by an oxidation of reduced Q (QH) 
by quinones: 

1. The E ' of the quencher varied 
widely: between + 0.18V (2,3-dimeth- 
ylbenzoquinone) and about —0.40V 
(alizarin); and many of these com- 
pounds should be unable to oxidize 

2. The quenching effect was inde- 

pendent of the intensity of the ex- 
citing light, and the regeneration of 
Q in the dark after an illumination 
period was found not to be affected 
by menadione. 

3. With relatively high concentra- 
tions of quencher the fluorescence 
yield was considerably lower than 
that observed (in the absence of 
quencher) under conditions where Q 
could be assumed to be fully oxidized 
(e.g., during efficient photosynthesis, 
or shortly after onset of illumina- 
tion) . 

4. In several experiments both 
with chloroplasts and with algae we 
did not see any reversion by quinones 
of the inhibition of oxygen evolution 
by DCMU. 

The above observations indicate 
that the quenching effect is not caused 
indirectly by stimulation of the rate 
of electron transport, but rather by a 
direct interaction of quinone and 
chlorophyll molecules, which then act 
as traps for the excitation energy. 




Porphyra perforata 


6xlO" 5 M 
1,4- Naphthoquinone 



DCMU 4.8xlO" 4 M 

1,4- Naphthoquinone 



I.2xl0- 4 M 

,4- Naphthoquinone 


Time, sec 

Fig. 17. Fluorescence kinetics of Porphyra perforata in the presence of DCMU (5 X 1 -5 A4) 
and 1 ,4-naphthoquinone. Fluorescence was excited by green light. The preceding dark intervals 
were 1 8 seconds for the traces on the left, and 30 seconds for the traces on the right. The traces 
on the right were recorded at about 20 times higher light intensity and lower sensitivity of the 
apparatus than those on the left. 



Since the fluorescence lifetime of 
chlorophyll is longer when Q is in 
the reduced state than when it is 
oxidized (Miiller and Lumry, 1965), 
a stronger quenching can be expected 
for the total (and thus for the vari- 
able) than for the initial chlorophyll 
fluorescence, even if the quencher acts 
only on the bulk of the pigment. It is 
doubtful, however, if this could ex- 
plain the large difference in the ex- 
tent of quenching of the initial and 
variable fluorescence, as observed at 
low concentrations of menadione 
(Fig. 16). 

Additional evidence that the qui- 
nones act at a site close to system 2 
was given by experiments shown in 
Fig. 18. Measurements of absorption 
changes at 420 ni/x in Porphyra in- 
dicated that menadione and other 
quinones inhibited the reduction of 

the /-type cytochrome by light ab- 
sorbed mainly by system 2. Quinones 
thus exhibited the same effect as well- 
known system 2 inhibitors such as 

In agreement with the results of 
earlier experiments of Wessels 
(1954) we found that a number of 
quinones inhibited the Hill reaction 
of chloroplasts with 2,6-dichloro- 
phenol-indophenol (DCPIP). In gen- 
eral, strong quenchers were more 
powerful inhibitors of cytochrome 
reduction and of the Hill reaction 
than were weak quenchers. However, 
a quantitative relation was not found. 
Menadione gave 50% inhibition of 
DCPIP reduction at 2 X 10" 4 M, 
about the same concentration as 
needed for 50% quenching of the 
initial fluorescence, but with 1,4- 
naphthoquinone, lawsone, and phthio- 









1 i i i i I i — i — i — i i i i — n — | — i 1 — i — i i i | — r~i — I — I — I th r 

Porphyra perforata 

No Menadione I0~ 4 M Menadione 
iRedh- -Dark — — I Red I Dark 

AI/t +0.005 

iRedl Green 

^" i »'* ' 

y >( in w «W >»>^> » 



l i i i i I ' i ' i l i ' i i I ' 

t »» i ' i i i i i I i i 


Time, sec 


Fig. 18. Effect of menadione on the cytochrome absorbance changes in Porphyra perforata. 
Red actinic light, to excite mainly system 1, had a half band between 670 and 780 m^t and green 
actinic light to excite system 2 from 550 to 570 m/x. A downward trace corresponds to the oxi- 
dation of cytochrome. 



col the effect on cytochrome and 
DCP1P reduction was stronger than 
that on fluorescence. 

Emission spectra. We measured the 
emission spectra of Porphyra with 
green exciting light in the presence 
and in the absence of 1, ^naphtho- 
quinone and DCMU, using the appa- 
ratus for automatic recording of 
fluorescence emission spectra con- 
structed by French (Year Book 65, 
p. 493). The emission spectra (Fig. 
19) showed three maxima at 658, 
684, and 731 m/i. The first one is pro- 
duced by phycocyanin and the second 
by chlorophyll a. The maximum at 
731 nijn was first observed by Duysens 
(1951). As Fig. 19 shows, naphtho- 
quenched the chlorophyll 
684 ni/x and, somewhat less 
that at 731 m/*, but had no 
effect on the intensity of phycocyanin 
fluorescence. This indicates that the 
quinone does not interact with the 
biliprotein chromophore, and that the 

band at 

transfer of energy between phyco- 
bilins and chlorophyll is not affected. 

DCMU alone caused about a three- 
fold increase of the 684-m^ band and 
a 2.5-fold increase of the 731-mju 
band, but had no effect on the phyco- 
cyanin emission. 

If it is assumed, according to cur- 
rent theories, that DCMU stimulates 
only the fluorescence of pigment sys- 
tem 2, then these data indicate that 
the pigment fluorescing at 731 m/x be- 
longs not only to system 1 (as indi- 
cated by experiments of Duysens, 
1951) but also to system 2. Naphtho- 
quinone then quenches system 2 
fluorescence, possibly by shortening 
the lifetime of the excited state of 
chlorophyll a fluorescing at 684 m^, 
and thus decreasing both the 684-m/x 
fluorescence and the extent of the 
transfer of energy to the pigment 
fluorescing at 731 m/x. 

Conclusion. Our experiments dem- 
onstrate that a number of substi- 



cS 75 








i 1 1 r 

1 1 1 r 

~i r 

Porphyra perforata 



700 750 


Fig. 1 9. Fluorescence emission spectra, excited by green light, plotted as relative number of 
quanta per frequency interval of Porphyra perforata in DCMU (5 X 1 0~ 5 M) and DCMU with 
1 ,4-naphthoquinone (9 X 1 -5 M). The broken line gives the difference between the two spectra. 


tuted quinones strongly quench the thetic Bacteria, Special Issue of Plant 
fluorescence of chlorophyll in vivo. and Cell Physiol, Jap. Soc. of Plant 
The quenching in vivo occurs at a Physiologists, The Univ. of Tokyo 
much lower concentration of quinone Press, p. 353, 1963. 
than when the chlorophyll is dis- Livingston, R., and C. L. Ke, J. Am. 
solved in organic solvent (Livingston Chem. Soc, 72, 909, 1950. 
and Ke, 1950). The quenching is Muller, A., and R. Lumry, Proc. Natl 
probably due to direct interaction be- Acad. Sci., U.S., 54, 1479, 1965. 
tween chlorophyll and quinone mole- Seelv > G - R > in The Chlorophylls, edited 
cules, rather than to stimulation of ^ L - P. Vernon and G. R. Seely Acad, 
electron transport. It is possible that Press > New York-London, p. 523, 1966. 
the actual quenching mechanism (see ^ {JL S * C *' Re °' TmV ' Ch%m '> 73 ' 
Seely, 1966, for a discussion) is the ' 
same in vivo as in chlorophyll solu- 
tion. The stronger quenching in vivo Some Essential Considerations in 
may be caused by a concentration of the Measurement and Interpreta- 
the quinone in the lipoid part of the tion of Absorption Spectra of 
chloroplast lamellae, and also may be Heterogeneous Samples 
explained by the fact that quenching James M> pickett and c & French 
of the same fraction of chlorophyll 

molecules, by formation of traps, The following discussion of well- 
could produce a stronger quenching in known but, of necessity, often neg- 
vivo than in a dilute chlorophyll solu- lected factors is given here primarily 
tion. The low activity of p'hthiocol as a review of basic principles essen- 
and l,2-naphthoquinone-4-sulfonic tial to the study of absorption spectra 
acid may be due to salt formation in of photosynthetic cells. These con- 
the water phase. siderations are particularly important 

There is no direct spectrophoto- for spectra that are to be analyzed in 

metric evidence (Amesz, 1964) for terms of th eir component forms of 

the hypothesis that a quinone is the chlorophyll. 

primary photooxidant for system 2. Sieve effect. The principal source 

However, the strong quenching of of error in measuring the shape of 

chlorophyll fluorescence in vivo indi- absorbance curves with thin suspen- 

cates that association of a quinone sions is that some light gets through 

with a chlorophyll molecule can pro- spaces between the cells. This makes 

duce an efficient trap for the excita- the valu e of transmitted light greater 

tion energy. It is possible that a than it would be if the pigment com- 

photosynthetic reaction center of plexes were uniformly dispersed in 

system 2 consists of a similar trap, solution. The sieve effect makes the 

supplemented by the enzymatic and P eaks of the absorbance curve lower 

structural arrangement necessary for than they should be, relative to re- 

the formation of stable photochemical g ions of lower absorbance. 

products. The error introduced by the sieve 

effect becomes smaller as the number 

References of particles increases because the 

Amesz, J., Biochim. Biophys. Acta, 79, absorbance of a given wavelength for 

257, 1964. various light paths then becomes more 

Duysens, L. N. M., Nature, 168, 548, uniform throughout the suspension. 

1951. However, at high suspension concen- 

Duysens, L. N. M., and H. Sweers, in trations the effective path length is 

Studies on Microalgae and Photosyn- reduced for wavelengths of high ab- 



sorbance relative to that for wave- 
lengths of low absorbance (Butler, 
1964). Thus peaks in absorbance 
spectra of dense algal suspensions are 
usually flattened because scattering 
(Latimer, Year Book 56, p. 259) and 
reflectance (Rabideau ctal, 1946) are 
not independent of wavelength, while 
the sieve effect is negligible. 

The absorbance of a suspension 
relative to that of the same pigment 
in solution can be calculated for a 
single layer of absorbing particles. 
The calculations are complex for sus- 
pensions of more than one layer. The 
case of oriented cubic particles has 
been described in detail by Duysens 
(1956). The analysis in the case of 
spherical particles is more complex 
because the absorbance is not uniform 
over the projected area of the spheri- 
cal particles. 

Sieve effect for a single layer of 
uniform particles. Let / be the frac- 
tion of the light beam that encoun- 
ters a particle of transmission T p . If 
mutual shading of the particles is 
negligible, the measured absorbance 
E s of a dilute suspension of uniform 
particles is 

E 8 = log ( ^ (1) 

\a-f) +fT P ) 

But the apparent absorbance of a 
single particle is 

have in the same volume of solution 
with a uniform path length. 

Spherical particles. Consider a 
sphere of radius r suspended in a 
medium of equal refractive index il- 
luminated by a circular beam of light 
whose radius is equal to that of the 
sphere. Assume that each spherical 
particle contains a concentration c of 
pigment whose absorption coefficient 
is e. The absorbance along a light 
path of length x through the sphere is 

E x — eCX (3) 

The transmission along the same path 

T B = 10- 


All paths of length x through the 
sphere are in a circle of radius y per- 
pendicular to the incident beam of 
light where 

/ x 

y* — r * .__ / — 

»2 — 


The absorbance E, which is propor- 
tional to the extinction coefficient of 
the pigment, and transmission T p of 
a single sphere are obtained by inte- 
grating the weighted values of E x 
and T x over all light paths (Duysens, 
1956, and Pickett, 1965) 


* 2irydy 

&x : — 



E v = log 



= — ecr 


The important conclusion is that the 
measured absorbance of dilute sus- 
pensions is not proportional to the 
apparent absorbance of one particle. 
Moreover, if all light paths through 
the particle are not equal, E p is not 
proportional to the extinction coeffi- 
cient of the pigment present in the 
particle. Therefore, we propose to re- 
view the types of errors generally in- 
troduced by absorbance measurements 
of suspensions relative to the absorb- 
ance E that the same pigments would 

T v = IT 


-n-r 2 

lQ2ecr _ 4 6 06 eCr _ i 
10.61 (ecr) 2 10 2£Cr 


from which 
E p = log 

10.61 (ecr) 2 10 2< 

lQ2ecr _ 4> 606 ecr — 1 


It is clear from equation 8 that E p 
is not proportional to e, the absorp- 



tion coefficient of the pigment within 
the sphere. The ratio of the apparent 
absorbance of a single sphere, E p , to 
the absorbance of the uniformly dis- 
persed pigment E is given in Fig. 20, 
for various values of 2ecr, the absorb- 
ance through the center of a single 

The transmission (10 _2ecr ) for 675 
ni/A of single Euglena chloroplasts as 
measured by Wolken et al. (Strother 
and Wolken, 1959; Wolken and 
Strother, 1963) falls in the range of 
0.4 to 0.6 (2ecr = 0.6 to 0.8) and for 
Chlorella, about 0.6 (2ecr = 0.8). 
The corresponding ratio E p /E is 0.94 
to 0.92. Flattening of the absorption 
peak may be much less than 6% to 
8% in these microspectrophotometric 
measurements because the absorption 
spectrum was determined by measur- 
ing the transmission of a single 
chloroplast with a beam of light whose 
cross section was small compared to 
that of the chloroplast. 

In practice it is generally more 
convenient to measure the transmis- 
sion of a relatively dilute suspension. 
The ratio of the measured absorb- 
ance E s of a suspension with negli- 
gible overlap of particles to the ab- 

sorbance E of the uniformly dispersed 
pigment is plotted as a function of 
the absorbance through the center of 
the sphere, 2ecr, in Fig. 21. For par- 
ticles with a maximum absorbance 

(2ecr) of greater than 0.5, the ab- 
sorbance maxima are reduced by more 
than 10%. The error becomes greater, 
the smaller the fraction / of the light 
beam intercepted by particles. Thus 
accurate absorbance spectra of dilute 
suspensions can only be obtained for 
particles of very low maximum ab- 

Dense suspensions. The sieve effect 
is relatively insignificant in very tur- 
bid suspensions, since the effective 
path length is much greater than the 
thickness of the suspension. However, 
the path length is not increased to 
the same extent at all wavelengths. 
The principal variable influencing 
effective optical path length is particle 
reflectance (Butler, 1964) . Anomalous 
dispersion, which is maximum on the 
long-wavelength side of absorption 
bands (Latimer, Year Book 56, p. 
259), is relatively insignificant except 
when a very small solid angle of light 
is collected by the spectrophotometer 

(Butler, 1964). 







1 1 1 


-*— * 


















1 T 

i r 

j L 

i i i 

1.0 2.0 

Absorbance through sphere center, 2ecr 

Fig. 20. The apparent absorbance E p of a single spherical particle divided by the absorbance 
E of the pigment in the same cylindrical volume of solution with path length 1 .33 scr, is plotted 
as a function of the absorbance through the center of the sphere 2 scr. The spherical particle of 
radius r contains a uniform concentration c of pigment with absorption coefficient e. 




1.0 2.0 

Absorbance through sphere center, 2ecr 

Fig. 21. The measured absorbance E 8 of a 
thin suspension of spherical particles divided 
by the absorbance E of the pigment in solution 
of equal cross section and path length, 1.33 
ecr[f), as a function of absorbance through the 
center of a spherical particle, 2 ecr. Each 
curve is characterized by f, the fraction of the 
light beam which encounters a particle. 

According to Butler (1964) accu- 
rate absorption spectra of dense sus- 
pensions can be obtained by correct- 
ing the measured spectrum for 
variation of effective path length with 
wavelength. The flattening may also 
be greatly reduced by suspending the 
pigmented particles in a much greater 
concentration of white powder. 

In summary, there are three prin- 
cipal means of measuring in vivo ab- 
sorption spectra which may closely 
approximate the absorption spectra 
of the pigments : (1) The absorbance 
may be measured along a fixed light 
path with a light beam whose cross 
section is small compared to that of 
the pigment complex (microspectro- 
photometry). (2) The absorption 
spectra of very weakly absorbing cells 
such as pale-green mutants may be 
measured by means of thin suspen- 
sions in opal glass containers. (3) 
According to Butler (1964) the ab- 
sorption spectra of very dense sus- 
pensions can be measured in the 
presence of large amounts of white 
powder or corrected for variation of 
effective path length with wavelength. 

The height ratio H. We need a rough 
test for the likelihood of serious dis- 
tortion in an experimentally measured 
curve. For this purpose with green 
plants we have used the height ratio 
H, defined as the absorbance at the 
secondary chlorophyll a maximum at 
about 625 ni/x divided by the height 
of the red absorbance peak at about 
675 m/x. 

As a standard of comparison we 
have taken the value of H for the 
alga showing the smallest ratio. 
This is Claes Chlorella mutant 871 
grown in the dark, which appears to 
have nearly all its chlorophyll as 
C a 672 and lacks chlorophyll b. For 
that alga H = 0.28 at -196°C. 

In an alga containing two forms of 
chlorophyll with absorption spectra 
of similar shape but shifted in wave- 
length, the value of H would be larger 
and would depend on the wavelength 
separation interval. This variation of 
H was obtained graphically for a hy- 
pothetical system having two identical 
spectra but shifted in wavelength by 
various amounts. The basic curve 
shape was taken as the sum of the 625 
and 671 components that fit the ab- 
sorption curve of Fig. 26 for dark- 
grown Chlorella mutant 871. This 
basic curve is the closest approxima- 
tion to the actual shape of the spec- 
trum of C fl 670 that we have. In 
calculating H for the hypothetical 
two-component spectrum, curves of 
the same height were used. For any 
other proportion the change of H with 
the separation interval would be 

The results in Table 4 show that 
the ratio in question would be only 
0.36 for the maximum separation in- 
terval of 13 expected between "C a 
670" and "C a 680." Ratios for experi- 
mental spectra of algae containing 
only chlorophyll a that are larger 
than 0.36 can therefore be taken as 
an indication of distortion by the 
flattening effects. 



TABLE 4. The ratio of the Orange, "625"- 
mfji, Peak to the Red, "675" m/x, Peak of a 
Hypothetical in vivo Chlorophyll a Spectrum 
Composed of Equal Parts of Two Identical 
Components at Several Wavelength 
Separation Intervals 




Ratio of Orange 


to Red Maximum, H 











Butler, W. L., Ann. Rev. Plant Physiol., 
15, 451, 1964. 

Duysens, L. M. N., Biochim. Biophys. 
Acta, 19, 1, 1956. 

Pickett, J. M., Thesis, U. Texas, Austin, 

Rabideau, G. S., C. S. French, and A. S. 
Holt, Am. J. Bot., 33, 769, 1946. 

Strother, G. K., and J. J. Wolken, Sci- 
ence, 130, 1084, 1959. 

Wolken, J. J., and G. K. Strother, Ap- 
plied Optics, 2, 899, 1963. 

A Spectrophotometer Primarily 

for Light-Scattering Samples 

at Low Temperature 

C. S. French and Mark Lawrence 

In Year Booh 65 (p. 498) a plan 
was discussed for modifying the de- 
rivative spectrophotometer so it could 
also be used as a double-beam instru- 
ment for absorbance measurements 
of highly scattering samples. The 
integrating sphere then under dis- 
cussion has been abandoned. Instead 
of the sphere, a different type of sam- 
ple chamber was constructed to be 
particularly suitable for absorbance 
measurements of cell suspensions. 

To reduce scattering errors the ma- 
jor requirement in the design of the 
sample holder was that it should 
gather a representative fraction of 
the transmitted light from the sample 

over a wide angle. Four other re- 
quirements were that it should (1) 
accommodate sample holders that 
could be used at liquid nitrogen tem- 
perature; (2) have a peak height of 
the plotted absorbance curve for a 
particular sample adjustable to any 
desired value and still have known 
scales in absorbance or transmission 
units; (3) by means of an entirely 
transistorized electronic measuring 
circuit eliminate maintenance prob- 
lems of vacuum tube amplifiers ; and 
(4) be simple enough for routine use 
by different people. The device, which 
meets these requirements reasonably 
well, has been in use since April 1967. 
A brief description is given here. 

Optical system. The sample holders 
shown in Fig. 22 use Shibata's opal 
glass diffusion principle for both 
windows. The aluminum spacers also 
serve as heat-conducting elements and 
supports. Since the opal glass win- 
dows are held on with silicone grease 
and oiled modeling clay only, they do 
not break when the sample freezes. 
A holder for the usual 1-cm-square 
liquid absorption cells can be used 
in place of the opal glass sandwiches 
when the material is not to be frozen. 

The sample chamber is shown in 
Fig. 23; just behind the sample holder 
is another opal glass window. Between 
this window and the photocell is a 
light-mixing tube with white walls. 
This tube helps to give uniform illu- 
mination on the photomultiplier tube 
for light falling on any part of the 
opal glass window at its entrance. 
The brass box in the sample chamber 
is thermally insulated by 2.5 to 4 cm 
of plastic foam. The lower part of 
this box can hold 100 cc of liquid 
nitrogen or other liquid for operation 
at a desired temperature. The en- 
trance window is lucite 2.54 cm thick, 
preceded by a double-walled glass 
chamber to reduce fogging at low 
temperature. A turret can be rotated 
to bring any one of three multiplier 


Sample cell 



yj \J 


Opal glass windows 


cone grease seal 

Liquid NL level 

Aluminum spacer » 

Fig. 22. The sample holder — an opal glass sandwich. The aluminum spacer and support is also 

a heat conductor. 

tubes into position for different spec- 
tral regions. A filter-holder attach- 
ment can be used in front of the 
photomultiplier for special experi- 

The optical system, long in use for 
derivative spectrophotometry, has 
been kept. This is a tungsten lamp 
with a Bausch & Lomb monochro- 
mator having a 10 X 10 cm grating 
blazed for green. The slits isolate a 
beam of 1 rrnx half -width. Glass color 
niters are used to reduce the stray 
light which is inevitable from a single 
monochromator. The large monochro- 
mator is particularly valuable in that 
plenty of light is available for meas- 
urements of scattering samples with 
three sheets of opal glass in the beam. 

A rotating half mirror alternates 
the beam from the sample to the ref- 
erence cell about 10 times a second. 
Adjustable sectors on the mirror 
shaft pass light from separate bulbs 

to photocells that activate the appro- 
priate electronic circuits while either 
the sample or the reference cell is 

Electronics. The electronic measur- 
ing system uses solid-state opera- 
tional amplifiers throughout. The first 
stage is a preamplifier with auto- 
matic gain control that makes opera- 
tion possible over a wide range of 
input currents from the photomulti- 
plier. In Fig. 24 two block diagrams 
show the operating principles and 
controls for recording absorbance or 
transmittance. The operating zero- 
position control, not shown, is ad- 
justable to facilitate recording of the 
difference spectra between two sam- 

In many spectrophotometers the 
logarithm needed for absorbance 
plots is generated by a logarithmic 
slide-wire in the recorder. Change of 
scale can then be done only by chang- 



Fig. 23. The sample chamber. 

ing slide-wires. In the instrument de- 
scribed the logarithm is produced by 
a transistor in the feedback circuit 
of an operational amplifier. The loga- 
rithmic signal appears as an electrical 
voltage that can be amplified or at- 
tenuated. This makes it possible to 
adjust the scale of the recording to 
compare curve shapes for different 

When the "curve size" control is at 
its extreme position, the 10-inch full- 
scale deflection of the recorder cor- 
responds to the value of absorbance 
or transmission selected by the range 
switch. On any scale the actual plot 
can be reduced in size by the curve 
size control. When this is done cali- 
bration marks for either absorbance 
or transmission can be put on the 
record by setting a "calibrator" dial 
to the values desired. 

Absorption Spectra of Chloro- 
phyll a in Algae 

C. S. French 

Much of the present-day research 
on photosynthesis is concentrated on 
two closely related questions. One of 
these is to find means by which the 
light energy absorbed by photosyn- 
thetic pigments is transferred from 
one pigment to another, then trapped 
and converted to chemical energy as 
a change in the oxidation state of a 
reaction center. The other problem is 
to define the succeeding reactions of 
specific substances in the pathways 
of electron flow from the primary re- 
action center that makes products 
useful to plants. 

There are two types of primary re- 
action centers believed to contribute 
energy to the electron flow system at 






















8JIM 8p||s japjODay 






























































two points in a series-linked system lated with well-measured absorbance 

of redox compounds. These two reac- spectra for the same cultures. 

tion centers are activated by different Ideally we would like to be able to 

pigment systems each of which con- separate each chlorophyll a form in 

tains predominantly, but not exclu- a pure state without change of its 

sively, one of the different "forms" spectral properties, measure its ab- 

of chlorophyll a and other pigments, sorbance and fluorescence, and also, 

Thus system 1 is largely powered by for a well-defined reaction, its action 

light absorbed by C (2 680 and system spectrum. Work on the difficult prob- 

2 by light absorbed by chlorophyll b, lem of separating the chlorophyll 

C«670, and the "accessory pigments," forms is progressing in several lab- 

carotenoids and phycobilins. oratories and some of it is described 

Considering the large amount of elsewhere in this report. 
intensive work being done on the The present report will discuss 
functioning of these photosynthetic some attempts to derive absorption 
systems it is remarkable that so little spectra of the individual forms of 
is known about the absorption spec- chlorophyll a by analysis of new meas- 
tra of the individual pigments ini- urements of absorption spectra of 
tially catching the light used to drive algae, and mainly of those algae lack- 
the two systems of photosynthesis, ing chlorophyll b. From past at- 
Most of the available information on tempts, largely by derivative spec- 
the partition of absorbed light be- troscopy, it seems that each of the 
tween the two systems has come, not forms designated as C a 670, C o 680, 
from absorption spectroscopy, but C a 695, etc., may represent classes of 
from measurements of action spectra, pigment complexes rather than defi- 
About all we know is that in green nite chemical compounds with identi- 
plants significantly more light is ab- cal spectra in different plants. This 
sorbed by system 1 than by system 2 conclusion is, however, by no means 
at wavelengths beyond 685 m/x. How- certain because small amounts of un- 
ever, system 2 still absorbs some recognized components could well 
light even at longer wavelengths be- shift the apparent position of the ob- 
cause of the overlapping spectra of served curves away from their true 
the pigments in the two systems, positions in the spectrum. 
Furthermore, at 650 and 480 m/x, the Some absorption spectra of intact 
bands of chlorophyll b, system 2 ab- algae. Spectra of some algae useful 
sorption is larger than at other wave- for characterizing the forms of chlo- 
lengths. Some system 2 action spectra rophyll a are given in Fig. 25; and 
show a significant contribution from Table 5 lists values of H, the peak 
C,;670 and some others appear to height ratio for comparison with 
involve C a 680. Beyond this qualita- chlorophyll a in solution. Figure 
tive information only a few attempts 25 (A) gives the spectra of Tribonema 
have been made to specify the exact at two temperatures. These curves 
proportion of any wavelength that are probably flattened but the sharp- 
goes to each functional system. Some ening by low temperature clearly 
few measurements of system 1 and shows the doublet structure of the 
system 2 action spectra have been main red band, which is not evident 
made with precision comparable to at 22 °C. The original record indicates 
that attainable with absorbance meas- a possible, but uncertain, doublet 
urements. However, action spectra structure of the 625 band, perhaps 
for the two systems in various species with broad components at about 620 
still have not been precisely corre- and 630 m/t. The bands at 496 and 



TABLE 5. Ratios of the Orange Band Height 
to that of the Red Band for Chlorophyll a 



Chlorophyll a 







Special Solvent* 


Claes Chlorella 

871 D 



520 D 



515 D 




0.30 (0.36) 








Highkin barley 






Ochromonas refrozen 


* Freed and Sander, 1951. 

463, and some of the absorption at 
shorter wavelengths, are presumably 
due to carotenoids, whereas the other 
labeled bands are attributed to forms 
of chlorophyll a. 

Figure 25(B) gives the spectra of 
Claes wild-type Chlorella at two tem- 
peratures. The low ratio of the peak 
absorbance at about 625 to about 670 
mfi (H = 0.31, at 22°) and the very 
small absorbance at 530 m/x shows a 
surprising lack of flattening in this 
spectrum for a green alga. The 486- 
and 651-m/x bands are attributed to 
chlorophyll b ; those at 492, 467, 460, 
and some shorter absorption bands, 
to carotenoids. The 433 and 437 peaks 
might be due to different forms of 
chlorophyll a. 

In Fig. 25 (C) , we have spectra for 
the alga Botrydiopsis and for a leaf 
of Highkin's barley mutant, both 
lacking chlorophyll b. The curves ap- 
pear strongly flattened but show the 
two major forms of chlorophyll a, 
and the barley has a band at about 
710 m/x. 

Claes Chlorella mutants 515, 520, 
and 871, when grown in the dark, con- 
tain very little chlorophyll and no 
chlorophyll b, thus providing excel- 
lent material for the study of the 

shapes of the chlorophyll a spectra. 
Furthermore, mutant 871 appears to 
be free of carotenoids absorbing in 
the visible spectrum. Figure 25(D) 
shows large blue peaks, believed to 
be zeta-carotene, which obscure the 
blue chlorophyll a spectrum. Figure 
25(E) for mutant 520 shows longer 
wavelength carotenoid spectra. 

In Fig. 25(F) we have a close 
approximation to the spectrum of 
C a 672 with only a small percentage of 
C a 680 and probably very little inter- 
ference from carotenoids. Its shape is 
discussed in detail below. An ethanol 
extract of chlorophyll from mutant 
871 was chromatographed by Dr. 
Michel- Wolwertz who found two frac- 
tions — ordinary chlorophyll a and a 
component spectrally similar to, but 
chromatographically distinct from, 
pheophytin. This suggests that the 
spectrum for mutant 871, while it is 
the simplest we have yet seen, may be 
composed of other components be- 
sides C o 670. She found the other 
Claes mutants grown in the dark to 
contain only chlorophyll a. 

Analysis of observed spectra. Our 
ultimate hope is to derive curve 
shapes for spectra of the individual 
chlorophyll forms that, when added 
together in correct proportions, 
would match any observed spectrum 
that was not distorted. Until that can 
be done it is helpful to compare ex- 
perimental spectra by the relative 
proportions, peak positions, and half- 
widths of the minimum number of 
normal probability curves that, when 
added together, match the measured 

We have previously used deriva- 
tives of probability curves to analyze 
derivative spectra in order to ap- 
proximate the relative content of 
different forms of chlorophyll a 
(Brown and French, 1959). The 
curves for the major components 
used for Chlorella were broader at 
673 than at 683 m/x. Cederstrand et al. 







500 550 600 




Fig. 25. The absorption spectra of various algae; all except the wild-type Chlorella lack 
chlorophyll b. (A) Tribonema Sp. (Starr 639). The curves at two temperatures are directly compa- 
rable. The sharpening at low temperature shows the red chlorophyll a band to contain two forms 
of chlorophyll a not evident in the room temperature spectrum. (With D. C. Fork.) (B) Claes wild- 


(1966) have matched absorbance ure and the table. The peak wave- 
spectra of Chlorella, plotted on a length position, the half -widths, and 
wave number scale, with probability the proportion of the total area for 
curves peaking at 668 and 683 m/x the gaussian curves of Fig. 26 are 
and having equal half -widths of 21 given in Table 6. 
m/x. For estimating the shape of the Mutant 520 appears to contain 
narrow bands here considered we more C a 680 than do 871 and 515. Its 
have used wavelength rather than best fit came from component curves 
wave number scales. The discussion nearly like those for 871 and 515 but 
will be limited to spectra measured slightly different, as shown in Table 
at — 196°C that appear to be free of 6 and Fig. 26(D). Whether or not 
flattening. these small differences of 1 to 2 m/x 

Starting with two spectra having in peak position and bandwidth re- 
most of the chlorophyll as C a 670 we quired to match the spectra of differ- 
see in Fig. 26 (A) that for mutant 871 ent mutants are significant remains a 
a single probability curve does not fit question. 

the main peak. With the same prob- These curve analyses have pointed 
ability curve at 672 m/x and 22 m^ out the previously disregarded fact 
half -width, an attempt was also made that a large fraction of the light inci- 
te match the spectrum of mutant 515 dent on a photosynthetic cell is ab- 
as shown in Fig. 26(B). The misfit sorbed by the secondary chlorophyll 
for both spectra is taken to mean that a bands having broad maxima in the 
components other than a single chlo- 625-m/x region. Table 6 shows that 
rophyll a form are present. 40 % of the total absorption from 

Therefore, enough gaussian curves about 570 nut to the red end of the 

were taken to give a close fit for mu- spectrum should be attributed to 

tant 871 (disregarding the slight these bands. 

error in the sensitive trough region The secondary orange band is pre- 
at 650 m.[i) . The curves used are given sumably different in peak position 
in Fig. 26(C) and Table 6. Then, with and width for the various forms of 
the same peak positions and half- chlorophyll. It is so wide that wave- 
widths as for 871, the heights were length peak differences probably 
adjusted to come as near as possible amounting to about one fifth of its 
to matching mutant 515 with unsat- width are obscured. There certainly 
isfactory results that are not shown, are, however, differences in the shape 
However, the slight changes in band of the orange bands in different algae 
positions and widths recorded in that may be worth further study 
Table 6 gave a good fit to 515, as after the variations in the red band 
shown in Fig. 26 (D) . For mutant 515 are better understood, 
the secondary component had to be A particularly interesting effect, 
slightly different, as shown in the fig- susceptible to curve analysis, is re- 
type Chloreila, dark grown. The curve heights are not directly comparable. The low value of 
H, 0.31 at 22°, indicates very little flattening in this sample. (C) Botrydiopsis afpina and High- 
kins mutant barley both lacking chlorophyll b. Both curves appear strongly flattened. (With D. C. 
Fork.) (D) Claes Chlorella mutant 515, dark grown. The large blue peaks are probably zeta- 
carotene. The red band is shown at about 20 X the scale for the blue. VV3/4 = 15 m/x, H = 
0.295. Probably free from flattening. (E) Claes Chlorella mutant 520, dark grown. The 51 5, 482, 
and 453 bands are probably carotenoids. W3/4 = 17 m/A, H = 0.345. Probably free from flat- 
tening. (F) Claes Chlorella mutant 871 at two temperatures. The red band is nearly all C a 672. 
W3/4 = 14 m/x, H = 0.29. Probably free from flattening. 














700 600 



Fig. 26. Analyses of the red chlorophyll a absorption bands of several Claes Ch/ore//a mu- 
tants grown in the dark and of Ochromonas danica, in collaboration with Dr. Brown. The char- 
acteristics of the gaussian component curves are listed in Table 1. The sum of these components 
is shown as a dotted line where it deviates from the measured curve being matched. 

ported in another section by Dr. 
Brown. This is in Ochromonas danica, 
an alga containing chlorophyll c 
rather than chlorophyll b. The three- 
peaked spectrum of the frozen sample 
changed to a two-peaked spectrum 
after thawing and refreezing. At- 
tempts to match the three-peaked 

curve with a small number of gaus- 
sian curves were not fruitful. How- 
ever, the spectrum of the ref rozen ma- 
terial was fitted successfully as shown 
in Fig. 26(F). By changing only the 
relative heights, but not the peak 
positions or widths of these curves 
matching the ref rozen sample, it was 



TABLE 6. Probability Curves Fitting Algal Absorbance Spectra 


band 620 

Chlorophyll c 

695 680 670 


Chlorella Mutant 871 

Peak, m/A 

695 681 672 


Half-width, m/x 

(11) 10 20 



Proportion, % 

1 3 55 

Chlorella Mutant 51 5 


Peak, m/jL 

695 683 672 


Half-width, m/x 

(11) 11 21 



Proportion, % 

1 4 53 
Chlorella Mutant 520 


Peak, mtt 

694 681 671 


Half-width, n\[x 

20 10 19.5 


• • • 

Proportion, % 

3 12 39 
Qchromonas Refrozen 


Peak, mfji 

696 686 671 



Half-width, m/i 

16 16 22 



Proportion, % 

14 7 39 



not possible to match the three-peaked 
curve of the same sample before 
freezing. The nearest approach to a 
fit is illustrated in Fig. 26(G). This 
difference suggests that the spectra of 
the components themselves were af- 
fected by the treatment. Just from 
looking at the curves before and after 
ref reezing we had the impression that 
the relative proportions of the bands 
but not their characteristics had been 
influenced. This does seem to be more 
nearly true for a repetition of this 
experiment with another culture 
shown in Fig. 34 of Dr. Brown's 

Because mutant 871 seems to be 
free of carotenoids we fitted the blue 
part of its spectrum with gaussian 
curves with the results given in Fig. 
27. Whether or not these represent 
only components of chlorophyll a re- 
mains to be decided when other ca- 
rotenoid-free samples are found in 
algae or in preparations such as the 
Si fraction shown in Fig. 30 of Mr. 
Michel's report. Analysis of the blue 
and of the red spectra of carotenoid- 
less samples are needed to determine 

which of the blue bands correspond 
to the forms of chlorophyll that are 
now identifiable only by their red 

In conclusion we may say that the 
present best available approximation 
to the red part of the spectra of C a 

' 1 ' 

Mutant 871 y 

i i i l 

I i 


// j 

'4I8\ \ 



^^^ j 


\ / 44l \\ 



/390\ / 





L-^i l i- 

-^r I i >- 

Vv 467 
CiV I """^ 

400 450 


Fig. 27. The blue spectrum of Claes Chlorella 
mutant 871 is approximately matched by the 
sum of the following gaussian curves described 
by peak position, half-width and percentage of 
the total area of their sum: 390 m/x, 27 m/x, 
27%; 418 m/x, 29 mti, 47%; 441 mtt, 18 mxt, 
24%; 467 m/x (18 m^), 2%. 



"670" is described by the sum of two 
gaussian curves, with the following 
characteristics: (1) A max = 622 mp, 
\\\ „ = 60 im t ; (2) A max = 672, 
W a s = 22 m,u, band height ratio (1)/ 
(2) = 0.27 (because of overlap, the 
band-height ratio, (1) / (2) , is not ex- 
actly the same as the total curve- 
height ratio, H) . 


Brown. J. S., and C. S. French, Plant 
Physiol, 54, 305, 1959. 

Cederstand, C. N., E. Rabinowitch, and 
Govindjee, Biochim. Biophys. Acta, 
126, 1, 1966. 

Freed, S., and K. M. Sancier, Science, 
114,275, 1951. 

Electrophoretic Study of the 


Complexes of Euglena 

J. M. Michel 

Many investigations, some from 
this laboratory, have shown that the 
red part of the absorption spectrum 
of green organisms results from the 
juxtaposition of at least three dis- 
crete forms of chlorophyll a, presum- 
ably lipoprotein complexes. They are 
identifiable only from differences in 
their absorption spectra. We know 
very little about their actual chemical 
properties. We do not know if the 
differences between these substances 
are due to different arrangements of 
chlorophyll bound to the same carrier, 
to different carriers for the same 
pigment molecule, or to some other 
set of conditions. 

Most of the attempts to separate 
these forms have been by differential 
centrifugation using different ways 
of solubilizing the chloroplast mate- 
rial. Only a few attempts have 
used electrophoresis, although zone 
electrophoresis methods are now 
available with properties particularly 
adapted for the study of such com- 

ponents. Among these methods, acryl- 
amide gel electrophoresis is of par- 
ticular interest because it is possible 
to tailor the size of the pores inside 
the gel to the size of the particles 
being studied. (Ogawa et al., 1966; 
Thornber et al., 1967). We used ac- 
rylamide gel electrophoresis to study 
the protein-chlorophyll complexes of 
chloroplast fragments of Euglena. 


Euglena gracilis, Indiana Culture 
Collection No. 752, grown for 21 days 
in low light intensity, was used. When 
so grown, Euglena contains an ap- 
preciable amount of a longer wave- 
length absorbing chlorophyll form, 
C a 695 (Brown and French, 1961). 
We prepared the chloroplast frag- 
ments using the needle-valve disinte- 
grator and the anionic detergent Na- 
deoxycholate (DOC) as a solubilizing 
agent according to the method of 
Brown et al. (1965). 

This method gives two fractions on 
centrifugation for 30 minutes at 
30,000 g; a sediment called here Sed 
30,000 and a supernatant called Sup 
30,000. Prior to electrophoresis, the 
fractions in tris-HCl buffer 0.01 M, 
pH 8.9 were mixed with an equal vol- 
ume of a solution of sucrose 40% in 
tris-glycine 0.022 M, pH 8.9. 

Disc electrophoresis was performed 
on the fractions Sed 30,000 and Sup 
30,000, according to the method of 
Steward (1965). Before electrophore- 
sis the sample was layered on top of 
the concentration gel; the buffer in 
the reservoir was tris-glycine 0.02 M, 
pR 8.9. 


The electrophoretic patterns of the 
two fractions Sed 30,000 and Sup 
30,000 are shown in Fig. 28. 

Sed 30,000 did not penetrate the 
gel. The stationary band is labeled S . 
The Sup 30,000 fraction, however, 
showed two clearly separated bands 



Sup 30,000 

Sed 30,000 

b,. , 

So ** 


No Amido- 
stain black 



Fig. 28. Electrophoretic patterns of Sed 
30,000 and Sup 30,000. 

Sx and S 2 moving to the anode and 
some material which did not move 
into the gel. This S band could be a 
contamination of Sup 30,000 by par- 
ticles of Sed 30,000. 

The Si band was blue-green and the 
S 2 band was green with a yellow 
front ; under an ultraviolet lamp only 
the S 2 band showed a visible red fluo- 
rescence. Figure 28 shows also the pat- 
tern after coloration of the proteins 
by the specific dye Amido Black B10. 
This shows the absence of other 
mobile proteins in the centrifugal 
fractions used for electrophoresis. 

The red part of the low-tempera- 
ture absorption spectra of the bands 
Si and S 2 as well as the sample stay- 
ing at the starting point for Sed 30,- 

000, are shown in Fig. 29. The absorp- 
tion spectra of the bands S x and S 2 at 
room temperature are given in Fig. 
30 for the entire visible spectrum. 

The absorption spectra of Si and S 2 
are almost identical in the red, hav- 
ing a maximum at 670 mp and lack- 
ing the "long-wavelength" absorbing 
form. There is, however, a minor 
difference in that Si has relatively 
higher absorbance than S 2 near 660 
and 690 m/x and less at about 670 m/ju 
Sed 30,000 has the same absorption 
spectrum as the original extract; it 
contains most of the chlorophyll b. 

Figure 30 shows clearly that the 
band S 2 contains most, if not all, of 
the carotenoids, whereas the band Si 
appears to be free of carotenoids. This 
result is similar to that of Ogawa et 
al. (1966) who found a differential 
separation of the carotenoids using 
sodium laurylsulfate as solubilizing 
agent for spinach chloroplasts. Brown 
et al. (1965) achieved a similar sepa- 
ration by differential centrifugation 
of DOC-treated Euglena chloroplasts. 

The fluorescence spectra of the 
bands Si and S 2 are reported in Fig. 
31; the maximum of emission is at 
680 m^. When compared with the 
original extract and Sed 30,000 the 
large emission band at 720 m/x dis- 
appears in the fluorescence spectra 
of Si and S 2 . Comparison of the ab- 

600 650 700 

Wavelength, m/j 

Fig. 29. Absorption spectra at liquid nitrogen temperature of electrophoretic fractions. 




-1 J 1 1 ! 1 J T— " 

-mri i 1 

— i J T — r 

—j. — , — , — , ,_ 






\ N 
\ \ 

\ \ 


! 1 ! ! 1 1 1 


! I t i i 


i i i i 

1 1 1 1 1 1 1 1 



500 550 600 


Fig. 30. Absorption spectra of electrophoretic fractions Si and S2 at room temperature. 


sorption and fluorescence emission 
spectra of Si and S 2 gives direct 
evidence that the 670-m^ absorption 
maximum is correlated with the 680- 
m/i emission maximum. 

From these data we can conclude 
that DOC solubilizes two components 
from the chlorolast lamellae of 
Euglena. The two components are 
different in their electrophoretic mo- 
bilities and in their pigments. Ca- 
rotenoids are associated with the 
S 2 band, but the chlorophyll portion 
of the pigment system is essentially 
chlorophyll a with nearly the same 
spectral properties in both cases. 
From the different electrophoretic 
mobilities it may be inferred that the 
protein portions of the complexes are 
different, whereas the spectra sug- 









1 1 

\ 1 

I 1 1 
\ 24 C C 







/ / 
/ / 

\ s 

\ N 
\ S 






1 1 



700 750 


Fig. 31. Fluorescence spectra of electropho- 
retic fractions at room temperature. 

gest that the chlorophyll a molecules 
are bound to the protein in a similar 
or identical manner. However, we 
cannot yet exclude the possibility that 
the S 2 band could contain some lipids 
dissolved in detergent micelles. 

Two more green bands have been 
separated by disc electrophoresis 
from material treated with sodium 
laurylsulfate. They are much less 
abundant than Si and S 2J which makes 
their study more difficult. 

To see if the different forms of 
chlorophyll are chemically separable 
entities or if the complexes exist 
only in the highly ordered structure 
of the lamellar membrane, we are now 
involved in attempts to separate the 
C a 672, C a 685 and the C a 695 forms as 
such. Preliminary experiments show 
that, with appropriate experimental 
conditions, it is possible to separate 
two different electrophoretic bands 
from Euglena chloroplasts without 
using any detergent. However, so far 
these two bands having identical 
spectra each contain all three in vivo 
chlorophyll a forms. 


Brown, J. S., and C. S. French, Biophys. 
J., 1, 539-550, 1961. 

Brown, J. S., C. Bril, and W. Urbach, 

Plant Physiol, hO, 1086-1090, 1965. 
Ogawa, T., F. Obata, and K. Shibata, 



Biochim. Biophys. Acta, 112, 223-234, 

Steward, F. C, R. F. Lyndon, and J. T. 

Barber, Amer. J. Bot., 52, 155, 1965. 
Thornber, J. P., R. P. F. Gregory, C. A. 

Smith, and J. L. Bailey, Biochemistry, 

6, 391-396, 1967. 

The Chlorophylls Extracted 
from Plants by Organic Solvents 

Marie-Rose Michel-Wolwertz 

Several studies reported in this 
Year Book and elsewhere have dem- 
onstrated the existence of different 
chlorophyll forms in vivo (for in- 
stance, Year Book 58, p. 278; Year 
Book 59, p. 330; Year Book 65, p. 
483). It is generally considered that 
these "in vivo forms" differ in the ar- 
rangement and (or) state of aggrega- 
tion of one single type of chlorophyll 
molecule with the lipoprotein carrier. 

Meanwhile, Sironval et al. (1965) 
showed that two chemically different 
pigments, namely, chlorophyllide a 
and chlorophyll a, were responsible 
for the two forms C 684 and C 672 pres- 
ent in vivo during the initial period 
of greening of etiolated leaves. 

Michel-Wolwertz and Sironval 
(1965) have isolated several chloro- 
phyllous pigments from ethanol ex- 
tracts of Chlorella by means of paper 
chromatography. These are in addi- 
tion to the common chlorophylls a and 
b (called a x and &i). They suggested 
that these "satellites" might be re- 
sponsible for the different absorbing 
forms in vivo. 

The purpose of the present study 
was to test this possibility. 

1. We tried to find out whether the 
light regime received by a plant 
would influence the pigment composi- 
tion of its extract. 

2. We studied the effect of the 
presence of air or of nitrogen during 
extraction and chromatography upon 
the pigment composition of extracts 
of different species. 


Chlorella pyrenoidosa, Pringsheim 
(211/8b) and Euglena gracilis were 
used for extraction of pigments. For 
comparison we also used spinach 
leaves obtained from local stores. 

Chlorella and Euglena were ex- 
tracted with boiling 95% ethanol. Im- 
mediately after extraction, the pig- 
ments were transferred to petroleum 
ether (b.p. 45-60 °C) by the addition 
of excess water. Spinach leaves were 
extracted by grinding with sand and 
pure acetone in a mortar. Pigments 
in acetone were directly transferred 
to diethyl ether by adding excess 
water. Extractions were made in dim 
white light. Pigments were separated 
by two successive paper chromato- 
grams. In this study, we considered 
pigments of the chlorophyll a type al- 
most exclusively, that is, pigments 
separated by rechromatography of 
the spot of chlorophyll a obtained 
from the first chromatogram. 

For some experiments, we used a 
chamber to do both extraction and 
chromatography under nitrogen. 


Pigment composition of extracts 
from Chlorella pyrenoidosa cultivated 
under various light exposures. Table 
7 compares the pigment composition 
of extracts from Chlorella pyrenoid- 
osa cultivated under these conditions : 
(1) 11 days in continuous light; (2) 
11 days in continuous light, then 2 
days in complete darkness. (In this 
case the extraction was made in the 
dark) ; (3) 11 days in continuous 
light, then two days in complete dark- 
ness, after which the Chlorella re- 
ceived a 4-msec light flash before 
extraction. (The extraction was made 
in the dark immediately after the 

The relative proportions of the dif- 
ferent pigments are about the same 
in all the extracts. These results show 
that in Chlorella pyrenoidosa the pig- 



TABLE 7. Pigment Composition of Extracts from Chlorella pyrenoidosa 
Grown Under Various Light Regimes 

Pigments, % of Total Chlorophyll a 

Light Regime 


°2 + 3 

a 4 


































Light 1 1 days 

Light 1 1 days + dark 2 days 

Light 1 1 days + dark 2 days 
+ 1 flash of 4 ms 

merit composition does not vary ac- 
cording- to the light regime received 
by the algae. 

Effect of oxygen during extraction 
and chromatography on the pigment 
composition of extracts of different 
species. Table 8 gives the relative 
proportions of the different pigments 
found in extracts of Chlorella, Eu- 
glcna, and spinach, when all the pro- 
cedures of extraction and chromato- 
graphy were done in air. The results 
show that the proportions of the sat- 
ellites are not the same for the three 
species. For example, the same solvent 
extracts of Euglena are rich in satel- 
lites compared to those of Chlorella. 
The fact that the proportions of the 
satellites vary from one organism to 
another had been interpreted as in- 
dicating either that these satellites 
were present in the living cells and 

the proportions varied in the different 
organisms studied or that the common 
chlorophylls a and b were differently 
altered from various plants during 
extraction and chromatography. 

In order to decide between the two 
possibilities we extracted and chro- 
matographed the extracts from Chlo- 
rella, Euglena, and spinach, both in 
the presence and in the absence of 
oxygen. The results (Table 9) show 
that extraction and chromatography 
in the absence of oxygen reduces or 
prevents formation of satellite chloro- 
phylls. This is true also for the chloro- 
phyll b type satellites. In the absence 
of oxygen we found only bi, but in the 
presence of oxygen, satellites b 2 , b 3 , 
etc., were also present. The absence 
of oxygen during extraction was 
especially effective in preventing the 
formation of satellites (Table 10). 

TABLE 8. Proportions of the Different Pigments in Extracts from 
Chlorella, Euglena, and Spinach Leaves 

Pigments, % of Total Chlorophyll a 



Q2 + 3 

a 4 
















7 A 






Chlorella pyrenoidosa 

Euglena gracilis 

Spinach leaves 





* Chlorella and Euglena were extracted with boiling 95% ethanol; spinach leaves, with pure 
acetone (extraction and chromatography were done in air). 



TABLE 9. Influence of the Presence or Absence of Air During Extraction and 
Chromatography on the Proportions of the Satellite Pigments 


ments, °/ 

9 of Total Chi 

orophyll a 







C*2 + 3 

a 4 



°2 + 3 

a 4 

a 5 















Spinach leaves 









* Chlorella and Euglena were extracted with boiling 95% ethanol; spinach leaves in pure 
acetone. For each extraction the same sample was used under air and under nitrogen. 

However, satellites were also formed 
in the presence of oxygen during 
chromatography (Table 11). In this 
experiment the same volume of a 
chlorophyll solution was deposited on 
six paper chromatograms in the 
presence of air. The time required for 
depositing the solution was varied 
from 1 to 10 minutes. It is clear from 
the results that longer exposure to 
air destroys a ± and consequently 
creates more satellite chlorophyll, 
especially a 2+3 and a 5 . 


From these considerations, the con- 
clusion is inescapable that all the sat- 
ellites may be formed from the parent 
compounds by chemical alterations 
during extraction and also during 
chromatography. This does not ex- 
clude the possibility that enzymatic 
alteration may occur during extrac- 
tion. Such alteration might account 

for the different proportions of satel- 
lites in extracts from different spe- 
cies. However, the extraction of spin- 
ach was made by grinding in acetone 
and this procedure was slower than 
extraction in boiling ethanol. 

It appears very unlikely that these 
satellite chlorophylls are present in 
living cells, since they are almost 
completely eliminated if extraction 
and chromatography are done in the 
absence of oxygen. From this work it 
appears that they are not related to 
the different in vivo forms of chloro- 
phyll a. 


Michel-Wolwertz, M. R., and C. Sironval, 
Biochim. Biophys. Acta, 9b, 330-343, 

Sironval, C, M. R. Michel-Wolwertz, and 
A. Madsen, Biochim. Biophys. Acta, 
94, 344-354, 1965. 

TABLE 10. Influence of Oxygen During Extraction and Chromatography on the 
Proportions of the Satellite Pigments from Euglena gracilis 


Pigments, % of Total Chlorophyll a 


1 st Chroma- 

2nd Chroma- 


02 + 3 

a 4 

a 5 






























TABLE II. Quantities of the Different Pig- 
ments Found on the Chromatogram as a Func- 
tion of the Time in Air between Application of 
the Chlorophyll Solution to the Paper and the 
Start of Chromatography, optical density* 

Relative Qi 

jantities of the Chlorophyll a 






































* The values given are the peak optical 
densities of the separated pigment bands dis- 
solved in 4 ml of ether, 1 -cm light path. 

Chlorophyll Fluorescence 
in Algae and Chloroplasts 

J. S. Brown 

Some fluorescence spectra of par- 
ticular algae whose spectra show great 
contrasts with each other were meas- 
ured at 20 °C and at — 190 °C. Very 
small amounts of the algae were sus- 
pended in a wet paste of BaS0 4 to 
reduce the difference in light scatter- 
ing between the room temperature 
and the frozen samples. The incident 
light was 436 im*. The curves of Fig. 
32 were all recorded at essentially 

the same size to facilitate comparison. 
However, the apparatus sensitivity 
adjustments were recorded to allow 
comparisons of the relative intensity 
of emission at the two temperatures. 
The ratios between the yields at 
— 190 °C and at 20 °C are given in 
Table 12. 

Any one of these curves is pre- 
sumed to be the sum of a number of 
overlapping fluorescence spectra each 
of which is characteristic of or 
emitted by a particular form of chlo- 
rophyll. My objective was to deter- 
mine the basic fluorescence spectrum 
for each of the pigment types present. 
The amount of fluorescent light radi- 
ated by any one pigment in compari- 
son with that from other pigments de- 
pends directly on the relative number 
of incident quanta it absorbs or re- 
ceives from other pigments, and in- 
versely, on the rate at which the pig- 
ment transfers quanta to other pig- 
ments or photochemical traps in the 
chloroplast. The efficiency of energy 
transfer and trapping can vary with 
temperature. Therefore a direct pro- 
portionality between the amount of 
any chlorophyll form, as determined 
from absorption measurements, and 
the intensity of its fluorescence emis- 
sion cannot be expected. The correla- 
tions sought are between the wave- 

TABLE 12. Ratios of the Fluorescence Yield at -190°C to the Yield at 20°C for Selected 
Wavelengths and for the Total Fluorescence between 650 and 770 m/x 


680 m/x 

691 mju, 

720 m/jt, 

740 m/x, 


High-light grown 
Low-light grown 







High-light grov/n 
Lov/-light grov/n 



























length positions of absorption peaks 
and their corresponding fluorescence 

Last summer Mr. Stephen J. 
Fulder attempted to derive the fluo- 

rescence spectra of the individual 
forms of chlorophyll by comparisons 
of a large number of fluorescence 
spectra at low temperature measured 
in different laboratories. His results 













750 650 700 



Fig. 32. Fluorescence emission spectra of several algae and spinach chloroplasts measured at 
20°C and -190°C. Excitation at 436 m/*. 


showed wide variations in the shapes (1964) ascribes a long wavelength 

of the individual bands when they (730 ni/x) emission maximum ob- 

were deduced from different sets of served from chlorophyll in solution. 

data. The present work should pro- Brody (1962) observed this band at 

vide a more coherent group of spectra low temperature and attributed it to 

for such curve analysis. a chlorophyll dimer. However, Goed- 

Some qualitative conclusions al- heer's experiments demonstrate that 
ready have become evident. In the this band occurs in dilute as well as 
first place the composite nature of the concentrated solutions and at both 
major 685-mp peak (Year Book 58, 20°C and — 196°C. It is relatively 
p. 328) as the sum of components at greater in more concentrated solu- 
longer and shorter wavelengths ap- tions because of self -absorption of the 
pears even more plausible. Secondly, main emission band at shorter wave- 
since these measurements were made lengths. Thus a significant proportion 
with a very dilute suspension of algae of long wavelength fluorescence in 
in a thick paste of BaS0 4 , it is not vivo may arise from the second vibra- 
likely that the difference between tional level intrinsic in the chlorophyll 
spectra at the two temperatures can a molecule. This hypothesis to account 
be attributed entirely to reabsorption for a long wavelength emission band 
distortion by greater light scattering requires no corresponding long wave- 
at low temperature. Therefore the length absorption band. On the other 
actual lowering of the fluorescence at hand, if the emission is from a dimer 
680 and 691 m/x shown in Table 12 is or aggregate as Brody suggests, an 
believed to show an increased effi- absorption band for the dimer must 
ciency in transfer of energy to longer exist. Possibly C a 695 is aggregated 
wavelength forms at low tempera- chlorophyll, but the nature of this 
ture. absorbing form of chlorophyll has not 

Long ivavelength emission. The yet been demonstrated. The wave- 
following evidence shows that there length position of the long wavelength 
are several different long-wavelength fluorescence band in a particular 
emission bands in addition to F-680 sample depends on the proportion of 
and F-690. In Euglena gracilis, Phae- the two types of fluorescence. 
odactylum tricornutum, and Ochro- Short wavelength emission. Duy- 
monas danica, chlorophyll a-695 may sens (1952) first provided evidence 
accumulate to a proportion of the that the chlorophyll a absorbing at 
total chlorophyll easily detectable in longer wavelengths was largely non- 
absorption spectra, but it is also more fluorescent in red and blue-green 
labile than the chlorophyll forms ab- algae. More recently Duysens and 
sorbing at shorter wavelengths. When Sweers (1963) correlated the major 
the amount of C a 695 is relatively portion of fluorescence with the ac- 
large, the fluorescence band F-720 is tivity of photosystem 2 in which the 
also more evident, and when it is par- shorter wavelength absorbing chloro- 
tially destroyed F-720 also decreases, phyll a is considered to be functional. 

No one has yet observed an absorp- These findings led to the assumption 

tion band that would be a likely that C a 670 is the source of F-685, 

source of emission beyond 725 m/x the main fluorescence emission maxi- 

which is so evident in spinach chloro- mum at room temperature. 

plasts and Euglena. Fluorescence be- The variability in position of the 

tween 720 and 740 m t i may result emission maximum in different algae 

from the second vibrational level of at room temperature shown in Fig. 

chlorophyll a to which Goedheer 32 and additional data from experi- 


ments on Ochromonas discussed in at 20°C in the beginning. On the other 

another section of this report indicate hand, the increases in yield at 720 and 

that other forms of chlorophyll may 740, which were originally 2.9 and 3.3 

contribute to the main emission band. (Table 12), were reduced to 1.2 and 

A discrete fluorescence emission band 1.5 after a second freezing. These 

between 690 and 700 m/x at — 190 °C yield changes are reflected in changes 

has been observed in several labora- in shape of the emission spectra (as 

tories. Goedheer (1964) has sug- illustrated in Fig. 34 for Ochro- 

gested that C a 680 is the chlorophyll monas). 

form which fluoresces at about 690 The disruptive thawing and freez- (F-690) both at room tempera- ing process has either lowered the 
ture and — 190 °C. In Ochromonas the possibility of energy transfer be- 
main emission maximum is at 690 m/x, tween pigment molecules or has de- 
suggesting that in this case C a 680 stroyed the longer wavelength ab- 
may be the major fluorescing chloro- sorbing pigment complex C a 695. Ab- 
phyll form. sorption spectra of Euglena, Phaeo- 

Thus we now have evidence that all dactylum, and Ochromonas, measured 

three forms of chlorophyll a may at — 190 °C and again after thawing 

fluoresce under physiological condi- and refreezing, do show a relative de- 

tions but that the proportion of fluo- crease in long wavelength absorption, 

rescence contributed by each form Figure 33 shows an example for 

varies considerably with the algal Ochromonas. 

species and growth conditions. The The structural or specially oriented 

same conclusion was reached several nature of C a 695 is suggested. Only 

years ago concerning the relative ab- those organisms that are rather eas- 

sorption by each chlorophyll a form ily disrupted show a large change in 

in different algae. shape of the emission spectrum at low 

Energy transfer. Although no temperature after a single thawing 

clearly consistent hypothesis concern- and refreezing. Chlorella or spinach 

ing energy transfer emerges from chloroplasts suspended in isotonic 

these studies as yet, some postulations sucrose buffer must be thawed and 

can be made. These corroborate and frozen several times before the yield 

extend those of Goedheer (1966). changes that are induced by cooling 

Upon cooling, the yield of the are different from those observed 

shorter wavelength fluorescence after the first measurement at 

bands always decreases and that of — 190 °C. 

the longer wavelength bands always An explanation for these observa- 
increases. After the cells are damaged tions may be that at very low temper- 
in some way such as heating to 50 °C ature the possibility of energy trans- 
for 10 minutes, placing in 10% fer between C a 670 and C a 695 in- 
ethanol overnight, or freezing and creases, resulting in a decreased 
thawing, the yield changes upon cool- fluorescence yield at 680 m^ and an 
ing are very different. For example, increased yield at 720 m/x. Possibly 
the Euglena grown in high light some energy transfer between C a 670 
showed slightly more than a fivefold and C a 695 occurs also at room tem- 
decrease in fluorescence yield at 680 perature, and C fl 695 acts as an energy 
m/x when the cells were first cooled collecting trap in the photochemical 
from 20 °C to — 190°C. But when this sense. 

sample was thawed and brought Last year we suggested that ac- 

again to — 190 °C, the yield at 680 m/x tivity during photosynthesis of both 

was twice as great as from the cells C a 670 and C a 695 might be observed by 



i I — r 

First frozen 
Thawed, refrozen v. 


■ i i 



500 550 600 

Wavelength, m/x 

Fig. 33. Absorption spectra of Ochromonas danica measured at — 190°C. Solid line curve 
measured with initially frozen cells; dashed curve from same sample, thawed and refrozen. 

monitoring their respective fluores- 
cence emission maxima. This should 
be possible with either Eitglena or 
Phacodactylum cultured in an appro- 
priate manner. Now we can add that 
with Ochromonas fluorescence from 
C a 680 may also occur as a major band 
at room temperature. 


Brody, S. S., and M. Brody, Trans. Fara- 
day Soc, 58, 416, 1962. 

Duysens, L. N. M., Dissertation, Utrecht, 

Duysens, L. N. M., and H. E. Sweers, in 
Studies on Microalgae and Photosyn- 
thetic Bacteria, Japanese Soc. of Plant 
Physiologists, University of Tokyo 
Press, Tokyo, p. 353, 1963. 

Goedheer, J. C, Biochim. Biophys. Acta, 
88, 304-317, 1964. 

Goedheer, J. C., in Biochemistry of 
Chloroplasts, edited by T. W. Goodwin, 
Academic Press, London, p. 75, 1966. 

Absorption and Fluorescence 
Spectra of Ochromonas danica 

J. S. Brown 

The algal flagellate Ochromonas 
danica has a red absorption spectrum 
with three major components at 670, 

682, and 692 m^. Furthermore, both 
its absorption and fluorescence spec- 
tra are very sensitive to mild distur- 
bances such as freezing or gentle 
warming (Allen et al. 1960, and Year 
Books 57 and 58) . The older measure- 
ments that produced these data were 
all done at room temperature and 
only derivative absorbance spectra 
were recorded. 

We have recently extended these 
experiments with improved equip- 
ment and have measured both absorp- 
tion and fluorescence at — 190°C as 
well as at 20 °C. The results, qualita- 
tively similar to those reported ear- 
lier, are more precise and useful for 
comparison with absorption and fluo- 
rescence spectra of other algae dis- 
cussed in other sections of this 

Figure 33 shows absorption spectra 
of an Ochromonas suspension. The 
three adjacent bands evident in a de- 
rivative absorptive spectrum are diffi- 
cult to distinguish in an absorbance 
spectrum at room temperature but 
are clearly visible in the low-tempera- 
ture spectrum. The experimental de- 
rivative spectrum measured at 20 °C 
was fitted previously with three 
gaussian derivative components with 



peaks at 670, 682, and 692 m/x. At- 
tempts to fit the low-temperature in- 
tegral absorption spectrum with simi- 
lar gaussian component bands have 
been unsuccessful. 

Figure 32 in the preceding section 
of this report shows that at 20 °C 
Ochromonas has a single fluorescence 
emission band at about 690 m^. At 
— 190°C, a second broad emission 
maximum appears at about 720 m^. 
The aliquot used for these measure- 
ments was from a culture in which 
essentially all of the cells were ac- 
tively motile. Occasionally spectra 
had a shorter wavelength emission 
band between 670 and 680 m^ in ad- 
dition to bands at 690 and 720 nut. 
The solid curve of Fig. 34 shows an 
example with a distinct shoulder on 
the short wavelength side of the 690- 
m/x band. We found that the relative 
height of the 675-imi band was cor- 
related with the number of rounded, 
nonmotile cells in the culture. 

Thus a particular culture may show 
all three fluorescence maxima at low 
temperature, the relative height of 
the 675-m/x band depending on the 
number of damaged cells. It is, of 
course, possible to damage all the cells 
by freezing and thawing or by homog- 
enizing with the needle-valve press. 
After such treatment the emission is 
nearly all at about 675 m^, as shown 
in Fig. 34 measured at — 190°C, and 
Year Book 58 (p. 329), measured at 

An absorption spectrum of cells 
that have been frozen, thawed, and 
cooled again to — 190 °C is shown in 
Fig. 33. The shortest wavelength 
band near 670 m^ increases, and the 
two longer wavelength bands de- 
crease after ref reezing. A comparison 
of the absorption spectra in Fig. 33 
with the fluorescence spectra in Fig. 
34 before and after thawing and re- 
freezing indicates that the three ab- 
sorption bands are the sources of 
emission of the three fluorescence 

1 L_J L 

650 700 750 


Fig. 34. Fluorescence emission spectra of 
Ochromonas danica at — 190°C, and after 
thawing and refreezing. Excitation at 436 mju.. 

bands. In addition, changes in absorb- 
ance throughout the visible spec- 
trum are evident. When more com- 
parative spectra are available, we 
hope, by correlating some of these 
changes with those in the red region, 
to determine other absorption max- 
ima for the forms of chlorophyll a. 

Chlorophyll a-680 may be the pri- 
mary fluorescent form in Ochro- 
monas. Since C a 680 is thought to be 
the major pigment in photosystem 1, 
the correspondence between fluores- 
cence and photosynthetic activity 
may be different from that in other 
algae where C o 670, which functions 
in photosystem 2, is the main fluores- 
cing form. 


Allen, M. B., C. S. French, and J. S. 
Brown, in Comparative Biochemistry 
of Photoreactive Systems, edited by 
Mary Belle Allen, Academic Press, 
New York, p. 33, 1960. 

The Action Spectrum for 

Blue-Light-Stimulated Oxygen 

Uptake in Chlorella 

James M. Pickett 

Emerson and Lewis {Year Book 
U0, p. 158; and 1943) found the rate 


of oxygen uptake by Chlorella py- of oxygen uptake stimulated by blue 
renoidosa to increase after an ex- light in Chlorella pyrenoidosa. 
posure to low-intensity blue light. Electrode method. We constructed 
Stimulation of oxygen uptake by 480- a Teflon-covered oxygen electrode 
m,a light was increased by keeping which has provision for repeated in- 
cells in the dark for 90 minutes. Fol- jection of samples from a reservoir of 
lowing an exposure to 480 m/a, the dark-adapted cells (Year Book 65, p. 
rate of oxygen uptake remained 487). The ability to compare measure- 
greater than the former dark respira- ments on aliquots of a single suspen- 
tion rate for approximately 1 hour, sion of Chlorella greatly facilitated 

Ried (Year Book 64, p. 399) re- determination of the action spectrum. 

ported that Chlorella pyrenoidosa Optical setup. The image of a tung- 

kept in the dark, then exposed to sten ribbon filament or a mercury arc 

flashes of blue light (A < 540 m/x), was focused on the platinum cathode. 

gave increased rates of oxygen up- Spectral bands were isolated by 

take. The maximum increase in rate Balzers interference filters (half-band 

of uptake occurred 6 to 10 minutes width approximately 10 m/x). Inten- 

after the flash. This increase was sities were measured with a cali- 

not inhibited by 5 X 10~ 6 M DCMU brated photocell. Far infrared was re- 

[3- (3,4-dichlorophenyl) -1,1-dimethyl- moved by 15 cm of water, 

urea]. Under optimal conditions the Culture conditions. Chlorella pyre- 

extra amount of oxygen consumed noidosa (Pringsheim 211-8b) was 

after a short-wavelength flash was grown at 20° C in 1-liter erlenmeyer 

more than 500 times greater than the flasks above cool white fluorescent 

amount evolved photosynthetically lamps giving 400 foot-candles. The 

during the flash. flasks, containing 200 ml of Knops 

Kowallik and Gaffron (1966) medium with added microelements, 

found no blue-light stimulation of were aerated with 2% C0 2 in air and 

oxygen uptake in the nonphotosyn- shaken. After 2 to 3 days the sus- 

thetic alga Prototheca, which has pension containing 3 to 5 /xl cells/ml 

little absorption in the visible. How- was centrifuged and the cells resus- 

ever, after dark adaptation for more pended in medium without microele- 

than 1 hour, the dark yellow mutant ments to give a final concentration of 

of Chlorella vulgaris (211-llh, 50 til cells/ml. This suspension pro- 

M.20), which also lacks chlorophyll, vided a layer 3 to 4 cells thick in the 

did show the blue-light stimulation electrode chamber. A 3-ml aliquot of 

of oxygen uptake. Kowallik and this suspension was placed in the 

Gaffron found that the rate of dark reservoir and kept dark at 20 °C for 

respiration decreased about 60% in at least 15 hours before the experi- 

algae kept in a glucose-free medium ment. 

for 6 hours. However, in blue light, Exploratory experiments to develop 

the initial rate of oxygen uptake was the procedure of measurement. A 

stimulated as much as 30 % for this typical response to a 5-second flash of 

time. blue light is shown in Fig. 35. The 

In the experiments the stimulation maximum rate of oxygen uptake, 

of oxygen uptake was limited to reached 10 to 15 minutes after the 

wavelengths shorter than 550 m>; flash, was measured by the maximum 

therefore, the responsible pigment difference between the extrapolated 

could not be a chlorophyll. These ob- dark current before the flash and the 

servations led us to attempt to deter- measured current after the flash, as 

mine the action spectrum for the rate shown in Fig. 35. The rate of oxygen 






















x +30 




Biue light 
5 sec,469m/j 

Stimulation of 
oxygen uptake 





Fig. 35. The rate of oxygen exchange before and after a 5-sec flash of 469 m/x (0.49 X 1 9 
ein/cm 2 /sec = 1250 erg/cm 2 /sec). The cells were in the dark 47.5 hr before the experiment. 

uptake remained high for at least 30 
minutes. The maximum rate of oxy- 
gen uptake stimulated by a 5-second 
flash of 469 m^ was about 1 % of the 
light-saturated rate of oxygen evolu- 

The stimulation of oxygen uptake 
by blue light developed slowly after 
the cells were placed in darkness, as 
seen in Fig. 36. Almost all data used 
for the action spectrum were meas- 
ured with cells which had been dark 

for 15 to 48 hours. Figure 36 also 
shows that it was necessary to correct 
the measured rates of uptake for the 
change in response which occurred 
during each 6- to 8-hour experiment. 
Preliminary experiments indicated 
that 469 m/x was near the peak of the 
action spectrum. Figure 37 shows 
that the response to 469 m/x was 
nearly linear up to 0.75 nein/cm 2 /sec. 
These data provided the basis for 
selection of approximately 0.5 nein/ 

Time in dark, hr 

Fig. 36. Effectiveness of 5-sec exposures to 0.5 X 1 -9 ein/cm 2 /sec of 469 m/i as a function of 
time in darkness before the flash. 










Light intensity, 
einsteins cm -2 sec -1 x 10~ 9 

Fig. 37. Light saturation curve for uptake after 5-sec flashes of 469 m/x for cells kept dark for 

48 hr. Corrected for change in response with time. 

cm 2 /sec (1300 ergs/cm 2 /sec) of 469 
nin. as the reference beam with which 
other wavelengths were compared. 

To compare the effectiveness at 
several wavelengths it was necessary 
to make measurements over a period 
of several hours. With the tungsten 
lamp the measuring sequence used 
was: 469 m/x, X u A 2 , A 3 , 469 m/x, A 4 , 
A 5 , As, 469 m/x. We attempted to ad- 
just the intensities at all other wave- 
lengths to give approximately the 
same rate of oxygen uptake as that 
measured at 469 m/x. The intensities 
available with the tungsten lamp were 
sufficient from 401 to 496 m/x. The 
mercury lamp gave sufficient inten- 
sity at 366, 407, and 436 m/x. 

Since the effectiveness at 469 m/x 
changed with time, an interpolated 
value was used to correct the results 
at other wavelengths. The corrected 
results were then scaled by normaliz- 
ing to the average effectiveness at 
469 m/x, or at 436 m/x when the mer- 
cury lamp was used. 

Action spectrum. The resulting 
action spectrum is shown in Fig. 

38(A) and is in general agreement 
with Ried's recent results (personal 
communication). The curve was 
drawn to fit the mean effectiveness at 
each wavelength. The bar at 469 m/x 
represents the mean ±3 standard 
deviation units (n = 39) . All but two 
measurements at 469 m/x were within 
two standard deviation units. The 
scatter at other wavelengths is gen- 
erally greater than at 469 m/x, which 
probably indicates that some of the 
scatter was introduced by the nor- 
malization process. If so, the shape of 
the action spectrum for a single sus- 
pension must change somewhat with 
time. The measurements with the 
mercury lines are in good agreement 
with the measurements made with 
bands (half -band width approxi- 
mately 10 m/x) isolated from tungsten 
emission. The effectiveness beyond 
550 m/x was always less than 5% of 
the effectiveness at 460 m/x. The ac- 
tion spectrum was not changed by us- 
ing suspensions of 25 /xl cells/ml in- 
stead of 50 /xl cells/ml. 

The spectrum in Fig. 38(A) sug- 




i r 

i 1 r 

i 1 1 1 1 1 r 




Wavelength, m/j 



Fig. 38(A). Action spectrum for blue-light-stimulated oxygen uptake in Chlorella. (B) Comparison 
of riboflavin absorption in water with the action spectrum corrected for shading by a 1 0-/x thick 
layer of cells. 

gests that the pigment responsible 
for the blue-light-stimulated increase 
in rate of oxygen uptake is a flavin. 
The near-ultraviolet peak at 375 nut 
distinguishes the action spectrum 
from that of a carotenoid. If a cis- 
carotenoid were responsible for the 
effect, this peak should occur below 
330 mil. 

Shading by chlorophylls and carot- 
enoids would shift the observed peak 
of the action spectrum toward the 
red. The absorption spectrum of 
whole cells was therefore measured in 
the Beckman DK-2 spectrophotometer 
with opal glass plates. The corrected 
action spectrum (Fig. 38B) was esti- 
mated by multiplying the observed 
effectiveness (Fig. 38 A) by the re- 
ciprocal of the transmission of a 10 

/x-thick layer of cells (absorbance at 
440 m.fx = 0.575). This correction for 
shading, by a layer half the estimated 
thickness used, shifts the visible ab- 
sorption peak from 460 to 440 m^ and 
introduces a shoulder at 470 m/x. The 
location of the near-UV peak remains 
unchanged at about 375 However, 
the ratio of the visible to near-UV 
peak is increased from 1.15 to 1.42. 
The corresponding value for ribo- 
flavin (Fig. 38B) is 1.17. Correction 
for shading by a thinner layer would 
have given smaller shifts. The fact 
that the corrected spectrum lies 
within the scattering of the measure- 
ments explains why dilution of the 
cell suspension from 50 /J cells/ml 
(20^-thick cell layer) to 25 A J /ml 
did not change the action spectrum 



significantly. It should be noted that 
the absorption spectrum of flavopro- 
teins is generally more complicated 
than that of riboflavin itself. The 
peak height ratio in different flavo- 
proteins is variable, and shoulders 
may occur on the short- or long-wave- 
length sides of the visible peak. 

Because of the scatter in the data, 
one can only conclude that the pig- 
ment responsible for the blue-light- 
stimulated increase in rate of oxygen 
uptake is probably a flavin. 

Effect of poisons. After adding 
10- 5 '" M DCMU in ethanol (final 
ethanol concentration, 1%) the 
steady-state rate of photosynthetic 
oxygen evolution from 0.62 nein/cm 2 / 
sec at 469 m^ was inhibited by 94%. 
The rate of oxygen uptake caused by 
a 5-second flash of 0.62 nein/cm 2 /sec 
was not inhibited under the same 
conditions (Fig. 39). 

By contrast with DCMU, cyanide 
poisons the light-stimulated uptake of 
oxvgen much more than photosyn- 
thesis. With 2 X 10- 4 M KCN the in- 
hibition of light-stimulated oxygen 
uptake was 97 % , but steady-state oxy- 

gen evolution only 28%. A further 
indication that the prolonged increase 
in oxygen uptake by blue light is re- 
lated to respiration is presented in 
Fig. 40. Addition of 1% glucose to 
starved cells stimulated the blue-light 
effect 63 % . In the presence of glucose 
the rate of dark respiration was in- 
creased by an amount 10 times 
greater than the maximum rate of 
light-stimulated oxygen uptake. The 
ratio of the rate of respiration before 
addition of glucose to that after addi- 
tion of glucose is unknown. Only 21 % 
of the increase in dark respiration 
was rapidly inhibited by cyanide. 
However, cyanide sensitivity of light- 
stimulated oxygen uptake and the 
sensitive part of respiration were 
identical (Fig. 40). 

Discussion. Our results probably 
concern the same stimulation of oxy- 
gen uptake by blue light measured 
by Emerson and Lewis (1943), Ried 
(Year Book 6h, p. 399), and also 
Kowallik and Gaffron (1966). 

Our data indicate that the pro- 
longed increase in the rate of oxygen 
uptake observed after exposure to 












° -30 








Control X 

t i 

+DCMU >v 



10 10 

Time, min 

Fig. 39. Effect of DCMU on oxygen uptake. Oxygen uptake after a 5-sec flash of 0.62 X 1 0~ 9 
ein/cm 2 /sec of 469 m/j, light is not inhibited by 1 X 1 -5 M DCMU, which inhibited oxygen evo- 
lution 94%. Reservoir dark 72 hr. 














•+— ' 














-5 -4 

KCN concentration, log M 

Fig. 40. Effect of cyanide on light-stimulated oxygen uptake and respiration. Light-stimulated 
oxygen uptake of starved cells (open circles) after 5-sec flash of 1 .2 X 1 -9 ein/cm 2 /sec 
of 469 mjx. Respiration (open circles) after stimulation by 1 % glucose. 

small doses of blue light is activated 
by a flavin. The action spectrum (Fig. 
38A) is almost identical to that of 
Kowallik (1967a) for increased oxy- 
gen consumption by a chlorophyll-less 
mutant of Chlorella exposed to con- 
tinuous blue light. The effect does not 
appear to be coupled to system 2 of 
photosynthesis because the uptake 
is not inhibited by DCMU. Kowallik 
(1967&) found that light-stimulated 
oxygen uptake in Chlorella had a re- 
spiratory quotient of one. These re- 
sults are consistent with our data for 
cyanide sensitivity. 

Although the data suggest that oxy- 
gen consumption stimulated by blue 
light is probably mediated by mito- 
chondrial respiration, there is no evi- 
dence that the responsible flavin is 
directly involved in respiration. 

Summary. The action spectrum for 
the stimulation of oxygen uptake by 
blue light was determined for Chlo- 
rella pyrenoidosa. The action spec- 
trum is limited to wavelengths 
shorter than 550 nut and has peaks at 
460 and 375 m/x. The peaks are 
separated by a definite minimum at 
400 The action spectrum indi- 
cates that the responsible pigment is 
a flavin. 

The effect is not closely coupled to 
system 2 of photosynthesis because 

the uptake was not inhibited by 
10- 5 M DCMU. Light-stimulated oxy- 
gen uptake may be closely related 
to cyanide-sensitive dark respiration, 
since both exhibit the same cyanide 


Emerson, R., and C. M. Lewis, Am. J. 
Bot., 80, 165, 1943. 

Kowallik, W., and H. Gaffron, Planta, 69, 
92, 1966. 

Kowallik, W., in Energy Conversion by 
the Photo synthetic Apparatus, Brook- 
haven Symp. Biol., Brookhaven Na- 
tional Laboratory, Upton, N.Y., p. 467, 

Kowallik, W., Plant Physiol, 42, 672, 

Intermediates in Phytochrome 

Transformation in vivo and 

in vitro 

Winslow R. Briggs and David C. Fork 

The plant pigment phytochrome, a 
chromophoric protein, mediates a 
wide range of physiological processes 
in green plants from the algae 
through the higher angiosperms. It 
can exist in either of two spectrally 
distinct forms : a red-absorbing form 
Pr with maximum absorption be- 
tween 650 and 670 im*, and a far-red- 



absorbing form Pfr with maximum 
absorption between 715 and 735 m^. 
The two forms are readily intercon- 
vertible by red and far-red light. Both 
have absorption peaks in the long 
ultraviolet or blue, attributable to the 
ehromophoric group, a linear tetra- 
pyrrole. These latter peaks show spec- 
tral shifts comparable to those in the 
red and far red upon appropriate 

Previous work {Year Book 61*, pp. 
406-412) has shown that a mixture 
of high-intensity red and far-red 
light causes rapid cycling of the pig- 
ment from one spectral form to the 
other, producing measurable steady- 
state levels of spectrally detectable 
intermediates in partially purified 
phytochrome from dark-grown seed- 
lings of Arena sativa L. Actinic light, 
provided by a Sun-Gun lamp, was 
passed through a 5-cm water filter, 
and a Corning 2030 cutoff filter to 
remove wavelengths below 620 m^. 
Absorbancy changes were measured 
at 513 m/x by means of previously 
described instrumentation (Year 
Book 63, pp. 435-441). The inter- 
mediates were detected by their in- 
crease in absorbancy at this wave- 
length when the actinic beam was 
turned on, and their subsequent ab- 
sorbancy decrease when it was turned 
off. Appropriate filters between the 
sample and the phototube prevented 
actinic light from reaching the photo- 
tube, but allowed passage of the low- 
intensity 543-m/>t measuring beam. 
Except for Q 10 experiments, sample 
temperature was maintained at ap- 
proximately 5°C. 

A kinetic analysis of absorbancy 
decay after illumination suggested 
that two intermediates with different 
rate constants were decaying in par- 
allel to Pfr. Time course studies for 
the formation of the intermediates 
in the light showed that the more 
rapidly decaying one increased to its 
maximum concentration within 2 sec- 

onds and then decreased to a steady- 
state level approximately one half as 
high by the end of 12 seconds of illu- 
mination. The more slowly decaying 
intermediate appeared less rapidly, 
reaching its maximum level at about 
12 seconds and then remaining there. 
The decay constant r for one inter- 
mediate was about 0.2 and for the 
other about 1.1 seconds at 5°C. The 
Q 10 for dark decay (between 5°C and 
25 °C) of the intermediates was be- 
tween 1.9 and 2.3, although the over- 
all Q 10 for transformation of the 
pigment in either direction was ex- 
tremely close to 1.0. Thus, under the 
conditions of the experiment, decay 
of intermediates could not have been 
limiting the rate of light-induced pig- 
ment transformation. The two inter- 
mediates studied appeared to corre- 
spond to the two slowest-decaying 
intermediates on the Pr or Pfr path- 
way, as described by Linschitz, 
Kasche, Butler, and Siegelman (1966) 
from flash photolysis experiments. 

Phytochrome intermediates in vivo. 
It was of some interest to determine 
whether the intermediates studied 
with partially purified phytochrome 
could also be detected for native phy- 
tochrome in vivo, and if they could, 
whether or not their kinetics were 
similar to those in solution. For this 
purpose we used approximately 1.5 
grams of coleoptile tips between 2 and 
4 mm in length, which were excised 
from oat seedlings grown for 5 days 
in complete darkness. Care was taken 
to eliminate any fragments of the 
primary leaves, since these contain 
substantial amounts of protochloro- 
phyll, which would interfere with the 
desired spectral measurement. The 
coleoptile tips were packed into a 
cuvette approximately 2 cm in diam- 
eter, forming a layer of tissue about 2 
mm thick. The sample was spread 
carefully in the cuvette to eliminate 
any small light leaks. The cuvette was 
then placed over the photomultiplier 



tube and kept surrounded with ice 
during the measurements. Although 
it was not possible to monitor tissue 
temperature directly, experience with 
liquid samples showed that the aver- 
age temperature was probably be- 
tween 5°C and 8°C. 

Figure 41 shows records of absorb- 
ancy changes during and after three 
different durations of actinic illumina- 
tion. It is obvious that with the short- 
est period of illumination the decay 
rate is substantially higher than that 
found after the two longer periods. 
A plot of decay half-time against 
concentration of intermediate at the 
end of illumination shows a steep 
increase in decay half-time with in- 
creasing concentration of interme- 
diate. These observations are in close 
agreement with those previously ob- 
tained with partially purified phyto- 
chrome. In the experiment shown in 
Fig. 41, light intensity was kept 
constant, and concentration of inter- 

mediates was varied by varying expo- 
sure time. When the reciprocal exper- 
iment was done (exposure time kept 
constant and concentration of inter- 
mediates varied by varying light 
intensity) decay half-times remained 
constant and were independent of 
intermediate concentration. These re- 
sults are also in agreement with those 
previously obtained with the in-vitro 
system. Since the signals obtained 
with the tissue sample were smaller 
and substantially noisier than those 
obtained with phytochrome solutions, 
it was not possible to do any further 
kinetic analysis on the decay patterns 
found. However, since the kinetic 
data that were available closely 
matched those previously obtained 
with solutions, it seems reasonable to 
conclude that the same two inter- 
mediates were being observed in vivo 
as had previously been observed in 
vitro. Absorption spectra of this and 
similar preparations of oat coleoptile 















n vivo 

+ =3 sec 

'•'II )ll ll 

t =.75 sec 


f =.38 sec 



6 8 

Time, sec 

Fig. 41. Absorbancy changes at 543 m/A induced in intact oat coleoptile tips by high-intensity 
actinic light (620 to 800 m/x, 4 X 10 s erg cm" 2 sec" 1 ). 


tips showed almost no protochloro- across the sample is far lower than 
phyll (before initial illumination) or that in previous experiments with 
chlorophyll (after illumination) but phytochrome solutions. 
did have very high absorbancy at 667 A difference spectrum for inter- 
niu after far-red irradiation or at 735 mediates in vitro. Partially purified 
m,u, after red irradiation. Thus the phytochrome was obtained from 
predominant pigment absorbing at dark-grown oat seedlings as de- 
the longer wavelengths was clearly scribed earlier (Year Book 6U, pp. 
phytochrome. 406-412) for investigation of the 
One important difference between spectral properties of the intermedi- 
the behavior of the intermediates in ates between 365 and 580 m/x. The 
vivo and that in vitro should be source of actinic light was the same 
noted : The steady-state level of in- as above, although the intensity was 
termediates in vivo, per unit of meas- somewhat lower (2.9 X 10 5 ergs 
urable phytochrome in the sample, is cm -2 sec -1 ) . Preliminary experi- 
approximately five times as high as ments showed that high concentra- 
that in vitro, although the measured tions of sucrose or glycerol sub- 
light intensity at the sample surface stantially slowed intermediate decay, 
in the current experiments was less allowing formation of considerably 
than one half that used previously higher steady-state levels than ob- 
with phytochrome solutions. One tainable in buffer alone. Figure 42, 
could account for this difference in top two tracings, shows absorbance 
steady-state levels in two ways: changes at 560 m/x for identical 
either slower decay or more rapid amounts and concentrations of phy- 
formation in vivo. A comparison of tochrome diluted 50% either with 
intermediate behavior in the tissue buffer or with glycerol (7°C). At 
and in solution (cf. Fig. 41 and Year this wavelength glycerol almost 
Book 6J+, Fig. 31) reveals that both doubles the height of the signal. Thus 
explanations obtain. Decay in vivo is the difference spectrum was obtained 
somewhat slower, and rise time is both with glycerol, to obtain maxi- 
significantly shorter. Since rise time mum signal size, and without, to 
is independent of length of exposure determine whether or not glycerol 
time for illuminations longer than 1 spectrally altered the intermediates, 
second, the comparison for this param- The bottom two tracings of Fig. 42 
eter is easily made: steady state is show the absorbancy changes found 
reached in vivo in approximately 0.3 at 418 and 385 m/x in the presence of 
second, while requiring well over 1 glycerol. A sharp absorbance decrease 
second in vitro. Hence for phyto- was found at 418 m/x and increases at 
chrome in the intact tissue, forma- 385 and 560 m/x. The complete differ- 
tion of intermediates is facilitated ence spectra with and without 50% 
and decay is hindered; however, glycerol are shown in Fig. 43. Each 
speculation on possible mechanisms point represents the average height 
is premature. The differences in rise of a minimum of six signals, obtained 
time are actually more dramatic than in every case with a 3-second light 
appears on first inspection of the period followed by 6 seconds of dark- 
figures, since the total absorbancy of ness (the regime illustrated in Fig. 
the tissue sample is far higher than 42). Though the signals with glyc- 
that of the phytochrome solution, erol are significantly larger, the two 
Thus in the current experiments the curves are otherwise essentially the 
average light intensity integrated same, with minima near 418 m/x, max- 












30 45 


Fig. 42. Absorbancy changes at several wavelengths induced in phytochrome solutions with or 
without 50% glycerol by high-intensity actinic light (620 to 800 m/i, 2.9 X 1 5 erg cm -2 sec -1 ). 

ima near 380 ni/x, and isobestic points 
near 398 and 485 m/x. 

Several points should be made 
about these difference spectra. First, 
they represent the difference in ab- 
sorbancy between the intermediates 
and Pfr, to which both decay. Second, 
it is not possible to obtain values 
above 580 m/x, since the technique re- 
stricts one to measurements at wave- 
lengths well below the range of the 
actinic beam. Third, the spectra are 
composite, representing both the fast 
and slow intermediates previously de- 
scribed. Though it should be possible 
to separate these two components by 
detailed kinetic analysis of the rec- 
ords, such analysis has not yet been 
done. Fourth, these difference spectra 

are not identical with the difference 
spectrum between Pr and Pfr. The 
latter spectrum has a minimum at 
about 410 m/x, a maximum at about 
385 Hi/*, and isobestic points at 398 
and 495 m/x. Finally, 3 seconds of il- 
lumination were not sufficient to 
cause the maximum possible absorb- 
ancy changes in the glycerol prepara- 
tion. Experiments that allowed the 
change to go to completion gave much 
larger changes but the spectra were 
the same. 


Linschitz, H., V. Kasche, W. L. Butler, 
and H. W. Siegelman, /. Biol. Chem., 
2^,3395-3403 (1966). 




450 500 



Fig. 43. Difference in absorbancy between intermediates and Pfr from 365 to 580 m/x. The 
absorbancy changes were measured from records similar to those shown in Fig. 42, by means 

of the same actinic light. 


Growth Responses of Mimulus 

Races and F 1 Hybrids at the 

Stanford, Mather, and 

tlmberline transplant stations 

Malcolm A. Nobs and William M. Hiesey 

A segment of the long-range in- 
vestigations on Mimulus is concerned 
with the comparative performance of 
parental and F a hybrids of a wide 
array of crossing combinations at the 
three altitudinal transplant stations. 
These studies, begun in 1961 and com- 
pleted in 1966, will be briefly sum- 

The questions. Earlier inquiries on 
the genetic structure of ecological 
races of Potentilla (Carnegie Publica- 
tion 615) and Achillea (Year Book 51, 
pp. 122-124) dealt mainly with 
crosses between extreme altitudinal 
and latitudinal races. Studies on 

crosses between ecological races only 
slightly or moderately differentiated 
from each other in comparison with 
crosses between ecological extremes 
have been mostly lacking. Because of 
its biosystematic structure, the Mim- 
ulus cardinalis-M . lewisii complex is 
particularly well suited to investiga- 
tions aimed at filling this gap in our 
knowledge (Year Book 64, pp. 427- 
428). Questions which this work was 
designed to clarify include the follow- 

1. How do first-generation hybrids 
between closely related yet distin- 
guishable ecologic races of the same 
species but from distinct habitats 
compare with hybrids between highly 
diverse races in their capacity to sur- 
vive and grow at the altitudinal 
transplant stations at Stanford, 
Mather, and Timberline ? 


2. Is there evidence of hybrid vigor Timberline transplant stations. These 
(heterosis) in such crosses and, if so, three combinations are chosen to rep- 
to what degree is its expression modi- resent the range of patterns of re- 
fied in these contrasting environ- sponse found in a number of intra- 
ments? cardinalis racial crosses. 

3. Does maternal (i.e., cytoplas- 1. Los Trancos X Yosemite. The 
mic) inheritance play a part in the Los Trancos race of M. cardinalis 
expression of any of the first-genera- from the coast of central California 
tion hybrid combinations in any of has been previously described (Year 
these environments ? Book 60, pp. 381-384). It survives 

In practice, it is difficult to obtain only at the Stanford station situated 

objective answers to these questions near sea level. The individual parent 

because of the nature of ecological plant used in the crosses, 6546-5, is 

races. Our data are biased because, not a heavy producer of dry matter, 

from practical experimental consid- as indicated by the arrow at the lower 

erations, we are impelled to assemble left of Fig. 44. The individual of the 

and establish our experimental cul- Yosemite race used as a parent in this 

tures at Stanford. This mild coastal cross (6694-105), originally from 

climate of central California favors 880 m elevation in the Sierra Nevada, 

the growth and establishment of low- grows with considerable vigor at 

altitude as compared with high- Stanford and produces a higher dry 

altitude ecological races. We are weight yield, as shown by the arrow 

nevertheless able to overcome this at the lower right of the figure. A 

bias to a considerable degree through sample population of 10 F t individuals 

extended observations at the moun- resulting from this cross (which in- 

tain stations. eludes reciprocals) shows consistently 

Analysis of the available data from greater vigor than either parent when 

reciprocal crosses of all hybrid com- grown at Stanford, as reflected by the 

binations made in Mimulus, either dry weights shown by the bar dia- 

intraspecific or interspecific, has thus gram at the bottom of Fig. 44. 

far failed to yield any evidence of The same population sample of 

maternal or cytoplasmic inheritance cloned F t hybrids grown at Mather 

at any of the three altitudinal sta- at 1400 m elevation displays a slight 

tions. The answer to the third ques- enhancement of vigor as compared 

tion appears, therefore, to be clearly with the parents at this same station, 

negative. The following discussion but the expression of heterosis is so 

will accordingly be concerned only much smaller than at Stanford that 

with the first two questions, and is it is barely evident. At Timberline, 

based on a study of the pooled data where the F x hybrids barely survive, 

of reciprocally produced hybrids in they may nevertheless be said to dis- 

each combination mentioned. play a slight degree of heterosis as 

Intra-cardinalis hybrids. A useful compared with the parent clones, 

measure of overall growth in Mimulus both of which are nonsurvivors. 

is the total dry weight of the above- 2. Los Trancos X San Antonio 

ground parts that are produced by a Peak. When the same Los Trancos in- 

plant during the course of its normal dividual used in the preceding cross 

growing season. The graphs in Fig. is combined with a clone of M. cardi- 

44 show the productivity of three nalis from San Antonio Peak, origi- 

different combinations of intereco- nally from 2200 m elevation (clone 

typic Fi hybrids of Mimulus car- 7120-15), the F x hybrid population 

dinalis at the Stanford, Mather, and when grown at Stanford displays 



§ 2 

■S -5 


■£ o 

*- <D 















O jt 
O oo 
m — 



J o 


< 2 

<1> £ 

•♦— o 





1 - 



jc > 

*■ o 

~0 to 

c c 
o .2 

8 1 

u «" 

o , — , 

u E 

o o 

§ s 

i- CO 

o —' 


.** := 

C _Q 

■5 £ 

<o c 


^ o 


1 T 

J2 <l> 

i. o 

«/> -rr 
.2 E 
o o 

">. CO 

t= c 
-£> o 

<D O 

-- i- 
D O 

o c 

s ^ 

c ■£ 


— c 

D D 

'■E E 

> o 



■» D 

o -a 
"I s 

£ 13 
D C 
«/> o 

C "to 

.2 o 

1 - 

Q. * 

2 -Q 

-o o 

■£. ° 

£ 2 

o — 

I "g 

3 ° 

a) _g 

c a 

D O 

*■ a 

■£ -c 

■*■* *- 







O X 

D) c 
.E D 

£ a 

o "> 

■c E 




D) 2 

I/I C 

o o 
« '-c 
?■ ° 

D> D 
J= C 

a> «» 
.E J 

,E x 
= o 








H "> 





(D >- 
-C O 
I— c 









about the same degree of hybrid vigor tions. All races of M. lewisii are es- 

as the preceding cross, as shown by sentially nonsurvivors when grown in 

the bar diagram in Fig. 44 (second the Stanford garden and can only be 

graph from the bottom) . At Mather maintained successfully there in con- 

the same population of cloned Fx hy- trolled growth chambers. First-gen- 

brids grows somewhat more vigor- eration hybrids between contrasting 

ously than the Fi hybrids between altitudinal or latitudinal races of M. 

Los Trancos X Yosemite, but this is lewisii do show some improvement in 

scarcely significant. The great reduc- growth over their parents at Stanford 

tion in heterosis expressed at Mather but they are essentially nonsurvivors. 

as compared with Stanford is of At Mather the response is only 

the same order of magnitude as in slightly better than in the Stanford 

the preceding combination. At Tim- environment. There the parents are 

berline the influence of the San An- likewise essentially nonsurvivors and 

tonio Peak parent in enhancing the although occasionally F x plants may 

vigor of the hybrids as compared with produce well during a single season, 

the parents is somewhat more marked none have survived more than two 

than in the preceding cross and may seasons. At Timberline all races of M. 

be significant. lewisii survive and the F t hybrids 

3. Los Trancos X Baja. The in- show definite heterosis as compared 

fluence of the genome of the Baja race with the parents, 

of M. cardinalis originally from the When any race of M. lewisii is 

San Pedro Martir at 600 m elevation crossed with any race of M. cardinalis 

(clone 7119-16) when crossed with the vigor of the F t progeny at Stan- 

the same Los Trancos parent used in ford generally exceeds the perfor- 

the two preceding combinations is mance of either parent. Moreover, 

striking in the high degree of hetero- unlike interracial F x hybrids within 

sis shown by the resulting F a progeny M. cardinalis, heterosis in cardinalis- 

grown at Stanford. This is shown by lewisii combinations is clearly ex- 

the bar diagram in Fig. 44 (third pressed at all three altitudinal trans- 

from the bottom), which also por- plant stations. This is shown in Fig. 

trays the high degree of individual 45 in which the dry yield performance 

variability in this F x hybrid popula- at Stanford, Mather, and Timberline 

tion. The same Baja genome when in three combinations of cardinalis- 

combined with other races of M, car- lewisii are shown in comparison with 

dinalis consistently yields Fi hybrids the intr n-cardinalis cross Los Trancos 

displaying more hybrid vigor in the X Yosemite. 

Stanford garden than hybrids pro- 1. M. cardinalis Los Trancos X M. 

duced from any other interracial com- lewisii Yosemite. The coastal M. car- 

bination within this species. dinalis clone 6546-5 from Los Trancos 

At Mather, however, even the effect when crossed with M. leivisii, origi- 

of the Baja genome in enhancing het- nally from Yosemite National Park, 

erosis is barely expressed. This is at 880 m elevation (clone 7121-5), 

consistently shown also by all other yields a population of Fx progeny 

F x combinations between ecological which grows with moderate vigor at 

races of M. cardinalis. At Timberline Stanford, well exceeding the Los 

also there is essentially the same re- Trancos parent in overall yield, as 

sponse. shown in Fig. 45 (the second bar dia- 

Mimulus cardinalis X M. lewisii Ft gram from the bottom). The overall 

hybrids in comparison with intra- growth of this F x population at Stan- 

car dinalis and intra-lewisii combina- ford nevertheless falls short of that 



At Timberlme 

(Graphed at X10 scale) 

Los Trancos X Timber line 
M cardmolis X lewisii 

Los Trancos X Yosemite 
M. cardinal is X lewisii 

Los Trancos X Yosemite 
M. cardmolis X cordinalis 






At Mather 

(Los Trancos M. Cardi 

is is a nonsurvivor 

Yosemite X Semite 
M. cardmolis X lewisii 

Los Trancos X Timberline 
M cardmolis X lewisii 

Los Trancos X Yosemite 
M cardinal is X lewisii 

Los Trancos X Yosemite 
M. cordinalis X cardmolis 

At Stanford 

Yosemite and Timberline races of 

M. lewisii are nonsurvivors 

Yosemite X Yosemite 
M cardmolis X M lewisii 

Los Trancos X Timberline 
M car di nali s X lewisii 

Los Trancos X Yosemite 
M. cardmolis X lewisii 

Los Trancos X Yosemite 
M cardmolis X cardmolis 

t t 

Y Y 



♦ * 









(Los Trancos and Yosemite races of 
M. cordinalis are nonsurvivors) 

00 200 300 

Dry weight of tops, grams 


Fig. 45. Graphs of dry weight yields of parents and first-generation hybrids between different 
races of M. cordinalis and M. lewisii in comparison with the parents and with an \n\ra-cardinalis 
hybrid. See legend of Fig. 44 for explanation of the graphs. The arrows with asterisks indicate 
parental races of M. lewisii; those without asterisks, parental races of M. cordinalis. 


of the Fx population of M. cardinalis lewisii (clone 7121-5) from the same 
Los Trancos X M. cardinalis Yosem- area is of particular interest in study- 
ite, which is shown in the lowermost ing the effect of combining the ge- 
bar diagram of the same figure for nomes of these two distinct but closely 
comparison. related taxa originally growing in 
At Mather the heterosis of the two essentially the same natural habitats 
combinations is drastically reversed in comparison with crosses between 
as compared with Stanford: M. car- ecological races native to different 
dinalis Los Trancos X M. lewisii altitudes and latitudes. 
Yosemite is strikingly more vigorous The most striking feature of the 
than M. cardinalis Los Trancos X M. F x population from this cross is its 
cardinalis Yosemite. At Timberline marked similarity to the responses of 
the relative performance of the the F a hybrids between the two car- 
two Fi combinations is approximately dinalis X leivisii combinations pre- 
the same as at Mather although the viously discussed at the three trans- 
actual total dry weight yield at Tim- plant stations (Fig. 45). 
berline in both instances is reduced Conclusions. From the available 
by a factor of nearly 10 because of data, the following conclusions re- 
the much shorter growing season. garding the genetic structure of nat- 

2. M. cardinalis Los Trancos X M. ural populations of Mimulus cardi- 
lewisii Timberline. When the Los nalis and M. lewisii appear to be 
Trancos clone of M. cardinalis (6546- valid: 

5) is crossed with the high-Sierran 1. The genetic differences that have 

M. lewisii originally from Timberline evolved between M, cardinalis and M. 

at 3300 m elevation (7405-4), the re- leivisii are of greater magnitude and 

suiting F x progeny show about the evolutionary significance than the 

same degree of vigor as the progeny genetic differences between ecological 

of the cross M. cardinalis Los Trancos races within either species. 

X M. leivisii Yosemite. The parallel- 2. The genetic differences between 

ism in the performance of these two ecological races of either species are, 

Fi populations when grown at the nevertheless, of sufficient magnitude 

Stanford, Mather, and Timberline to be of ecological significance in nat- 

transplant stations is striking (Fig. ural selection. 

45) . This result fails to support an 3. Hybrid vigor, or heterosis, is the 

earlier hypothesis that the F 1 between general rule in crosses within the 

M. cardinalis Los Trancos X M. lew- Mimulus cardinalis-M . lewisii com- 

isii Timberline would be more sue- plex in both intraspecific and inter- 

cessf ul at Timberline than the F x be- specific crosses. The expression of het- 

tween M. cardinalis Los Trancos X erosis in a given Fi population may 

M. lewisii Yosemite. The effect of be profoundly influenced by climate 

combining the genomes of M. cardi- as, for example, in mtYSL-cardinalishy- 

nalis and M. lewisii clearly over- brids grown at Stanford as compared 

shadows genetic effects due to recom- with Mather, or in intra- lew is ii hy- 

binations of different races of M. lew- brids grown at Timberline as com- 

isii on a common parent of M. cardi- pared with Mather. In other instances 

nalis. heterosis may be almost equally ex- 

3. M. cardinalis Yosemite X M. pressed over the entire range of 
lewisii Yosemite. The combination climates at Stanford, Mather, and 
between M. cardinalis originally from Timberline, as in combinations be- 
Yosemite National Park at 880 m tween M. cardinalis and M. leivisii. In 
elevation (clone 6694-105) and M. such combinations heterosis is espe- 


dally evident at Mather where most at approximately 100,000 ergs/cm 2 / 

of the parental races of both M. car- sec, or about one third the intensity 

dinaiis and M. lewisii fail to be sue- of full sunlight. Leaves of the same 

cessful. individual plants grown under a weak 

4. There is no evidence to support light of 25,000 ergs/cm 2 /sec intensity 

the hypothesis that the survival of F x have a considerably lower light-satu- 

hybrids at contrasting altitudes is re- rated rate, and the differences in 

lated in a simple way to the altitudi- light-saturated photosynthetic rates 

nal origin of the particular ecological of ecological races and hybrids tend 

races that enter into the crosses. to disappear. The ratio between the 

light-saturated photosynthetic rate 

PHOTOSYNTHETIC KATES OF ° f * ^ ?}°™ . ^° WX [ "nder high 

MhnulHS RACES AND HYBRID ™ raU ? loW ^ ht llltenSlt y C&n *?? 

■p. fore be used as one measure of its 


William M. Hicscij, Malcolm A. Nobs, Tab]e 13 showg light-saturated 

and OUe Bjorkman photosynthetic rates of M. cardinalis, 

The higher rates of light-saturated subalpine M. leivisii, their reciprocal 

photosynthesis of coastal Mimulus F x hybrids, and two F 2 individuals 

cardinal is as compared with subalpine that differ widely from each other in 

M, lewisii described last year (Year their morphology and in their growth 

Book 65, pp. 461-468) have been con- responses at the Stanford, Mather, 

firmed by means of different precon- and Timberline transplant stations. 

ditioning treatments. All of these clones were previously 

As previously emphasized, such grown under the same conditions, 

differences in photosynthetic capacity under high light intensity in con- 

of contrasting races of Mimulus and trolled cabinets, before measurement. 

hybrids are revealed only when the The F 2 clone 7135-35 resembles M. 

experimental plants are previously car dinaiis both morphologically and 

conditioned by growing them under a in its survival, and in interstation 

high light intensity as, for example, transplant responses. It is a vigorous 

TABLE 13. Photosynthetic Rate and Chlorophyll Content of Mimulus Parents and Hybrids 

Gross Light- Ratio, P Rate chlorophyll Content of Leaves 

Saturated Photo- of Leaves Grown 

synthetic Rate under High and mg/dm 2 mg/g fresh 

mg C02/dm 2 /hr* Low Light Intensity surface weight 

Mimulus cardmalis, Los Trancos, 

Mimulus lewisii, Timberline, 


Mimulus cardinalis X lewisii, 
Ft Hybrid, 6546-3 

Mimulus lewisii X cardinalis, 
Ft Hybrid, 65A7A 

Mimulus cardinalis X lewisii, 

F2, caraVnal/s-like, 71 35-35 
Mimulus cardinalis X lewisii, 

F 2 , lewisii-Uke, 71 1 1 -1 6 

























* Plants previously grown under high light intensity. All values are means of several indepen- 
dent measurements. 


survivor at Stanford but dies at the different leaf-surface to leaf -volume 

Mather and Timberline stations. In ratios. 

contrast, the F 2 clone 7111-16 The major points of interest shown 
strongly resembles the M. lewisii by the data in Table 13 may be sum- 
parent. The resemblance is so close marized: 

that it would be classified as a mem- 1. Coastal M. cardinalis has a sub- 

ber of this species, and is a nonsur- stantially higher light-saturated 

vivor at Stanford and Mather. photosynthetic rate than does subal- 

The two F 2 individuals yield segre- pine M. lewisii. Likewise, the ratio 

gating F 3 progenies that differ widely between light-saturated rates of 

from each other in overall genetic photosynthesis of propagules grown 

composition. The growth responses at high light intensities to those 

and survival of samples of these F 3 grown at low light intensities is 

progenies are currently being tested much higher in the coastal clone. 

at the altitudinal transplant stations. 2. The two reciprocal F t hybrids 

The ratio between the light-satu- have a light-saturated photosynthetic 

rated photosynthetic rates of propa- rate that is intermediate between the 

gules of these clones grown under parents, as do also the two very unlike 

high light intensity to those grown F 2 individuals. The ratio between 

under low light intensity is given in high and low light-grown propagules 

the second column of figures from the differs markedly between the F 1 and 

left in Table 13. These ratios differ, the two F 2 individuals, being as high 

especially in that M. cardinalis has a as in the M. cardinalis parent in both 

higher ratio than M. lewisii. The F 2 plants. 

ratios of the F 1 hybrids are inter- 3. Chlorophyll content is markedly 
mediate between the parents, but both higher in the coastal M. cardinalis 
of the contrasting F 2 individuals have parent than in M. lewisii, and inter- 
ratios as high as the M. cardinalis mediate in both reciprocal F x individ- 
parent. uals. In the cardinalis-like F 2 chloro- 

The bulk chlorophyll content of the phyll content is essentially the same 
leaves also differs greatly between the as in the M. cardinalis parent, and in 
two parent clones and can be re- the leivisii-like F 2 it is nearly as low 
garded as another segregating char- as in the M. lewisii parent, 
acteristic in the hybridization studies These data suggest that differences 
independent of the light-saturated in photosynthetic characteristics in 
photosynthetic rate. The chlorophyll Mimalus are inherited in much the 
content affects the capacity of the same manner as morphological mark- 
leaves to absorb light and therefore ers {Year Book 63, pp. 432-435), and 
influences the rate at low intensities that one may expect the principle of 
of incident light, but not at saturating genetic coherence (Year Book 62, pp. 
light intensities {Year Book 65, pp. 387-389) to apply to the inheritance 
461-468). The chlorophyll content of of factors controlling various steps of 
the leaves is shown in the two col- the photosynthetic processes that 
umns at the right of Table 13 and is determine observed measured rates 
expressed on both a leaf -area and a under specified controlled conditions, 
fresh-weight basis. The differences Studies such as those described 
in relative values between these two above and biochemical investigations 
methods of computing chlorophyll comparable with those reported below 
content are due primarily to heredi- for Solidago are being extended to 
tary differences in leaf thickness be- include M. nelsonii from Mexico, F x 
tween the clones, which result in hybrids between nelsonii and subal- 


pine M. lewisii, and the tetraploid conditions were identical and as de- 
amphiploid derived from these spe- scribed earlier (Year Book 64, p. 
cies as described in last year's report. 420). In all measurements of photo- 
Data from these studies are still too synthesis, single rosette leaves, at- 
incomplete to be reported. tached to the intact plants, were used. 

Recently matured leaves (about 80% 

of full expansion) were used both for 

CARBOXYDISMUTASE Activity in Sun photosynthesis measurements and 

and Shade Ecotypes of Solidago enzyme determinations, except where 


OUc Workman and Monika Bjorkman ■ rtl , ' , 7 „. , 

' J Photosynthesis measurements. 

Preliminary studies, reported last Light-saturated rates of photosyn- 
year. indicated that ecotypes of Soil- thetic C0 2 uptake were measured 
dago uirgaurea originally from open with the apparatus described in Year 
as contrasted with shaded habitats Book 63 (pp. 430-431) modified to 
may differ in their capacity to pro- meet the requirements of the present 
duce the photosynthetic enzyme car- measurements. Light intensities, pro- 
boxydismutase (ribulose diphosphate viding up to 1.5 X 10 -7 absorbed ein- 
carhoxylase) . Of the two Solidago steins cm -2 sec -1 at 665 imi, half- 
clones investigated, the clone from an bandwidth 35 m/x, were obtained from 
open habitat showed a higher activity a quartz-iodine lamp (DWY) , appro- 
of the enzyme than the one from a priate lenses, and water and inter- 
shaded habitat, in extracts prepared ference filters. To minimize undesir- 
from the leaves when grown under a able gradients in C0 2 concentration 
high light intensity. Unrelated spe- in the leaf chamber its volume was 
cies limited in natural distribution to kept small (15 ml), and high flow 
open locations, such as Plantago Ian- rates (0.5 to 1.0 1/min) were used. 
ceolata, showed a higher activity of The average C0 2 concentration of the 
carboxydismutase than species lim- air in the leaf chamber was 0.030%. 
ited strictly to shaded habitats, such Enzyme determinations. Crude en- 
as Lamium galeobdolon. zyme extracts were prepared from 

This year we have studied this each separate leaf by homogenizing 
problem in greater detail. An im- the pre-weighed leaf sample (approx- 
proved assay procedure for enzyme imately 200 mg fresh tissue) in 10 ml 
determinations has been developed, of a mixture of 0.04 M tris-HCl, 0.01 
additional Solidago clones have been M MgCl 2 , 0.25 mM EDTA, and re- 
studied, and photosynthesis measure- duced glutathione (GSH). The final 
ments and enzyme determinations pK was 7.8 at 23 °C. Addition of GSH 
have been made on the same leaves. to the mixture used for homogeniza- 

Plant materials. The recent meas- tion was essential for high and repro- 

urements were made with cloned ducible enzyme activities. Although 

individuals of two populations of two in most cases 1.0 mM GSH was suffi- 

races of Solidago virgaurea, "Bes- cient for maximum activity, higher 

kades," native to an exposed alpine- concentrations were required with 

arctic heath in northern Norway, and some samples. A concentration of 5.0 

"Hallands Vadero," native to a dense mM GSH proved to be sufficient for 

oak forest in southern Sweden. Sev- all clones and was used throughout 

era] clones of each race were grown the investigation. The supernatant, 

in controlled cabinets at two different obtained after spinning the homoge- 

light intensities, 25,000 and 110,000 nate at 30,000 g for 20 minutes, was 

erg cm- 2 sec -1 (400-700 m/x). Other used for the enzyme assays without 



further purification. All preparative 
procedures were carried out at 0° to 

The enzyme assays were consist- 
ently made 40 to 60 minutes after the 
start of the homogenization. As a 
standard procedure, the reaction was 
started by the addition of 0.1 ml of 
the enzyme extract to 0.4 ml of a 
freshly prepared mixture of 2.5 /miole 
NaHC 14 3 (0.5 /xcurie//miole) , 0.14 
/miole ribulose-l,5-diphosphate, 1.25 
/miole GSH, 0.1 /.mole EDTA, 10 
/miole tris-HCl, and 2.5 /miole Mgd 2 . 
The pB. was 7.8 at 23 °C. The sodium 
salt of ribulose-l,5-diphosphate was 
obtained by conversion of the barium 
salt. This was either purchased from 
Sigma or was kindly supplied by 
Dr. Daniel McMahon, University of 
Chicago. The latter preparation 
yielded somewhat higher activities 
with given enzyme preparations in 
comparison with the former and was 
used as a standard throughout the 

After two minutes' incubation at 
23 °C the reaction was stopped by the 
addition of 0.1 ml 6 I acetic acid. 
A 0.25-ml aliquot was pipetted into a 
liquid scintillation counting vial and 
dried at 90° C for 2 hours. Then 0.25 
ml of water was added to each vial, 
followed by 10 ml of Bray's solution. 
The radioactivity measurements were 
made with a Packard Tri-Carb 
Liquid Scintillation Spectrometer. 

The rate of C 14 incorporation was 
entirely linear with enzyme concen- 
tration in the range used, but declined 
somewhat with increasing incubation 
time. No change in the activity of the 
enzyme preparation was observed 
during a period of 1 hour if the en- 
zyme was kept at 0° to 2°C. At least 
four replicate assays were run for 
each preparation, the typical varia- 
tion among replicates being about 
5%. The rate of C 14 incorporation in 
the absence of ribulose diphosphate in 
all cases was less than 1 % of the rate 
in the presence of this substrate. 

Results. The carboxydismutase ac- 
tivities in preparations from selected 
clones of the Beskades population 
from an open habitat, and of Hal- 
lands Vadero from a densely shaded 
habitat are compared in Table 14. 
The three clones of each population 
were selected to include the extremes 
of variation with regard to leaf anat- 
omy as well as to their photosynthetic 
characteristics as found in previous 
extensive work. 

It is evident that the Beskades 
plants grown at a high light intensity 
show the highest enzyme activities 
regardless of the basis used for ex- 
pressing them. The enzyme activities 
of the Hallands Vadero clones are 
relatively low when preconditioned to 
either high or low light intensities. 
Two of the Beskades clones (033 and 
039) seem to require a high light in- 

TABLE 14. Carboxydismutase Activity in Contrasting Races of Solidago virgaurea from 

Sunny and Shaded Habitats* 

Clone /xmole C02/min/g Tissue /xmole C02/min/mg Protein 

Light intensity for growth 










Beskades (sun race) 
















Hallands Vadero (shade race) 











* Each value represents the mean of independent experiments with two or more different plants 
of the same clone. 



tensity to produce maximum enzyme 
activity, whereas in the remaining 
clone (076) the activity is high 
whether the clone has been previously 
grown in strong light or weak light. 

The specific activity of the enzyme, 
that is. the activity on the basis of 
soluble protein, shows considerably 
less variation among individual 
clones and the differences between 
the two races become more distinct 
than when the activity is expressed on 
the basis of fresh weight. When 
grown at high light intensity the 
Beskades clones have nearly twice the 
specific activity of the Hallands 
Vadero clones. This is not associated 
with a lower level of soluble protein 
in the Beskades clones; the total as 
well as the soluble fraction of the leaf 
protein is generally higher rather 
than lower in the Beskades as com- 
pared with the Hallands Vadero 

It is interesting to note that even 
though the assays were carried out at 
bicarbonate concentrations several 
times below saturation, and also at a 
suboptimal temperature and ribulose- 
diphosphate concentration, the spe- 
cific activities of the Beskades plants 
are among the highest reported in the 
literature for any plant species. It 
should be pointed out, however, that 
if compared under physiological C0 2 
concentrations the light-saturated 
rate of C0 2 fixation by the intact leaf 
is much higher than the rate of C0 2 
fixation catalyzed by the enzyme ex- 
tract of the same leaf. 

We investigated the possibility that 
the apparently lower specific activ- 
ities of Hallands Vadero in compar- 
ison with Beskades leaves might be 
caused by the presence of an inhibitor 
in Hallands Vadero not present in 
Beskades, or conversely, to an activa- 
tor present in Beskades but not in 
Hallands Vadero. For this purpose 
homogenates of leaves from the two 
populations were mixed, and the re- 

sultant activity compared with the 
activities of the separate extracts. 
The results presented in Table 15 
show that the activity of the com- 
bined homogenates is very close to 
the calculated mean of the activities 
of the separate extracts. This is evi- 
dence for the absence of any such 
inhibitor or activator. Also, the pres- 
ence of a competitive inhibitor would 
be expected to shift the apparent 
value of the K m . No differences of K m 
have been found among extracts from 
the different clones. 

As has been reported earlier 
(Bjorkman and Holmgren, 1963), 
pigment bleaching may result when 
shade habitat plants of Solidago are 
grown under light of high intensities. 
This effect increases with leaf age 
and, at advanced stages of bleaching, 
deterioration of the chloroplasts is 
often apparent. Precautions were 
therefore taken to avoid complica- 
tions that might result from such sec- 
ondary effects. The light intensity 
was kept at a somewhat lower level 
than in previous investigations, and 
only recently matured leaves showing 
no visible evidence of chlorophyll 
bleaching or chloroplast disturbances 
were used for enzyme determinations. 

To test to what extent a low car- 
boxydismutase activity may be the 
result of light damage of the chloro- 
plasts, enzyme activity was deter- 
mined in extracts of leaves in which 

TABLE 1 5. Test for Presence of Inhibitors and 

Activators of Carboxyclisrnutase in Extracts of 

Leaves Grown in Strong Light 




Hallands Vadero 132 


± 56* 


Beskades 076 


± 140 


Mean of (1 ) and (2) 



Hallands Vadero 132 + 
Beskades 076 1 :1 


± 175 

* Mean and standard deviation, four inde- 
pendent determinations. 


increasing degrees of bleaching had an attempt to gain some information 

been induced. As shown in Table 16, on this problem we determined the 

the specific enzyme activity remains relationship between enzyme activity 

the same even to a stage where the and photosynthesis in Solidago 

chlorophyll content is less than one clones. Figure 46 shows the results 

third of the normal, and severe struc- obtained when the rate of light- 

tural disturbances of the chloroplasts saturated photosynthesis of leaves 

have become apparent. from a number of different Solidago 

It thus seems highly probable that clones of both races is plotted against 
the lower enzyme activity found in the specific activity of the enzyme in 
the Hallands Vadero as compared extracts prepared from the same in- 
with the Beskades plants when grown dividual leaves. The two variables 
at high light intensities is caused by a show a very high correlation. The 
genetically determined lower capacity results of a detailed study of two 
of Hallands Vadero to produce the clones of each population when grown 
enzyme. It is tempting to speculate under high and low light intensities 
that these differences have arisen as a (Table 17) further support the con- 
result of an adaptation of the photo- elusion that there is a strong relation- 
synthetic machinery to the light en- ship between the light-saturated rate 
vironment of the plants in their of photosynthesis and the specific 
native habitats. Under the low light activity of the enzyme. Both show 
intensities prevailing on the floor of a a parallel variation regardless of 
dense forest, efficient utilization of whether the differences are due to 
light for photosynthesis would re- genetically distinct clones or to pre- 
quire a high capacity of those steps conditioning by light, 
concerned with the absorption and If the assumption is valid that the 
utilization of the light in primary enzyme concentrations in vivo are 
photoacts, whereas the capacity of proportional to the enzyme activities 
enzymic steps could be relatively low. measured in vitro, these findings 
With abundant light, on the other strongly suggest that the failure of 
hand, the capacity of nonphotochem- the shade clones to adjust to the effi- 
ical steps must be high for efficient cient use of strong light for photo- 
photosynthesis, synthesis may be caused, at least in 

However, the extent, if any, to part, by a limited capacity to produce 

which the comparatively low level of the carboxylation enzyme, 

carboxydismutase of the Hallands Whether Solidago clones from 

Vadero leaves actually does limit the shaded and exposed habitats differ in 

light-saturated rate of photosynthesis their capacity to produce photosyn- 

under conditions of normal — and thus thetic enzymes other than carboxy- 

limiting — carbon dioxide concentra- dismutase is not known. In consider- 

tions is a question yet unanswered. In ing the question of what internal 



Carboxydismutase Activity in Relation to Chlorophyll Bleaching by Exposure to 
Strong Light. Clone, Hallands Vadero 132 

Leaf No. 

Total Chlorophyll, Activity, 
mg/mg protein fimole C02/min/mg protein 


0.17 0.14 
0.12 0.17 
0.05 0.17 




c 0.2 












F 0.1 






0.1 0.2 0.3 

Specific activity, pmol C0 2 /min/mg protein 

Fig. 46. Rate of light-saturated photosynthe- 
sis in vivo in relation to the activity of carboxy- 
dismutase in extracts from the same leaves. 

factors may limit the rate of light- 
saturated photosynthesis, it should 
also be kept in mind that with nor- 
mal CO j concentrations these factors 
include not only the capacity of en- 
zymic steps but also the barriers to 
the diffusion of C0 2 . Of great interest 
in this connection are the findings by 
Dr. Holmgren in Uppsala, Sweden, 
that shaded habitat clones of Solidago 
when grown in strong light show a 
greater stomatal resistance to diffu- 
sion than do exposed habitat clones. 
Perhaps a simple explanation for 
these differences can be found in the 
physical structure of the leaves. How- 
ever, the stomatal aperture is under 
metabolic control and may very well 

be strongly influenced by the rate at 
which biochemical steps of photosyn- 
thesis operates. The possibility can- 
not be ruled out, therefore, that the 
higher stomatal diffusion resistance 
found in the shaded habitat clones is 
associated with a lower capacity of 
one or several enzymic steps in 

Further Studies of the Effect 

of Oxygen Concentration on 

Photosynthetic C0 2 Uptake in 

Higher Plants 

Olle Bjorkman 

Last year we reported that under 
atmospheric C0 2 concentrations the 
rate of photosynthetic C0 2 uptake 
was about 30% inhibited by the oxy- 
gen in the air in a number of species 
of higher plants. The degree of inhi- 
bition was remarkably constant over 
a wide range of light intensities and 
among various species differing 
widely in their maximum photosyn- 
thetic rate. However, two species of 
green algae tested showed no inhibit- 
ing effect of 21 % oxygen when photo- 
synthesis was measured under con- 
ditions similar to those used for the 
higher plants. Although no conclusive 
evidence was obtained with regard 
to the mechanism of the inhibitory 

TABLE 17. Light-S 

aturated Rate of 

Photosynthesis and Carboxydismutase 
Sun Clones of Solidago 

Activity in Shade 

Grown in 

Strong Light 

Grown in 

Weak Light 

of Clone 




Activity, fxmole 


Activity, [imole 

Beskades 1 jJJ* 

1 076 

Hollands J 132 
Vcidero 124 









' All values represent means obtained from several ramets of each clone. Photosynthesis and 
enzyme determinations were carried out on comparable but not necessarily the same leaves. 


effect, some of our results suggested effective inhibitors of the enzyme 

that the inhibition might be caused glycolic acid oxidase. Recently he re- 

by a leakage of electrons from highly ported (Zelitch, 1966) that the addi- 

reduced photosynthetic intermediates tion of this inhibitor to leaf discs of 

back to molecular oxygen, possibly at tobacco stimulated the rate of net 

a site between the two photosystems. C0 2 uptake in air, at least temporar- 

That an enhancement of the rate of ily. Moreover, if the leaves were fed 

the normal respiratory process by an with C 14 -labeled glycolate, the rate of 

increased oxygen concentration C0 2 evolution in light and C0 2 -free 

would be responsible for the apparent air increased, as did the radioactivity 

inhibition of C0 2 uptake was not con- of the C0 2 evolved. Thus, there are 

sidered as a likely explanation, since several lines of evidence that strongly 

it is well established that respiration indicate a close similarity between 

is saturated at a low partial pressure the effects on C0 2 exchange obtained 

of 2 . when the oxidation of glycolate is in- 

As part of their extensive work hibited, and those obtained when oxy- 

directed toward uncovering the rela- gen concentration is reduced. Also, it 

tionship between photosynthesis and has been shown by Hess and Tolbert 

respiration in higher plants, Krotkov (1967) that the green alga Chlorella, 

and co-workers investigated the de- in which the C0 2 exchange was found 

pendence of the C0 2 compensation not to be affected by 21% oxygen, 

point on oxygen concentration. They lacks glycolate oxidase. Some of our 

found that this point (i.e., the C0 2 studies this year have therefore been 

concentration at which there is no concerned with the question of 

net uptake or release of C0 2 ) was whether a stimulation of glycolate 

close to zero at very low oxygen con- oxidation is, in fact, the primary 

centrations and that it increased cause of the inhibitory effect of oxy- 

linearly with concentration up to gen on photosynthetic C0 2 uptake in 

100% 2 (Tregunna, Krotkov, and higher plants. 

Nelson, 1966; Forrester, Krotkov, Quantitative measurements of C0 2 

and Nelson, 1966 a, h) . One exception exchange and glycolic acid content 

to this rule was corn (Zea mays) in under air and low-oxygen concentra- 

which the C0 2 compensation point is tion. If the increase in the rate of 

zero even at an 2 concentration of C0 2 uptake, which takes place when 

normal air. From these and other oxygen concentration is reduced, is 

data they concluded that the main ef- caused by inhibition of the rate at 

feet of increasing oxygen on C0 2 ex- which glycolate is oxidized to glyoxy- 

change is a stimulation of the rate of late, glycolate should accumulate in 

C0 2 evolution in light. They also con- the same way as it does when the 

eluded that the process underlying enzyme mediating this step is inhib- 

this evolution of C0 2 in light operates ited. If the rate of glycolate formation 

through a pathway different from is not strongly affected by oxygen 

that of the normal respiratory pro- concentration, the rate of glycolate 

cess. Fock and Egle (1966), working accumulation would be expected to be 

with bean plants and the liverwort proportional to the difference between 

Conocephalum, arrived at similar the rates of C0 2 uptake in low oxygen 

conclusions. and in air. 

Several years ago Zelitch showed To test whether such an accumula- 

that glycolic acid accumulates in the tion actually takes place, the follow- 

leaves of higher plants in the pres- ing experiment was carried out : The 

ence of hydroxysulfonates, which are rate of steady-state C0 2 uptake in 



air at 22 °C was determined on a de- 
tached leaf of Solidago mtUtiradiata 
whose petiole was placed in a small 
volume of water, and the effect of 
changing the oxygen concentration 
between 21 % and 1.5% was meas- 
ured repeatedly. The leaf was then 
kept at 1.5% 0„ and the rate of C0 2 
uptake recorded continuously for 1 
hour. After this time the leaf was 
killed by rapid immersion (<2 sec) 
in boiling 1% NaHS0 3 solution. The 
water in which the petiole was in- 
serted during the experiment was 
added to the solution, which was then 
kept at 95 °C for 5 minutes. The leaf 
was homogenized, and the superna- 
tant obtained after centrifugation 
was assayed for glycolate, by the 
technique described by Zelitch(1958). 
The controls were matched leaves 
kept in air during the experiments. 
The results in Table 18 show that no 
increase in glycolate content took 
place when the leaf was kept under 
low 2 concentration, even though the 
rate of C0 2 uptake was 41% higher 
than in air. 

Experiments in which Mimulus 
plants were grown under 4% and 
21% oxygen concentration over a 10- 
day period (this report, p. 232) like- 
wise yielded negative results. No sig- 
nificant difference in glycolate content 
of the leaves was found between 
plants grown under 4% and 21% 2 . 

Effect of temperature on the inhi- 

bition of CO z uptake by oxygen. With 
saturating light and normal CO- pres- 
sure most higher plants have a flat 
temperature optimum, usually be- 
tween 15° and 30 °C. An attractive 
hypothesis to explain why the rate of 
C0 2 uptake declines at a relatively 
low temperature is that the rate of 
CO- evolution in the light increases 
more strongly with increasing tem- 
perature than does photosynthesis. 
Zelitch's findings that the stimulation 
of C0 2 uptake by the addition of a 
glycolate oxidase inhibitor increases 
strongly with temperature appears to 
be in good agreement with this ex- 
planation. In Zelitch's experiment 
with tobacco leaf discs, C0 2 uptake 
was stimulated several fold by the in- 
hibitor at 35° C but was not affected 
at 25°C. 

If such a strongly temperature-de- 
pendent photorespiratory process un- 
derlies the observed inhibitory effect 
of oxygen on C0 2 uptake, this effect 
should be small or nonexistent at a 
low temperature (in tobacco leaves at 
25° C) and should increase strongly 
with temperature. The data shown 
for Solidago and Mimulus in Fig. 47 
indicate that this is not the case as 
far as the steady-state rate of C0 2 
uptake is concerned. The degree of in- 
hibition caused by 21% 2 is almost 
independent of temperature. 

The results thus obtained both in 
the experiments on glycolate accumu- 

TABLE 18. Glycolate Content in Detached Leaves of Solidago multiradiata, T. L. 7621-2, 
in Relation to CO2 Uptake in Low O2 and in Air 


(1) C0 2 fixed during 1 hour in 1.5% 2 

(2) C0 2 fixed during 1 hour in 21 % 2 

(3) Increase in CO2 fixation due to low O2 

concentration (1) minus (2) 

(4) Glycolate content expected if 1 mole CO2 is 

formed for each mole of glycolate oxidized 

(5) Glycolate content found in 1.5% O2 

(6) Glycolate content found in 21 % O2 (controls) 



> 14 

< 1 

< 1 



pj'aBuDipxa Z Q^) JO 3-p^j 


"* °? 5 

,D> io CO 
U. CO O 



lation in low 2 concentration and in 
those on temperature dependence of 
CO_> uptake fail to support the suppo- 
sition that the primary cause of the 
observed inhibition of C0 2 uptake by 
oxygen in normal air is stimulation 
of a "photorespiratory process'' medi- 
ated by the enzyme glycolate oxidase. 

It is noteworthy, however, that the 
time course of the change in C0 2 
uptake that results when C0 2 concen- 
tration is altered is markedly affected 
by temperature. The curves shown in 
Fig. 48 typify the time course ob- 
tained with Mimulus and Solidago 
leaves, but it should be pointed out 
that the magnitude of the transient 
changes varies considerably among 
different leaves, and also varies with 
the detailed experimental conditions. 

Effect of wavelength on the inhibi- 
tory effect of C0 2 uptake by oxygen. 
The steady-state rate of C0 2 uptake 
in normal air seems to be a remark- 
ably constant fraction of the rate in 
low 2 . The degree of inhibition 
caused by 21% 2 was found to be 
about the same among diverse species 
of higher plants even though their 
light-saturated rate of C0 2 uptake 
varied greatly. Furthermore, the de- 
gree of the inhibition is constant 
over a wide range of light intensities 
and, as already mentioned, also of 
temperatures. The results shown in 
Tables 19 and 20 indicate that the in- 
hibition is also about the same at 
different wavelengths of the light 
over the range investigated. Thus 
there is no indication that the ob- 
served inhibition of C0 2 uptake, at 
least under rate-limiting light inten- 
sities, is due to a photorespiratory 
process sensitized by pigments other 
than those operating in photosynthe- 
sis. Whether or not the spectral dis- 
tribution of the light influences the 
degree of inhibition at or beyond sat- 
urating light intensities is not yet 

Absence of an inhibition of C0 2 

uptake by 21% oxygen in certain 
higher plant species. The thermo- 
philic grasses corn and sugar cane, 
which exhibit unusually high light- 
saturated rates of photosynthesis, are 
capable of reducing the C0 2 content 
in a closed system to zero. The results 
of Forrester et al. (1966b) also indi- 
cated that the rate of C0 2 uptake in 
corn was little affected when 2 con- 
centration was changed from 1% to 
21% 2 . The absence in corn of an 
inhibitory effect of 2 in this range 
was confirmed in this laboratory. No 
significant inhibition was found at 
temperatures ranging from 10° to 
40 °C, light intensities from 2 X 10 3 
to 1.2 X 10 6 erg cm 2 sec -1 , and wave- 
lengths from 430 to 700 niju. Simi- 
larly, the dicotyledonous species 
Amaranthus edulis, which, like corn 
and sugar cane, is capable of high 
rates of C0 2 fixation and has a C0 2 
compensation point close to zero in air 
(El-Sharkawy et al., 1967), was 
found to lack an apparent inhibition 
of C0 2 uptake by 21% 2 . 

Figure 49 shows the different re- 
sponse pattern in the rate of C0 2 ex- 
change to changes between light and 
dark in the three species Mimulus 
cardinalis, corn (Zea mays), and 
Amaranthus edulis. Mimulus was 
chosen to represent species whose C0 2 
uptake is inhibited by 21% 2 . 

Particularly noteworthy are the 
differences among the species with re- 
gard to the presence of a marked 
transient high rate of C0 2 evolution 
during the first minute after the light 
is turned off. The presence of such a 
post-illumination burst of C0 2 has 
been taken as evidence for a stimula- 
tion of the rate of C0 2 production in 
the light (e.g., Tregunna et al., 1964). 
The fact that the extent of the burst 
decreases with decreasing 2 concen- 
tration (Fock and Egle, 1966, and 
Tregunna et al., 1966) and that it is 
absent in corn, even in the presence 
of air, at first sight seems to be con- 



sistent with the hypothesis that there 
is a close connection between the C0 2 
burst and the inhibitory effect of 2 
on net C0 2 uptake. It is known, how- 
ever, that glycolysis is inhibited in the 

light. Darkening relieves this inhibi- 
tion and may lead to an overshoot of 
oxidation, which may, in part, explain 
the observed C0 2 burst. Moreover, 
the present finding that Amaranthus 












Time, minutes 

Fig. 48. Time course of the rate of CO2 exchange in response to changes in oxygen con- 
centration at different temperatures. Same leaf and conditions as in Fig. 47(A). 


TABLE 19. Ph< 

Dtosynthetic CO2 


under 0.2% 

2 c 

ind under 

Air at 

Different Wave- 

lengths in 


lid a go 

virgaurea, H< 


s Vdderb* 


Rate of CO 

'2 Upt< 



Half band- 


cm -2 : 

sec -1 

length, m/i 



In low O2 

In Air 

Inhibition, % 









































* Measurements were made at 22 °C and 0.03% CO2. Light intensity at 665 m/x was 6 X 

1 3 erg cm -2 sec -1 . At other wavelengths the intensity was adjusted to give the same rates of 
CO2 uptake as at 665 m/x. The rates were linearly related to light intensity. 

edtrtis leaves exhibit a pronounced inhibitory effect of 2 on C0 2 uptake 

burst of C0 2 immediately following il- show a temperature dependence con- 

lumination, despite the fact that 21 % sistent with the finding that inhibi- 

2 does not inhibit the net C0 2 uptake tion by glycolate oxidase inhibitors of 
in these same leaves, shows that such the rate of C0 2 production in light is 
a burst does not necessarily indicate strongly increased with temperature. 
an inhibitory effect of 2 . Measurements of the inhibition of 

Conclusions. Experiments designed net C0 2 uptake caused by 21% 2 at 
to test the hypothesis that the inhi- different wavelengths of light give no 
bition of photosynthetic C0 2 uptake indication that a "photorespiratory 
by 2 in normal air can be explained process" whose action spectrum dif- 
simply by an enhancement of the rate fers markedly from that of photosyn- 
of glycolate oxidation mediated by thesis causes the inhibition. 
the enzyme glycolate oxidase have Although no new evidence support- 
yielded negative results. An increased ing such a conclusion has been ob- 
rate of C0 2 fixation in low 2 concen- tained, it seems reasonable that the 
tration was not correlated with ac- inhibition is caused primarily by a 
cumulation of glycolate. Nor does the back-reaction between a highly re- 

TABLE 20. Photosynthetic C0 2 Uptake under 0.2% 2 and under Air at Different 
Wavelengths in Mimulus cardinalis 71 20-7 

Rate of C0 2 




Half band- 

nano-mol cm -2 i 

sec -1 

length, i 




In low O2 

In Air 

Inhibition, % 


























* Measurements were made at 23 °C and 0.03% CO2. Light intensity at 650 m/x was 4 X 
1 4 erg cm -2 sec -1 . At other wavelengths the intensity was adjusted so that the rates of CO2 
uptake were approximately the same as those at 650 m/x. The rates were partially light saturated. 



Light Dark Light Light Dark Light 

3 C 

6 8 

Time, minutes 

Fig. 49. Time course of the rate of CO2 exchange in response to light and dark in Mimulus 
cardinalis, 7211-4; Zea mays, Ferry Morse hybrid 901; and Amaranthus edulis. Seeds of Ama- 
ranthus, native to Argentina, were kindly supplied by Dr. R. S. Loomis and Dr. W. A. Williams 
of the University of California at Davis. Measurements were made at 24 °C and 0.03% CO2. 
White light of an intensity of 3 X 1 5 and 5 X 1 5 erg cm -2 sec -1 (400—700 mfx) was used 
with Mimulus and the two other species, respectively. 



dueed photosynthetic intermediate 
and molecular oxygen that increases 
with 2 concentration. Hydrogen pe- 
roxide resulting from such a back- 
reaction could have several secondary 
effects. For example, it could increase 
the rate of oxidation of glyoxylate ; it 
also could stimulate the production 
of glycolate from intermediates of the 
Calvin cycle as proposed by Coombs 
and Whittingham (1966). 

Comparative studies of photosyn- 
thesis among species differing in their 
response to oxygen concentration ap- 
pear to be a promising approach to 
gaining new insight into the mecha- 
nism of inhibition by 2 . Of particular 
interest in this connection are the re- 
sults of Hatch and Slack (1967), 
which indicate that certain tropical 
grasses, including corn and sugar 
cane, which lack an apparent inhibi- 
tion of C0 2 uptake by 21% 2 , utilize 
a different pathway for photosyn- 
thetic C0 2 fixation than most other 


Coombs, J., and C. P. Whittingham, Proc. 
Roy. Soc, B., 164, 511, 1966. 

El-Sharkawy, M. A., R. S. Loomis, and 
W. A. Williams, Physiol. Plant., 20, 
171, 1967. 

Fock, H., and K. Egle, Beitr. Biol. Pflan- 
zen, 42, 213, 1966. 

Forrester, M. L., G. Krotkov, and C. D. 
Nelson, Plant Physiol, 1*1, 422, 1966a. 

Forrester, M. L., G. Krotkov, and C. D. 
Nelson, Plant Physiol., 41, 428, 1966&. 

Hatch, M. D., C. R. Slack, and H. S. 
Johnson, Biochem. J., 102, All, 1967. 

Hess, J. L., and N. E. Tolbert, Plant 
Physiol, 42, 371, 1967. 

Tregunna, E. B., G. Krotkov, and C. D. 
Nelson, Can. J. Bot., 42, 989, 1964. 

Tregunna, E. B., G. Krotkov, and C. D. 
Nelson, Physiol Plant., 19, 723, 1966. 

Zelitch, I., J. Biol. Chem., 233, 1299, 

Zelitch, I., Plant Physiol, 41, 1623, 1966. 

Effect of Oxygen Concentration 

on Dry Matter Production in 

Higher Plants 

OUc Bjorkman, William M. Hiesey, 

Malcolm Nobs, Frank Nicholson, and 

Richard W. Hart 

The fact that the rate of photosyn- 
thetic C0 2 uptake in most higher 
plants is strongly enhanced by low 2 
content in the atmosphere propounds 
the question of whether or not the 
rate of dry matter production is also 
enhanced. An increase would be ex- 
pected if (1) the enhancement of C0 2 
uptake is due to a net increase in the 
yield, and not merely an effect caused 
by a greater portion of the reducing 
power and phosphate bond energy 
formed in photosynthesis being used 
for the reduction of C0 2 at the ex- 
pense of other vital endergonic proc- 
esses, and (2) if low 2 concentration 
does not adversely affect secondary 
growth processes. 

Although a great deal is known 
about the effects of 2 partial pres- 
sure in the root medium, few investi- 
gations have been concerned with the 
dependence of growth on the 2 con- 
centration of the atmosphere. In re- 
cent years Siegel et al. (1963) have 
shown that a number of plant proc- 
esses such as germination, root, and 
coleoptile elongation are essentially 
unaffected or sometimes even en- 
hanced, by subatmospheric 2 levels, 
whereas senescence is suppressed. In 
some cases early seedling growth was 
found to be somewhat greater at 10% 
as compared with 21 % 2 , and young 
seedlings were able to grow at con- 
centrations as low as 5%. Before any 
conclusions can be drawn as to 
whether the enhancement of C0 2 up- 
take by low 2 is matched by an en- 
hancement of dry matter production, 
however, further experimentation 
under precise control not only of 2 
concentration but also of other ex- 
ternal variables, particularly the con- 
centration of C0 2 , is needed. 



To make such studies feasible, new 
growth chambers were developed 
during the year. A brief description 
of their design and the results of ex- 
ploratory growth experiments under 
different 2 concentrations are given 

Growth cabinets with controlled 2 
and C0 2 concentration. The basic 
construction of the cabinets is the 
same as described previously (Hiesey 
and Milner, 1962) , except that the air 
circulation system is closed to the ex- 
ternal atmosphere. In addition, the 
cabinets are connected with large col- 
lapsible plastic bags (approximately 
500 1) to equalize the internal and 
external pressures. This prevents 

leakage of external air into the cab- 
inets, which would otherwise take 
place when the external pressure in- 
creases or when the internal pressure 
decreases due to decreases in temper- 

A schematic diagram of the control 
system is given in Fig. 50. Thermis- 
tor-operated controllers provide pre- 
cise temperature control. Gas from 
the cabinets is continuously pumped 
through a paramagnetic 2 -analyzer 
(Beckman, Model F3) and an infra- 
red C0 2 -analyzer (Beckman, Model 
15 A or LIRA, Model 300) . Both the 
2 -analyzer and the C0 2 -analyzer are 
equipped with electronic controllers, 
specifically designed for this purpose 

Light Bank 











Cabinet A 

>U th $k 

Cabinet B- 

n^ m w 




N 2 or0 2 in 

C0 2 in 

Fig. 50. Block diagram of gas control system. The valve positions are shown for the control 
cycle of cabinet A. 



by Mark Lawrence. Each consists of 
a solid-state operational amplifier used 
in a voltage comparator circuit which 
compares the output of the gas ana- 
lyzer with an internal reference. 
When the gas concentration deviates 
from the preset level, a transistor 
switch activates a solenoid in the con- 
trol ling gas lines to correct this 

To avoid undesirable differences in 
COj or O a concentration between the 
two cabinets, the same analyzers are 
used for both cabinets. An electronic 
timer switches the gas sampling and 
controlling circuits between the two 
cabinets at preset intervals (usually 
1 min) . When a cabinet is being op- 
erated at or above atmospheric 2 
concentrations, C0 2 -free air is slowly 
fed into the cabinet to prevent the 
CO 2 concentration from increasing 
beyond the preset value ; at subatmos- 
pheric 2 concentrations, the C0 2 - 
free air is replaced with C0 2 -free N 2 . 
The absolute accuracy of C0 2 control 
is about ±5 ppm in the range 200 to 
400 ppm. The maximum difference 
between cabinets (if both are set to 
the same value) is only about ±1 
ppm. The accuracy of the 2 control 
is about ±0.5% in the range 0% to 
25% 2 . 

Light is provided by 96-inch Syl- 
vania VHO fluorescent tubes (2.4 m 
long) , supplemented with incandes- 
cent lamps. Since a single light bank 
is used for the illumination of both 
cabinets, differences in light intensity 
and quality between the two cabinets 
can be kept very small. 

The root medium can be aerated 
with gas whose composition is inde- 
pendent of that of the cabinet atmos- 
phere. Gas-tight seals between root 
and shoot are obtained with foam 
neoprene gaskets (4 mm thick). 
Prior to an experiment, cuttings or 
seedlings are inserted through a tiny 
hole in each gasket at an early stage 
of development. As the stems expand, 

tight seals are obtained by the neo- 
prene gaskets without undesirable 
effects on the plants. 

Transpiration water is condensed 
on the cooling coils of the refrigera- 
tion system and drained through a 
trap. The cabinets work well, and 
several successful runs have been 
completed with Phaseolus, Mimulus, 
and Zea. 

Effect of 2 concentration on 
growth rate. In all experiments de- 
scribed below, the plants were grown 
under continuous light and 0.03% 
C0 2 . Except where specified, the roots 
were immersed in a large volume of 
nutrient solution which was aerated 
with normal air (21% 2 , 0.03% 
C0 2 ). Within each experiment all 
factors but 2 concentration were 
kept constant. 

The short-term response of growth 
to different 2 concentration was 
studied on bean seedlings (Phaseolus 
vulgaris, var. Mexican Red). Care- 
fully matched 5-day-old seedlings 
were grown under 2.5% and 21% 2 
at 29 °C and a light intensity of 5 X 10 4 
erg cm- 2 sec -1 . Plate 1 illustrates the 
response obtained in a 6-day period. 
Clearly the growth was strongly en- 
hanced at the low 2 concentration. 
As shown in Table 21, the net in- 
crease in the total dry weight of the 
seedlings was about twice as high 
with 2.5% as compared with 21% 
2 . All organs showed an enhanced 
growth in low 2 ; there were no 
major differences in their relative pro- 
portions. Root growth was strongly 
increased, even though the roots 
themselves were aerated with normal 

The effect of low 2 concentration 
on the growth of bean seedlings over 
an extended period was studied in 
another experiment. In this case 12 
pairs of 6-day-old seedlings (Phase- 
olus vulgaris, var. Ferry Morse Bush 
869), sampled at random, were 
grown in perlite with nutrient solu- 



TABLE 21. 

Effect of O2 Concentration on Dry Weights of the Bean Seedlings, 
6 days' growth (Plate 1 ) 


21%0 2 , 2.5% O,. 2.5% 

grams grams 21 % 

(1) Leaves, final weight 

(2) Roots, final weight 

(3) Stem, final weight 

(4) Entire plant, final weight 

(5) Entire plant, original weight 





Growth in 6 days, (4) minus (5) 

0.333 0.700 


TABLE 22. Effect of 2 

Concentration on Dry Weights of Mimulus 

21%0 2 , 5%0 2 , 
grams grams 

5% 2 
21 % 2 



Entire plant, final weight* 
Entire plant, original weight 
Growth in 1 days, (1 ) minus (2) 

0.675* ± 0.029 1.186* ± 0.043 
0.110 0.110 
0.565 1 .076 


* Means are significantly different at P = 0.01. 

tion fed automatically. The roots 
were kept at the same 2 concentra- 
tion as the shoots. Light intensity was 
8 X 10 4 erg cm -2 sec -1 , and tempera- 
ture was 27 °C. Plate 2 illustrates the 
typical appearance of the plants after 
17 days under 5% and 21% 2 . No 
adverse effects of growing the plants 
under low 2 concentration were 
found. The plants grew very rapidly 
under these conditions, and produced 
an abundance of flowers. The net in- 
crease in shoot dry weight was about 
twice as great under 5% 2 as under 
air. No quantitative measurements of 
root weights were made, but it was ob- 
vious that root growth also was gen- 
erally greater in the plants grown in 
low 2 . With beans chlorosis became 
increasingly apparent in developing 
shoots after about 10 days in normal 
air. Presumably the effect is caused 
by growing the bean plants under 
continuous light. Interestingly, no 

signs of chlorosis could be detected in 
the plants grown under 5% 2 . 

Another noteworthy effect of grow- 
ing the bean plants under low 2 is 
that the rhythmic movement of pri- 
mary leaves was arrested after about 
two days. In air this movement con- 
tinued for at least five more days. 

Mimulus cardinalis, Jacksonville 
1211-4;, provides an experimental ma- 
terial much better suited for the pres- 
ent studies than Phaseolus. Uniform 
cuttings of a single clone can easily 
be obtained, and the problem of ge- 
netically variable seedling material 
can be avoided. Moreover, continuous 
light has no adverse effects on Mimu- 
lus; on the contrary, it results in a 
very rapid growth. 

The results obtained with Mimulus 
thus far indicate that the yield in dry 
matter production is strongly en- 
hanced by low 2 . As shown in Table 
23 the net gain in dry weight was 
much higher in 5% and 2.5% 2 than 



TABLE 23. Effect of 2 Concentration on Mimulus, dry weights 

21% 2 , 


2.5% 2 , 

2.5% Q 2 

21%0 2 

(1) Entire plant, final weight* 

(2) Entire plant, original weight 

(3) Growth in 10 days, (1) minus (2) 

0.589 ± 0.017 



0.887 ± 0.036 1.51 


0.817 1.58 

* Means are significantly different at P = 0.01. 

in air. under the conditions tested 
(7.2 X 10 4 erg cnr 2 seer 1 , 24°C). 
It appears that 2.5% 2 might be too 
low for maximum enhancement of 
dry matter production. 

Growing Mimulus under 2.5% and 
d c c Oo resulted in a greater root-to- 
shoot ratio, and a lower water content 
of the shoots, than in air (9.8% dry 
matter in low 2 versus 8.4% in air). 
The plants grown under low 2 had 
thicker leaves and a more compact 
growth habit than those grown in air. 

Photosynthesis measurements on 
Mimulus and Phaseolus leaves 
showed that the degree of inhibition 
of net C0 2 uptake by 21% 2 was 
unaffected by growing the plants un- 
der different 2 concentrations. The 
inhibition was about 30% regardless 
of preconditioning. The absolute rate 
of light-saturated C0 2 uptake on a 
leaf area basis was higher in leaves 
developed in low 2 than in air. This 
was associated with a greater leaf 
thickness of the former. 

As was mentioned in the preceding 
section of this report, Zea mays 
(corn) differs from most other 

higher plant species in that it lacks 
an apparent inhibition of C0 2 uptake 
by 21% 2 . It seemed of interest 
therefore to compare its response to 
different 2 concentrations during 
growth with that of Mimulus and 
Phaseolus. In a preliminary experi- 
ment, five matched pairs of corn seed- 
lings, var. Ferry Morse Hybrid 901, 
were grown under 5% and 21% 2 , 
with other conditions the same as 
those given for Mimulus. As shown in 
Table 24, the enhancement of dry 
matter production in low 2 , if any, 
was much less than in Mimulus and 
Phaseolus. Interestingly, however, 
the root-to-shoot ratio and the per- 
centage of dry matter in the shoots 
(10.8% in low 2 , 6.3% in air) were 
higher in the plants grown in low 2 . 
This was also true of the other two 


Hiesey, William M., and Harold W. Mil- 
ner, Bot. Gazette, 124, 103-118, 1962. 

Siegel, S. M., L. A. Rosen, and C. Giu- 
marro, Nature, 198, 1288, 1963. 

TABLE 24. Effect of O2 Concentration on Dry Matter Production of Corn Seedlings 

21 % 2 , 


5% 2 


5% 2 
21%0 2 




Entire plant, final weight* 
Entire plant, original weight 
Grov/th in 1 days, (1 ) minus (2) 

1.479 ± 0.123 



1.683 ± 0.129 




* Difference between means is scarcely significant: 0.3 < P < 0.4. 


Hybridizations in Solidago California, and between these two 

** i 7 a -KT t. sMi t>- i i and lowland counterparts from the 

Malcolm A. Nobs, Olle Bjorkman, and , . . , „ ~ , , 

William M.Hiesey same tw0 continents, all from about 

39°N. latitude. In addition, races 

The exploratory crossings reported from approximately 70 °N in Scan- 
in Year Book 65, p. 471, between dif- dinavia and in Alaska are being 
ferent forms of the Solidago virgau- crossed with each other, and also with 
rea-S. multiradiata complex have the southern races just mentioned, 
yielded a number of F t hybrids now Observations on the responses of pa- 
growing in the Stanford garden that rental races and their F a and F 2 prog- 
show interesting and complex char- enies at the Stanford, Mather, and 
acteristics. Among these, for ex- Timberline transplant stations are 
ample, is a cross between the shade essential to interpret the genetic corn- 
form of Solidago virgaurea from Hal- position and potential of the parental 
lands Vadero in southern Sweden, races. 

mentioned in earlier sections of this We are indebted to Dr. John Ko- 

report, and a subalpine from S. randa and Dr. William Mitchell of the 

multiradiata from near our Timber- Alaska Experiment Station at Palmer 

line transplant station in California, for collections of living material of 

The success of this and other hybrid- Alaskan forms, 
izations has prompted us to follow a 
more extensive program this year 

utilizing other intercontinental com- Comparative Physiological Studies 
binations featuring interaltitudinal on Solarium dulcamara 
and interlatitudinal crosses between 
contrasting races of what appear to 
be members of the Solidago virgau- The recent developments in plant 
rea-S. multiradiata species-complex ecological research aimed at uncover- 
that is widely distributed over the ing internal physiological mecha- 
northern hemisphere. nisms that have evolved in plants 

Our first objective is to determine native to different climates excite in- 

the degrees of biosystematic relation- quiries concerning species groups 

ship between key ecological forms other than Solidago and Mimulus. The 

that have been referred to various combined results from independent 

named taxa, yet appear to be closely studies on a diversity of species are 

enough related to be capable of gene more likely to lead to the discovery of 

exchange. general principles than the study of 

A second objective is to compare only one or two species-complexes, 

the genetic constitution of parallel The choice of material to be used 

ecological races that are found in for this kind of work is of utmost 

corresponding kinds of habitats in importance because of the exacting 

widely separated continents. Such practical experimental requirements 

races apparently were evolved inde- for such investigations. Essential re- 

pendently over long periods of geo- quirements are (1) that the species 

logical time from a common ancestral occur over a wide geographical range 

stock. having a great diversity of habitats, 

The hybridizations that are being (2) that its races be easy to grow and 
made include combinations between be propagated as clones in order to 
high-altitude forms from the Sierra facilitate experiments with geneti- 
Nevada of Spain and high-altitude cally identical material in various con- 
forms from the Sierra Nevada of trolled environments, and (3) that its 

Eckard Gauhl 



leaf and stem structure be such as to 
facilitate measurements of gas ex- 
change in intact leaves while still 
attached to the living plant. 

Solatium dulcamara L., the bitter- 
sweet, meets these requirements 
well. It is distributed in Europe from 
the Atlantic coast in France north- 
ward to central Norway, and south 
to northern Africa. It extends to 
India and Asia in the east and occurs 
in scattered areas throughout China 
and Japan. In central Europe it grows 
in shady swamps with alder (Alnus 
glutinosa), or in deep shade within 
stands of reed grass (Phragmites 
communis) along borders of lakes. 
Other races occupy exposed habitats 
on coastal sand dunes near sea level, 
or on rocky slopes of southern ex- 
posure in the alps up to altitudes of 
1,700 m. The most contrasting eco- 
logical races are readily distinguish- 
able morphologically. 

A number of select clones have 
been collected in central and southern 
Europe, some of which have been 
brought to Stanford under a U.S. De- 
partment of Agriculture import per- 
mit. Current work is being confined 
mainly to three clones that represent 
extremes of habitat origin and of 
growth responses to controlled con- 
ditions. One clone originally from a 
reed grass marsh near Frankfurt 
grows better under low light intensity 
than under high. Another from a 
sand dune along the Baltic Sea in 
northern Germany that has thick, 
fleshy leaves and short internodes be- 
comes prostrate when grown under 
strong light, and semi-erect under 
weak light. A third clone, originally 
from a dry, exposed rocky habitat 
near Rovinj along the northern Dal- 
matian coast of Yugoslavia, resembles 
in growth habit the inland marsh 
form, but grows better under strong 
than under weak light. 

Comparative studies on rates of 
C0 2 uptake of intact leaves when 

grown under high and low light in- 
tensities have been started along lines 
described above for clones of Solidago 
and Mimulus. It is planned to pursue 
these studies in other directions also, 
with special consideration of inter- 
relations between water metabolism 
and photosynthetic performance in 
contrasting climatic races of Sola- 

Clusters of Tree Species on Both 
Sides of the Pacific 

Jens Clausen 

Trees are suitable for use in gross 
surveys of the vegetations of the 
world. They are conspicuous enough 
to be easily seen and fairly good rec- 
ords of the limits of tree species are 

Records and experimental data in- 
dicate that the greater number of the 
world's tree species are arranged in 
clusters of species, which experimen- 
tal taxonomists call "cenospecies." 
Certain taxonomic sections of genera 
are on the size-order of species clus- 
ters. The species of such a cluster are 
closely enough related to permit occa- 
sional interchange between their 
heredities, either at present or in the 
past. The evolutionary significance of 
the species cluster has generally been 

Species clusters of this kind circle 
the earth within definite latitudinal 
belts. Modern taxonomists, however, 
constantly look for minor differences 
between the taxa within each cluster 
to apply different species names to 
them in various parts of the world. By 
splitting the species clusters, some- 
times even into separate genera, sense 
has been lost of the deep-seated phys- 
iological and ecological character- 
istics that are built into the heredities 
of each cluster and that can be ex- 
pressed in morphological characters. 
Heredity holds the natural clusters 


subunits. The principle here discussed species of the larch cluster are now 

characterizes also herbaceous species being commercially intercrossed, 

complexes around the world. Two species clusters belong to the 

Such facts suggest that genes that genus Pinus, namely a two-needle and 

govern morphology and those govern- a five-needle complex of species. Spe- 

ing physiology are correlated in their cies belonging to different clusters do 

inheritance. Such correlations are not interbreed. This hereditary sepa- 

based on the principle of interlocking ration enables the species of two clus- 

heredity (coherence) that we have ters to maintain their identity when 

shown exists between climatic races growing together in the same local 

of plants. Similar although much area. 

stronger correlations probably limit The two-needle cluster of species 

families and orders to their climatic includes the North American lodge 

zones. pole pines, P. contorta Doug., P. mur- 

Developing a sense of close rela- rayana Grev. et Balf., and P. bank- 

tionship among the members of a siana Lamb., with the Japanese P. 

species cluster brings a sense of per- thunbergiana Franco, P. densi flora 

spective that is vital to the study of Sieb. et Succh., and with the scotch 

taxonomy. A study of the tree species pines of Siberia to Europe, P. silves- 

clusters on both sides of the Pacific tris L., P. montana Mill., and the 

illustrates the principles involved in dwarf P. mugho Turra. Pinus silves- 

this kind of modern plant taxonomy, tris is the type species of the cluster. 

Based on worldwide field experi- Many natural hybrids have been sus- 

ences, Misao Tatewaki's comprehen- pected among the latter three and 

sive report (1958), supplemented by among the Japanese species. Righter 

personal discussions with him on field and Stockwell (1949) described the 

tours in Hokkaido in 1966, have fertile species hybrid P. murrayana 

deeply influenced the report to follow. X P- banksiana, and intermediates 

Trees of high latitudes. On both are known where the two species meet 
sides of the Pacific approximately 12 in the wild. Like the larches, the two- 
clusters of tree species follow a wide needle pine cluster circles the earth 
belt, extending from about 70° to 35° and apparently shares a common gene 
North latitude (Table 25). The indi- pool. 

vidual species of each of the 12 clus- The five-needle pine cluster in west- 

ters are species of high latitudes and era North America is represented by 

high altitudes. Interestingly enough, the hardy whitebark pine P. albi- 

6 of the 12 clusters belong to the caulis Engelm. On the Asiatic side of 

pine family Pinaceae. Carolus Lin- the Pacific are the species P. pumila 

naeus (1753) included all of the 6 Regel, P. koraiensis Sibth. et Succh. 

clusters in the genus Pinus. of Northeastern Asia to Japan, and 

The top cluster in Table 25 in- P. sibirica du Tour, from the Altai 

eludes all species of the larch genus to the Ural Mountains and Peninsula 

Larix. There are North American L. Kola. The type species of the cluster 

laricina Koch, the Japanese L. lep- P. cembra L., occurring in the Car- 

tolepis Gord., the East Asiatic L. pathian Mountains to the Swiss Alps 

kamtschatica Carr., L. gmelini Gord., (Critchfield and Little, 1966), com- 

and L. sibirica Ledeb., and the type pletes the circle. P. cembra, P. sibi- 

species of the cluster, the Euro- rica, and P. koraiensis are generally 

pean L. decidua Mill. C. Syrach Lar- erect trees; P. pumila is always an 

sen (1937) found that all these spe- elfinwood bush, but P. albicaulis is a 

cies are able to interbreed, and the tree in subalpine environments and 


TABLE 25. Trees of Highest Tolerance, Northern Hemisphere, High Latitudes, 35°— 70°N 

Eastern Hemisphere 



1 . Larches 

2. Scots pines 

3. Cembras 

4. Spruces 

5. Firs 

6. Hemlocks 

7. Junipers 

8. Aspens 

9. Willows 

10. Birches 

11. Alders 

1 2. Mountain 

PINACEAE, n = 12 

LARIX decidua sibirica-kamtshatica 


PINUS silvestris silvestris-densiflora 


PINUS cembra sibirica-koreensis-pumila 

PICEA excelsa .... obovata-jezoensis-glehnii 

ABIES alba . . . .sibirica-sachalinensis-veitchii 


TSUGA (extinct) diversifolia-sieboldii 


JUNIPERUS communis 


POPULUS tremula tremula-jezoensls 


SALIX phylicifolia pyrifolia-pulchra 


BETULA tortuosa cajanderi-ermani 

ALNUS incana maximovicsii-inokumae-crispa 

ROSACEAE, n = 17 

SORBUS aucuparia amurensis-scopulina 






mertensiana canadensis 




incana-crispa .... 



Dots indicate that the particular species is continuous over the region. The second lines of the 
clusters indicate the geographic location of certain southern species. 

an el fin wood cushion type above the 
tree line (Year Book 62, pp. 394-398; 
and Clausen, 1965). Members of this 
cluster of species extend to and be- 
yond the tree line in North America 
and Eurasia. Intermediates are re- 
corded where the species meet. This 
cluster of species and the members of 
the two-needle cluster occur together 
in many populations but do not inter- 

As far as is known at present, the 
spruces, firs, and hemlocks constitute 
three other high latitude circumpolar 
clusters of trees, as shown in Table 
26. The spruces (Picea), firs (Abies), 

and hemlocks (Tsuga) are composed 
of several geographic species that 
have many intermediates. Species 
within the Picea and Abies clusters 
are known to interbreed (Wright, 
1953) . Like the other members of the 
pine family, all appear to have 12 
pairs of chromosomes, and they were 
included by Linnaeus as species of the 
genus Pinus. The presently recog- 
nized species represent geographic, 
climatic, and edaphic adaptations to 
many environmental niches within 
the belt covered by the three genera 
around the world. The species of a 
cluster share a common gene pool. 


Juniperus communis L. is still found a diploid Japanese species, 

recognized as one species around the Alnus inokumae Mur. et Kus., having 

world by conservative taxonomists. n = 7. Chiba was able to cross it with 

Its lowland and inland forms are of tetraploid forms of A. hirsuta Turcz., 

erect columnar shape, whereas forms A. glutinosa Gaertn., and A. japonica 

from windy coasts and alpine habi- Call., n = 14, and the hybrids have 

tats are low elfinwood cushions. In 2n = 21 chromosomes (Chiba, 1966: 

transplant experiments in Denmark, 1-16). The genus is commonly sup- 

C. Syrach Larsen has shown that the posed to have many hybrids around 

differences in growth form are hered- the earth. 

itary. The junipers of this species The birches, Betula, also intercross 

have 11 pairs of chromosomes and irrespective of differences in chromo- 

cover approximately the same wide some number and of differences in the 

belt as the six clusters of conifers appended species name. Helms and 

previously discussed. Jorgensen (1925) found that Betula 

Five nonconif erous high-latitude alba, n = 28, growing in a wet Danish 
clusters of tree species include the moor, crossed with B. verrucosa, Ehr., 
aspens, willows, birches, alders, and n = 14, along drier, sandy edges of the 
mountain ashes (Table 25). Each of moor. Partially fertile hybrid prog- 
these is composed of closely related enies ranging from approximately 
species that together circle the earth 21 pairs to intermediate numbers be- 
at high to medium North latitudes, tween 21 and 27 were found along the 
All are well known for their inter- borderline of the two populations, 
specific crossability ; the willows, suggesting that possibly 7 is the basic 
birches, alders, and mountain ashes chromosome number also in Betula. 
cross irrespective of differences in The variation in birch is so great that 
number of chromosomes. classification is very difficult, but 

The quaking aspens constitute a Table 25 lists some of the names ap- 

circumboreal natural group of spe- plied to its species around the high 

cies, differing from other poplars by latitudes. 

their ability to spread vegetatively An equal number of species names 

through root shoots. A stand of 20 have been applied to the Sorbus aucu- 

to 50 trees may consist of one indi- paria cluster, the mountain ashes of 

vidual. The aspens have 19 pairs of the rose family, having 17 pairs of 

chromosomes, and the North Amer- chromosomes. In North America are 

ican Populus tremuloides Michx. is Sorbus americana March., S. decora 

being commercially intercrossed with Sarg., S. scopulina Greene, and S. 

the Asiatic-European P. tremula L. sitchensis Roem; in Japan are S. 

The hybrids are fertile and possess sambucifolia Roem., S. matsumurana 

hybrid vigor. The Japanese aspens, Koehne, and S. commixta Hedl. ; in 

P. jezoensis Nakai and P. davidiana northeastern Asia is S. amurensis 

Dode, have rhizomes like the other Koehne, and in Europe, the type spe- 

quaking aspens, and P. davidiana has cies S. aucuparia L. The growth 

been crossed with P. tremula (Chiba, forms vary greatly among the mem- 

1966). ber species, but a member of this 

The North American-Eurasiatic highly variable cluster can always be 

alders, Alnus, were thought to follow expected to occur at high latitudes 

a 14-series in chromosome number and high altitudes. 

(Gram, Muhle Larsen, C. Syrach The 12 cold-tolerant species clus- 

Larsen, and Westergaard, 1941). ters discussed here are on both sides 

Chiba, however (1966: 155-156) , of the Pacific, but the member species 


within each cluster are difficult to dis- monticola Dough, and possibly the 

tinguish because they share common sugar pine, P. lambertiana Dougl. On 

gene pools. Members of distinct clus- the western side of the Pacific are the 

tors, on the other hand, are clearly Armand pine, P. armandi Franch, 

separated from each other. native to Taiwan, South China, and 

Within the southern hemisphere Vietnam, the Indian blue pine, P. 
there are no species clusters that in griffithi McClelland of the Himalayas 
tolerance to extremely cold climates to Kashmir, and the Balkan pine, 
match the 12 uniquely cold-tolerant P. pence Griseb., native to southeast- 
northern hemisphere species clusters, ern Europe. Little and Righter(1965) 

Clusters of tree species of medium intercrossed these species in all direc- 

cold tolerance. A different and much tions, producing successful hybrids, 

larger group of tree species occupies The white pines accordingly form an 

medium latitudes from about 55° to evolutionary cluster that spans the 

25 °N. on both sides of the Pacific, earth at medium latitudes. 

Some of these species and genus clus- Several genera of the Taxodiaceae- 

ters grow at comparable latitudes on Cupressaceae group circle the earth, 

both sides of the Pacific in both the and they follow an 11-series in chro- 

northern and southern hemispheres, mosome numbers both north and 

The several thousand medium tolerant south of the equator and across the 

tree species are less tolerant to cold oceans (Table 26). These genera con- 

than the species native to the high tain the cypresses, the incense cedars, 

latitudes, but they are much more thujas, sequoias, and swamp cy- 

tolerant than species native to the presses. Specialists have split each of 

low latitudes on both sides of the these clusters into distinct genera by 

equator. At latitudes where species calling attention to small differences, 

clusters of both high and medium but have neglected the gross simi- 

tolerance overlap, those of medium larities and adaptations and chromo- 

tolerance grow at lower altitudes than somal heredity that unite them across 

the species of the highly cold-tolerant the oceans and across the equator, 

clusters. One of these clusters is the cy- 

Like the high latitude species, presses, Cupressus-Chamaecyparis. 
those of medium tolerance are orga- Along the west coast of North Amer- 
nized in evolutionary clusters of ica five to six Cupressus and two 
closely related species. Sufficient ex- Chamaecyparis species are recog- 
perimental evidence is on hand to nized. Osborn (1941) described, how- 
indicate that the species of several of ever, a hybrid between the Monte- 
these clusters share common gene rey cypress, Cupressus macrocarpa 
pools on both sides of the oceans. Hartw., native to the central Cali- 
Table 26 lists examples of tree species fornia coast, and the Alaska cedar, 
clusters of medium tolerance on both Chamaecyparis nootkatensis Spach., 
sides of the Pacific. a mountain species growing from 

The top cluster is a belt of white northern California to Alaska. This 

pines of medium tolerance having five "intergeneric" hybrid is fertile, and 

needles and elongated cones that con- its progeny segregates the parental 

stitute the taxonomic section Strobus. characters, suggesting that at least 

On the North American side this clus- these two species belong to one genus 

ter is marked by the type species, the and that in this group the prevailing 

eastern white pine, P. strobus L., the taxonomy is misleading. In Japan 

Great Basin limber pine, P. flexilis Chamaecyparis pisifera Lindl. was 

James, the western white pine, P. previously considered a Thuja, and C. 



TABLE 26. Examples of Trees of Medium Tolerance, Latitudes 20°-55°, Both Hemispheres 

(several thousand species) 

Pacific Western 

Eastern Hemisphere Ocean Hemisphere 

PINACEAE, n = 12 

White pines: PINUS peuce-griffithii-armandi monticola-strobus 


,.~*^r-<r. n ..r. I Taiwan Sierra Nevada 

Incense cedars: UBOCEDRUS ... j New ZeaIand ai|e 


Cypresses: CHAMAECYPARIS obtusa nootkatensis 

CUPRESSUS sempervirens macrocarpa 

Thujas: THUJA orientaiis stondlshii plicata 

FAGACEAE, Northern Hemisphere, n = 1 2 

White oaks: QUERCUS robur crispula lobafa-alba 

Chestnuts: CASTANEA sativa crenata . . .americana 

CASTANOPSIS cuspidata chrysophylla 

Beeches: FAGUS silvatica crenata grandifolia 

NOTHOFAGUS, n = 13, Tasmania, New Zealand Chile, Patagonia 


MAGNOLIA obovafa . . . grandiflora 

TILIACEAE, x = 41 

Basswood: Till A cordata amurensis-japonica . . .americana 

ACERACEAE, x = 1 3 

Maples: ACER platanoides mono macrophylla- 


obtusa was shifted from Cupressus to 
Chamaecyparis. The European spe- 
cies is referred to as Cupressus sem- 
pervirens L. 

Excessively detailed taxonomic 
studies have been even more mislead- 
ing in presenting relationships be- 
tween the incense cedars, the old 
genus Libocedrus, which spans the 
Pacific both north and south of the 
equator (Table 26). Based on differ- 
ences in the arrangements of stomata 
and leaves, and disregarding the re- 
markable gross similarities, the genus 
has been split into five. The chromo- 
some number is 11 in L. bidwilli 
Hook, f ., and L. plumosa Sarg. of New 
Zealand (Hair and Beuzenberg, 
1958) , and also n = 11 in the Chilean 
L. chilensis Endl. (Hunziker, 1961) 
on the other side of the Pacific, evi- 

dence that this odd number of chro- 
mosomes persisted through epochs of 
time and wide spatial isolation. 

The Sequoia-Taxodium group 
(Table 26) represents another 11- 
chromosome conifer cluster. It is now 
split into the genera Cryptomeria and 
Metasequoia on the west side of the 
Pacific, and Sequoia, Sequoiaden- 
dron, and Taxodium on the east side. 
Miki (1965) showed that counterparts 
of Sequoia and Sequoiadendron are 
present in Tertiary beds in Japan. In 
past ages Sequoia relatives covered 
north latitudes between 34° and 58° 
around the earth (Florin, 1963:201 
and maps 28-30) . Embryologies differ 
in this primitive conifer group but 
one senses great similarity between 
the assumed genera of this cluster al- 
though experimental evidence is 



Among the broad-leaved trees of 
the medium latitude and altitude the 
members of the beech family, Faga- 
ceae, stand out (Table 26). The oak 
genus. Que reus, circles the earth at 
medium northern latitudes, although 
a few species spill over to the vicinity 
of the equator and beyond at high 
altitudes in Central America and 
Indonesia (Oersted, 1871). Applying 
conservative classification, between 
300 and 500 species are known in the 
genus, which is divided into six to 
eight subgenera, each having several 

Hybridization is common between 
oak species around the world, but the 
genus contains many genetically sep- 
arate species clusters. The clusters 
are deciduous in the northern parts 
of the oak belt, and evergreen in the 

The chromosome number is uni- 
formly n = 12 in about 58 species of 
Qua reus, Castanea, Castanopsis, and 
Fagus, a situation similar to that of 
the pine species. 

The white oaks in Table 26 ex- 
emplify one widely distributed de- 
ciduous cluster of oak species, namely 
Que reus alba L. of eastern North 
America, Q. lobata Nee of California, 
Q. crispula Nee of Japan, and Q. 
robur L. of Europe. Members of dis- 
tinct oak clusters may grow adja- 
cently without intercrossing. 

The beeches, Fagus, have probably 
only one species cluster around the 
earth, but hybrids between the Euro- 
pean Fagus silvatica L. and the North 
American F. grandifolia Ehrh. are be- 
ing produced by shipped pollen at the 
Horsholm Arboretum in Denmark, 
which suggests a close genetic rela- 

The southern hemisphere genus 
Nothofagus is morphologically a close 
counterpart to Fagus. It crosses the 
Pacific between South America, New 
Zealand, and Australia. The southern 
hemisphere beeches differ from Fagus 

in having 13 instead of 12 pairs of 
chromosomes as seen in the geneti- 
cally closely related red and mountain 
beeches N. menziesii (Armstrong and 
Wylie, 1963). Nothofagus possibly 
differentiated chromosomally from 
the northern beeches very long ago, 
but their evolutionary history is still 

Other genera that cross the Pacific 
and have closely related species on 
both sides are the magnolias, the bass- 
woods, and the maples. Their chromo- 
some numbers follow unusual odd 
basic series that remain the same on 
both sides of the Pacific (Table 26). 
Such innate constancy, correlated 
with the generic morphologies and 
persistence within the latitudinal belt 
on both sides, testifies to coherence in 
heredity that persisted through geo- 
logic ages as plant families moved 
around the world. 

Low latitude tree species. The 
greatest number of the world's recog- 
nized tree species, possibly some 
60,000 to 80,000, have remained 
within a belt of some 50° in latitude 
on both sides of the equator. They 
are here because most of them are 
tropical and unable to tolerate even 
occasional mild frosts. This charac- 
teristic contrasts sharply with the 
trees of the high latitude group that 
exist in regions that are subjected to 
frosts eight to nine months of the 

The low latitude tree species occupy 
the forests on both sides of the Pacific 
and the oceanic islands, extending 
through southeastern Asia to Africa 
and to both of the Americas, includ- 
ing the Caribbean Islands. Many fam- 
ilies and orders are restricted to these 
latitudes. Those families that extend 
into the temperate zones have evolved 
genera of different kinds within the 
temperate zones. 

Within the low latitudes the tree 
lines are at remarkably low altitudes 
(Clausen, 1963), and species native 


TABLE 27. Examples of Trees of Low Tolerance, Low Latitudes, 25 'N to 25 °S 
(Major Portion of World's Trees, Possibly 50,000 to 70,000 Species) 

Eastern Hemisphere Pacific Ocean Western Hemisphere 


ARAUCARIA species in Australia, New Zealand Peru, Brazil 



New Guinea to 3600 m, New Zealand Peru, Brazil 

RANALES, buttercup-magnolia relatives; 

DRIMYS, Tasmania, New Guinea to 4000 m Peru, Brazil 


Cunonias: WEINMANNIA, New Zealand, Hawaii Andes, Brazil 

Rosaceae, POLYLEPIS leptophylla dry Andean, 3000 m 

Legumes, SOPHORA, New Zealand, Hawaii, 3000 m high Andes 

PARIETALES, camellia relatives 

TERNSTROEMIA, CLUSIA Brazil, tree line 1900 m 

MYRTALES, myrtles, melastomes 

EUGENIA, ca. 1 000 species 

Africa, India, Indonesia, Australia, Pacific Islands Mexico, Andes, Brazil 

RUBSALES, coffee and camphor relatives 
COFFEA, Congo, Ethiopia 
CINCHONA Peruvian Andes, 3700 m 

Missing at low latitudes: willows, walnuts, birches, alders, oaks, beeches, chestnuts. 

there may show apparent damage terials from which trees capable of 

even before the freezing point of surviving at high altitude might have 

water has been reached. evolved. Characters used in plant 

At the equator only very few tree classification are generally morpho- 
species occur at altitudes beyond 3000 logical. One must therefore conclude 
meters (Table 27). Such trees include that physiological adjustments to cli- 
certain Araucarias and Podocarps, matic zones are somehow tied to the 
Drimys, which is remotely related to morphology of the major plant orders 
the buttercups and magnolias ; the so- and families. These gross ad just- 
called pepper tree, Schinus molle, of ments are retained as members of 
the Anacardiaceae ; the camphor tree, such families migrate across oceans. 
Cinchona, of the Peruvian cloud for- The genus Eugenia of the myrtle 
ests ; and Polylepis leptophylla, a tree- family provides an illuminating ex- 
like plant of the rose family, through- ample. It has evolved about 1,000 spe- 
out the drier Andes. The base of cies that cover both of the Americas 
Table 27 indicates also that typical at low latitudes — besides 100 in the 
mid-latitude tree families miss the Caribbean Islands, 45 in the Pacific 
low latitudes completely. Islands, about 34 in Australia and 

The many thousands of tree species Southeast Asia, 150 in the Indies, and 

that occur only below the frost line 45 in Africa. It has been able to move 

suggest that there is no lack of ma- across oceans, but has not extended 



into cooler latitudes. The large num- 
ber of species indicates a high degree 
of morphological variability, but this 
and many other genera appear to be 
trapped climatically. 

In the discussion above it has been 
necessary to amplify somewhat the 
concept of vegetation in relation to 
latitudinal and altitudinal zones. Veg- 

• ions overlap, permitting mem- 
bers o( the hardiest group to adjust 
to habitats within the medium lati- 
tude zone by occupying the higher 
altitudes or cooler local habitats. Sim- 
ilar overlapping is found between the 
medium and low latitude plants, so 
that certain species of a cluster may 
spill over to a different belt. 

It is necessary in studies of this 
kind to apply a relatively conserva- 
tive concept of species and genera. 
Modern tendencies to split species, 
genera, and families obscure the 
broad, deep-seated physiological and 
ecological relationships that are built 
into the various heredity patterns 
and hold the natural entities together 
even on their worldwide migrations. 


Armstrong, Jean M., and Ann P. Wylie, 
A new basic chromosome number in 
the family Fagaceae, Nature, 205, 
1340-1341, 1963. 

Iba, Shigeru, Studies on tree improve- 
ment by means of artificial hybridiza- 
tion and polyploidy in Alnus and Pop- 
. Hull. Oji Inst. Forest Tree 
Improvement, l, 70-165, 1966. Oji 
Paper Co., Kuriyama, Hokkaido, Ja- 

n, Jens, Tree lines and ^erm plasm 
-a study in evolutionary limitations, 
Proc. Natl Acad. Sci., 50, 860-868, 

Clausen, Jens, Population studies of al- 
pine and subalpine races of conifers 
and willows in the California hiprh 
Sierra Nevada, Evolution, 19, 56-58, 

Critchfield, William B., and Elbert L. 
Little, Geographic distribution of the 
pines of the world, U.S. Dep. Agr. 
Forest Ser., Misc. Pub. 991, iii + 93 
pp., maps 1-51, 1966. 

Florin, Rudolf, The distribution of coni- 
fer and taxad genera in time and 
space, Acta Hort. Berg., 20, 122-312, 

Gram, K., C. Muhle Larsen, C. Syrach 
Larsen, and M. Westergaard, Contri- 
butions to the cytogenetics of forest 
trees, Roy. Vet. Agr. Coll. Yearbook, 
Copenhagen, 44-58, 1941. 

Hair, J. B., and E. J. Beuzenberg, Con- 
tributions to a chromosome atlas of 
the New Zealand flora, New Zealand J. 
Sci., 1, 617-628, 1958. 

Helms, Anna, and C. A. Jorgensen, Bir- 
kene paa Maglemose, Botan. Tidsskr., 
39, 57-134, 1925. 

Hunziker, Juan H., Estudios cromosomi- 
cos en Cupressus y Libocedrus, Rev. 
Invest. Agr. 15, 169-185, Buenos Aires, 

Larsen, C. Syrach, The employment of 
species, types and individuals in for- 
estry, Roy. Vet. Agr. Coll. Yearbook, 
Copenhagen, vii + 151 pp., PI. I-VI, 

Little, Elbert J., and Francis I. Righter, 
Botanical descriptions of forty artifi- 
cial pine hybrids, U.S. Dept. Agr. 
Forest Service Tech. Bull., 1345, 47 pp., 
Washington, B.C., 1965. 

Miki, Shigeru, Sequoiadendron prima- 
rium Miki n. sp. and Sequoia couttisiae 
Heer from Tertiary beds in Japan, 
Bull. Mokugawa Women's Univ., 13, 
1-7, 1965. 

Oersted, A. S., Bidrag til Kundskab om 
Egefamilien., Vidensk. Selsk., Skr. 5 
Raekke, Naturv. Matem. Afd., 9, 335- 
506, Tab. I-VIII + 1 map, 1871. 

Osborn, A., An interesting hybrid coni- 
fer, Cupresso-Cyparis leylandii, J. Roy. 
Hort. Soc, 66, 54-55, 1941. 

Righter, Francis I., and Palmer Stock- 
well, The fertile species hybrid, Pinus 
murraybanksiana, Madrono, 10, 65-69, 


Tatewaki, Misao, Forest ecology of the Wright, Jonathan W., Summary of tree- 
islands of the North Pacific Ocean, breeding experiments, U.S. Dep. Agr., 
J. Fac. Agr., Hokkaido Univ., 50, 371- Forest Serv., Northeast. Forest Exp. 
484, PL I-XXX, 1958. Sta., Sta. Paper No. 56, 47 pp., 1953. 


Several botany classes from the visiting" with friends and colleagues 

University of California at Davis and of up to 40 years' standing, Clausen 

at Santa Cruz, and from Stanford, was introduced to Japanese investi- 

visited the laboratory and the moun- gators in cytogenetics, ecology, taxon- 

tain stations during the year. Demon- omy, and plant breeding at the 

strations and talks were given by the universities in Tokyo, Kyoto, and 

Experimental Taxonomy group. The Sapporo, and the National Institute 

Bay Area Biosystematists group met of Genetics at Misima. 

at the laboratory in January. Studies of the Japanese forest veg- 

The greenhouses and the garden at etation were made possible by a spe- 

the laboratory were used by Dr. Peter cial traveling symposium September 

Raven of Stanford for his studies of 4 to 9 through the Japan Alps. The 

the evening primrose family. Mrs. symposium subject was "The Ecologi- 

Mary Mantuani, a graduate student cal Basis of Nature Conservation in 

at Duke University, started a com- Alpine and Subalpine Zones." About 

parative study of water relations of 70 biologists from different countries 

ecological races of Solidago at participated in the symposium orga- 

Mather. nized by Professor Makoto Numata of 

The Northern California Photobi- Chiba University. After the tour, Dr. 

ology and Photochemistry Group vis- Clausen and Professor Misao Tate- 

ited the Department in January and waki of the University of Hokkaido 

heard short talks from each of the studied the forests of Hokkaido for 

Photosynthesis group. Dr. Brown about a week. 

served as vice president of the Group On September 17 and 18 Dr. Shi- 

this year. geru Chiba, Director of the Kuriyama 

Dr. Jens Clausen attended the Institute for Improvement in Forests, 

Eleventh Pacific Science Congress in and Dr. Clausen discussed principles 

Tokyo. He also attended the Interna- of forest genetics. Dr. Chiba showed 

tional Symposium on Plant Biosys- the modern Institute and the exten- 

tematics at Tokyo University, sive experiment grounds owned and 

arranged by the International Organi- operated by the Oji Paper Company, 
zation of Plant Biosystematics. While 


Amesz, Jan, and David C. Fork, Quenching Bjorkman, Olle, and Paul Holmgren, Photo- 

of chlorophyll fluorescence by quinones in synthetic adaptation to light intensity in 

algae and chloroplasts, Biochim. Biophys. plants native to shaded and exposed 

Acta, 1US, 97-107, 1967. habitats, Physiol. Plantarum, 19, 854-859, 

Amesz, J., see also Fork, David C. 1966. 

Anderson, Jan M., see also Fork, David C. Clausen, Jens, Stability of genetic charac- 

Bjorkman, Olle, Comparative studies of ters in Tragopogon species through 200 

photosynthesis and respiration in ecologi- years, Trans. Proc. Botan. Soc. Edinburgh, 

cal races, Brittonia, 18, 214-224, 1966. UO, 148-158, 1966. 



C, David C., and Jan Amesz, Energy 
ansfer between photosynthetic units of 
pigment system 1 in algae, Biocliim. 
phys. Acta. k:. 366-368, 1967. 
Fork. David C, Jan Amesz. and Jan M. An- 
derson, Light-induced reactions of chloro- 
phyll /> and P700 in intact plants and 
chloroplasi fragments, Brookhaven Syvnp, 
\. 19, 81-94, 1966. 
Fork, David C, and Yaroslav de Kouchkov- 
sky. The 518-m/i change and its relation 
the photosynthetic process, Photochem. 
PhotobioL, 5,609-619, 1966. 
Fork. David C, sec also Amesz, Jan. 

':.. C. S., Review of Molecular PJiysics 
in Photosynthesis, by Roderick K. Clayton, 
A . S ntist, 54, 332A-334A, 1966. 
French. C. S. (Inventor), and Charlton M. 
Lewis (Registered Patent Agent), Plane 
Table Plotter, a map making device, U.S. 
Patent =3,302,293, 1967. 

rich, C. S.. Chloroplast pigments, in 
Biochemistry of Cliloroplasts, Vol. I, 

edited by T. W. Goodwin, Academic Press, 

London, pp. 377-386, 1966. 
French, C. S., Review of The Cldorophylls, 

edited by Leo P. Vernon and Gilbert R. 

Seely, Science, 154, 1318, 1966. 
French, C. S., Die Photochemische Nutzung 

der sichbaren Strahlung durch zwei Licht- 

reaktionen der Photosynthese, Novo Acta 

Lcopoldina, N.F., 31, 169-187, 1966. 
French, C. S., see also Pickett, J. M. 

Hiesey, William M., Review of The Genetics 
of Colonizing Species, edited by H. G. 
Baker and Ledyard Stebbins, Quart. Rev. 
Biol., 41, 418-419, 1966. 

Holmgren, Paul, see also Bjorkman, Olle. 

de Kouchkovsky, Yaroslav, see also Fork, 
David C. 

Lewis, Charlton M., see also French, C. S. 

Pickett, J. M., and C. S. French, The action 
spectrum for blue-light-stimulated oxygen 
uptake in Chlorella, Proc. Natl. Acad. Sci. 
U.S., 57, 1587-1593, 1967. 


Amesz, Jan, Spectrophotometry measure- 
ments of reactions of photosynthetic inter- 
mediates, Seminar, Department of Bot- 
any, University of Illinois, Urbana, 
Illinois, October 11, 1966. 

Bjorkman, Olle, Photosynthetic adaptability 
to light in plants native to shaded and 
exposed environments, Seminar, Botany 
Department, University of California, 
Davis, California, December 6, 1966. 

Bjorkman, Olle, Adaptation of photosynthe- 
to light intensity in sun and shade 
ecotypes of Solidago virgaurea, Seminar, 
Department of Biological Sciences, Simon 
Praser University, Burnaby, B. C, Can- 
ada, January 26, 1967. 

Bjorkman, Olle, Ecological adaptation of 
photosynthesis, Seminar, Department of 
Biology, University of California, Santa 
Cruz, California, May 17, 1967. 

Brown, Jeanette S., Chlorophyll fluorescence 
emission of a diatom, Pacific Slope Bio- 
chemical Conference, University of Cali- 
fornia, Davis, California, June 17, 1967. 

Clausen, Jens, Biosystematic consequences 
of ecotypic and chromosomal differentia- 
tion, International Symposium on Plant 
Biosystematics, Tokyo, Japan, August 
25, 1966. 

Clausen, Jens, Species clusters in tree vege- 
tations around the Pacific, Eleventh Pa- 

cific Science Congress, Tokyo, Japan, 
August 27, 1966. 

Fork, David C, Studies on components in 
the electron-transport chain of photosyn- 
thesis, Seminar, Botany Department, Uni- 
versity of California, Davis, California, 
November 22, 1966. 

French, C. S., Forms of chlorophyll in algae 
and mutants, Pacific Slope Biochemical 
Conference, University of California, 
Davis, California, June 17, 1967. 

Hiesey, William M., Genetic vs. environ- 
mental effects on photosynthetic rate of 
clones of contrasting climatic races of 
Mimulus, American Society of Plant 
Physiologists, AIBS Meeting, College 
Park, Maryland, August 17, 1966. 

Hiesey, William M., Cytogenetic and eco- 
physiological studies in Mimulus, Seminar, 
Department of Biology, California Insti- 
tute of Technology, Pasadena, California, 
May 16, 1967. 

Michel-Wolwertz, M. R., The different 
chlorophylls extracted from plants, Pacific 
Slope Biochemical Conference, University 
of California, Davis, California, June 17, 

Nobs, Malcolm, Cytogenetics and climatic 
selection in Mimulus, Seminar, Depart- 
ment of Genetics, University of Califor- 
nia, Davis, California, October 27, 1966. 




Biochemical Investigations 

Staff: C. Stacy French, Director; Jean- 
ette S. Brown, David C. Fork; James 
H. C. Smith, Emeritus 

Carnegie Corporation Fellows: Jan 
Amesz, 1 Ulrich W. Heber 2 

Institution Research Fellows: Marie- 
Rose Michel-Wolwertz, 3 Jean-Marie 
Michel, 4 James M. Pickett 5 

Technical Assistants: Alan de Schwein- 
itz, 6 Marion A. Koerper, 7 Mark C. 
Lawrence, Suzanne Parmelee, 8 William 
T. Rhodes 9 

Part-time Laboratory Helpers: Christine 
M. Anderson, 10 Stephen J. Fulder 11 

Experimental Taxonomy 

Staff: Jens C. Clausen, Emeritus; Olle 
Bjorkman, William M. Hiesey, Mal- 
colm A. Nobs 

Institution Research Fellow: Eckard W. 
Gauhl 12 

Technical Assistants: Frank Nicholson, 
Pamela Radford 13 

Summer Research Assistants: Oakley 
Shields, 14 Stephen G. Wood 15 

Clerical Assistant: Marylee Eldredge 

Part-time Garden Helpers: Robert M. 

Baker, 16 James M. Barnes, 17 John M. 

Emmel, 18 Lawrence W. Hart, 19 Ralph 

R. Liddle, 20 John C. Nobes 21 

Accountant-Secretaries : Richard F. Gill, 22 
Clara K. Baker 23 

General Department Secretary: Lena R. 

Mechanical Engineer: Richard W. Hart 

Custodian: Jan Kowalik 

1 Through April 21, 1967. From Biophysical 

Laboratory, State University, Leiden, 
The Netherlands. 

2 From May 3, 1967. From Botanisches In- 

stitute, Universitat Dusseldorf, Bonn, Ger- 

3 From Jan. 4, 1967. From Centre de Re- 

cherches de Gorsem, Belgium. 

4 From Jan. 4, 1967. From Centre de Re- 

cherches de Gorsem, Belgium. 

5 From Oct. 1, 1965. From Department of 

Zoology, University of Texas, Austin. 

6 From Oct. 12, 1966, through May 30, 1967. 

7 To September 28, 1966. 

8 Aug. 22, 1966, through June 16, 1967. 

9 To September 30, 1966. 

10 Through August 19, 1966. 

11 July 5, 1966, through Aug. 26, 1966. 

12 From April 11, 1967. From Botanisches 

Institut, Universitat Frankfurt, Ger- 

13 October 24, 1966, through June 30, 1967. 

14 Through September 10, 1966. 

15 From June 15, 1967. 

16 To September 1, 1966. 

17 From June 8, 1967. 

18 Through July 29, 1966. 

19 Through August 15, 1966. 

20 May 9, 1967, through June 5, 1967. 

21 Through October 21, 1966. 

22 Through July 31, 1966. 

23 From August 6, 1966. 


Plate 1 . Growth of bean seedlings in air containing low versus normal oxygen concentration. 
Above. Plants at the beginning of a growth experiment (top row) and after 6 days in normal 
air with 21 % O2 (second row from the top). Be/ow.- Corresponding matched plants before and 
after growth in 2.5% 2 - 

Plate 1 

Department of Plant Biology 









Plate 2 

Department of Plant Biology 

Plate 2. Bean seedlings after growing for 17 days in normal air (left) and corresponding 

plants grown in 5% O2 (right). 

Mount Wilson and Palomar 

Operated by Carnegie Institution of Washington 
and California Institute of Technology 

Pasadena, California 

Horace W. Babcock 


Horace W. Babcock 

Carl D. Anderson 
Jesse L. Greenstein 
Eobert B. Leighton 

Allan E. Sandage 
Olin C. Wilson 

Carnegie Institution Year Book 66, 1966-1967 


Introduction 253 

Observing Conditions 254 

Solar Physics 255 

Solar magnetograms .... 255 
Oscillatory motions in the solar 

atmosphere 256 

High-resolution magnetograms . 258 

Cinematography 258 

Eclipse observations .... 259 

Coronal physics 259 

Continuous observations . . . 259 
Spectroscopic study of sunspots . 259 

Planets, the Moon, and Comets . . 260 

Mars 260 

Infrared photometry and colorim- 

etry of planets 260 

Differential infrared spectrometry 

of the Moon 261 

Daytime infrared . colorimetry 

program 261 

Comets 261 

Stellar Spectroscopy and Photom- 
etry 261 

White dwarfs ,261 

Dwarf K and M stars .... 262 
Photoelectric flux measurements 
at the center of stellar H and 

K lines 262 

Dwarf Me stars 263 

Dependence of rotation on age for 

stars of solar type . . . 263 
Stellar evolution and the conserva- 
tion of angular momentum . 264 
Structure and evolution of W 

Ursae Majoris stars . . . 264 

Faint blue stars 264 

A Bootis stars 265 

Peculiar A stars 265 

Stellar envelopes 265 

Binaries 266 

Helium abundance 267 

Stellar composition 268 

Variable stars 271 

Infrared stellar spectroscopy . . 272 
Infrared photometry .... 272 
Temperature scale 272 

Absolute Spectrophotometry . . . 273 

Star Clusters 273 

Color-magnitude diagrams . . 273 
Helium deficiency 274 

"Blue stragglers" 275 

NGC 6171 275 

Other globular clusters . . . 275 

Praesepe cluster 276 

Cluster Stars with Anomalous 

Spectra 276 

Interstellar Gas and Gaseous Nebu- 
lae 276 

Interstellar absorption lines . . 276 

Nova envelopes 277 

Crab nebula 277 

Extinction in emission nebulae . 278 

High-latitude planetary nebulae . 278 

Infrared Sky Survey 279 

Galactic Structure 279 

Galaxies 280 

Distance modulus of NGC 2403 . 280 

Variables in dwarf galaxies . . 281 

Seyfert galaxies 282 

Compact galaxies 282 

Counter-jet in M 87 . . . .284 

Ultrafaint blue stars .... 284 
Catalogue of galaxies and clusters 

of galaxies 285 

Statistical analysis 285 

Classification of clusters of gal- 
axies 286 

Special cluster 286 

Supernovae 286 

Radio Galaxies 287 

Cosmic-ray sources 287 

Redshifts of radio galaxies . . 287 

Energy distribution .... 287 

Direct photography .... 287 

X-Ray Sources 288 

Scorpius X-l 288 

Sources in Cygnus 289 

Quasi-Stellar Sources .... 289 

Spectroscopy 289 

Parkes source 0237-23 .... 289 
Spectral energy distribution . .291 
Optical survey for radio-quiet 
quasars or quasi-stellar ob- 
jects 291 

Space distribution 293 

Optical variability 293 

Infrared emission 294 

Theoretical Studies 294 

Heating of the solar atmosphere 294 
QSO spectra 294 

Observational cosmology . . . 294 

Star formation 295 

Guest Investigators 295 

00-Inch Photometric Telescope . . 302 

Optical design 302 

Optical work 302 

Mechanical parts 302 

Site 303 

Instrumentation 303 

Image-tube spectrograph . . . 303 

Multichannel spectrometer . . 303 

Prime-focus corrector . 

Astro-electronics laboratory 

Other instrumentation for 
large reflectors 

Solar equatorial telescope . 

Photographic laboratory . 
Site Investigation .... 

Image, motion at sites in Chi 

Microthermal data . 


Staff and Organization . 


. 303 

. 304 

. 305 

. 306 

. 306 

. 306 




It is worthy of note that an increas- achieved during the past year. The 

ing convergence is apparent between central region of our Galaxy, seen 

the fields of stellar physics and solar from our position far out toward the 

physics. The study of the sun, the edge, is heavily obscured by clouds of 

only star that can be observed in de- dust and gas that are concentrated 

tail, has provided facts about such near the plane of the spiral system. In 

phenomena as sunspots, flares, the visible light it has been impossible to 

chromosphere and corona, rotation, penetrate these clouds and to discern 

internal circulation and mixing, mag- anything of the structure of the cen- 

netic fields, the solar magnetic cycle, tral region. But in longer wave- 

and transfer of angular momentum, lengths the Galaxy is far more trans- 

both internal and external. As im- parent, and some years ago radio 

proved tools and techniques are ap- astronomers detected a strong radio 

plied, and as knowledge of the evolu- source, Sagittarius A, in the general 

tion of stars accumulates, we find that direction of the center. Recently, Eric 

stellar analogues of solar phenomena Becklin, a graduate student at Cal- 

are becoming the subjects of specific tech, using an optical telescope with 

researches. The example of flare a special detector, discovered that the 

stars is well known. The spectro- galactic center is an emitter of far 

scopic study of chromospheres of infrared radiation. Subsequently, he 

dwarf M and K stars has been fruit- and Dr. G. Neugebauer, working in 

ful, and efforts are under way to de- the wavelength region 1 to 3 microns 

tect the stellar analogues of the solar (some 2 to 6 times the wavelength of 

magnetic cycle. Convection zones, in- visible light), have mapped an area 

ternal circulation, and mixing are ac- about equal in apparent size to that 

tive subjects. It was the study of the of the moon in which they have 

ever-expanding magnetic patterns on plotted the observed infrared radia- 

the surface of the sun, scarcely a tion-intensity pattern including some 

decade ago, that gave proof of the six intensity peaks of differing shape 

continuous outflow of plasma (ion- and sharpness; the most intense peak 

ized gas with associated magnetic is coincident with Sagittarius A. This 

fields) from the sun — an effect that is the first evidence on the structure 

has since acquired the convenient de- and brightness of the nucleus of the 

scriptive name of solar wind. Loss Galaxy, a region that one suspects 

or transfer of angular momentum by may have been the site of extraordi- 

a star, with its resultant effects on nary physical activity in the early 

stellar rotation and "equatorial ac- stages of the development of our 

celeration," is a subject near the fore- stellar system. 

front of research. The origin of the At various observatories, spectra 

planets of the solar system has long have now been obtained for approxi- 

been a subject for speculation or for mately 120 quasars, including both 

elaborately structured theories re- radio-emissive (QSS) and radio-quiet 

mote from the opportunity for deci- (QSO) types. Data on photometry 

sive observational tests, and this is a and color exist for a comparable num- 

situation that is slowly but surely ber. QSO candidates can be distin- 

changing. guished by their colors, and new ones 

In the field of galactic structure, can be discovered almost at will. Data 

a significant new result has been by Luyten and Sandage suggest that 




the total number of QSOs over the 
entire sky as faint as magnitude 
19.7 (blue) is about 100,000. Quasars 
are receiving the best efforts of a 
growing number of observational as- 
tronomers and theoreticians at many 
places, and several significant find- 
ings have emerged during the past 
year from the work of the staff and 
visitors at Mount Wilson and Palo- 
mar. Among these findings, a few 
have been selected for brief mention. 

Schmidt has investigated the space 
distribution of quasars, and he esti- 
mates that their "local" space den- 
sity, including those observed as radio 
sources, is 1 per 10 s cubic megapar- 
secs. From a study of 33 brighter 
sources from the 3C Catalog for 
which redshifts are known, he finds 
that the space density of quasars 
must be increasing with redshift. The 
excess density factor has been placed 
at 15 for redshift of 0.5 and at about 
100 for redshift of 1. Such a result 
has great implications for cosmology. 

New redshifts have been observed 
for a number of QSSs and QSOs; the 
largest yet found is for the Parkes 
radio source 0237-23, a candidate 
suggested by Bolton and first ob- 
served optically by Arp. Later spec- 
troscopic studies by Greenstein and 

Schmidt show that there are, in ad- 
dition to the emission lines for which 
AX/X = z = 2.22, two sets of ab- 
sorption lines giving z = 2.202 and 
1.956, respectively. Lines from ex- 
cited states occur, suggesting rela- 
tively high density and some colli- 
sional excitation. Interpretive studies 
based on spectroscopy of other qua- 
sars showing absorption features 
have provided some preliminary indi- 
cation of temperature and density in 
these remarkable objects. 

Practically all quasars vary in lu- 
minosity, and two that are highly 
variable have been investigated in 
detail by Oke, who measured absolute 
energy distributions in their spectra. 
The source 3C 279 varied as much as 
0.25 magnitude per day and by nearly 
2 magnitudes during the year. Oke's 
new measurements confirm a recent 
finding by Sandage, Westphal, and 
Strittmatter that it is the continuous 
spectrum that varies, while the emis- 
sion lines remain constant. 

These and other observational find- 
ings are gradually providing the con- 
ditions that circumscribe the de- 
velopment of a physical model for 
these bizarre objects that were un- 
known only seven years ago. 


Rainfall at Mount Wilson was 55.59 
inches, as compared with the 63-year 
mean of 35.33 inches. Total snowfall 
was 68 inches. The temperature was 
highest (90°F) on August 5 and low- 
est (19°F) on April 12 and 20. The 
heaviest snowfall in one day was 10 
inches on April 11, and precipitation 
was registered on 17 days of that 

At Palomar Mountain 45.53 inches 
of rain fell, of which 24 inches oc- 
curred during a 5-day period begin- 

ning on December 3, 1966. Total 
snowfall was 37 inches. 

Observations with the 100-inch and 
200-inch telescopes were scheduled on 
363 nights of the year, and no lost 

TABLE. 1. 














time because of mechanical failures scopes, as limited by sky conditions, 
was reported. The time actually used is shown in Table 1. The total number 
for observations with the major tele- of nighttime hours was 3800. 


Routine solar observations were 
made by Thomas Cragg, Merwyn 
Utter, and Robert Howard on 307 
days. The records of various kinds 
made between July 1, 1966, and June 
30, 1967, were as follows: 

Direct photographs 


Ha spectroheliograms, 30-foot 



K2 spectroheliograms, 30-foot 





Sunspot drawings 


Magnetic classifications of sunspot 
groups were made visually on 226 
days during the year. 

Alan Title assembled and arranged 
for publication a series of the best 
spectroheliograms that had been ob- 
tained in the course of the six-year 
program of the study of solar velocity 
fields initiated by Leighton. 

Solar Magnetograms 

The new magnetograph, image 
scanning, and data systems are in 
operation at the 150-foot tower. Dur- 
ing a magnetogram scan, data are re- 
corded on magnetic tape. The data 
are analyzed by means of the IBM 
7094 computer at the California In- 
stitute of Technology. Partial support 
of this program is provided by the 
Office of Naval Research. 

From the velocity data the differ- 
ential rotation of the sun is deter- 
mined along with the limb redshift. 
This is done by means of a least- 
squares solution for the coefficients of 
a sine-squared, sine-fourth-power ex- 
pansion of the solar latitude. For the 
standard magnetogram with 17 arc- 
second resolution, there are about 
11,000 data points on the solar disk, 

each with a very accurate line-of- 
sight velocity measurement. Thus the 
differential rotation of the scan can 
be determined with precision. It is 
planned to continue observing the 
solar differential rotation throughout 
the current solar cycle. J. W. Harvey 
of the High Altitude Observatory is 
collaborating with Howard on this 

Another quantity that is analyzed 
is the integrated magnetic field over 
various parts of the solar disk. Dur- 
ing this year the magnetic fields in 
both polar regions have been meas- 
ured and prove to be quite weak — 
about 1 gauss negative. The magnetic 
flux in each polar region is about 10 20 
maxwells. It is expected that contin- 
ued observations of these quantities 
and the magnetic flux in other lati- 
tude zones will provide valuable data 
about large-scale solar activity. 

Solar magnetograms are computer- 
plotted for each day's observation. 
Figure 1 shows an example of a full- 
disk magnetogram. These magneto- 
grams are being published monthly 
along with Mount Wilson sunspot 
magnetic classifications and solar 
data from other observatories in the 
publication Solar -Geophysical Data 
issued by the Environmental Science 
Services Administration of the U. S. 
Department of Commerce, and ob- 
tainable through the Superintendent 
of Documents, Government Printing 

"Dopplergrams" are plotted from 
the velocity data after subtracting 
the background solar rotation. An ad- 
ditional Dopplergram is plotted after 
the data have been averaged over 
area in order to smooth out some of 
the small-scale features and to dis- 




DATE (RS.T.) 6/18/67 21 
TIME(UT) 22.29-23.78 


Fig. 1 . This full disk magnetogram was drawn by a computer-controlled mechanical plotter from 
data obtained at the 1 50-fool solar tower. Solid contour lines enclose regions of positive field and 
dotted contour lines enclose regions of negative field. This map shows many bipolar regions in 
various stages of development. All magnetic activity on the surface of the sun first appears as 
compact, bipolar regions. As these magnetic regions age, they expand in characteristically in- 
clined patterns. The remnants of the trailing parts drift slowly toward the poles to give the sun its 
apparent dipolar magnetic field. 

close large-scale low-amplitude mo- 
tions on the solar surface. Both of 
these Dopplergram programs were 
started near the end of the year and 
no results are yet available. 

Oscillatory Motions in the Solar 

Howard has completed an exten- 
sive study of oscillatory motions in 
the solar atmosphere. He used the 
solar magnetograph, which can be 
adapted as a very sensitive detector 

of Doppler shifts of a spectrum line. 
Using the line A5250 (Fe I) and ob- 
serving at fixed points on the sun's 
image, he and the other solar observ- 
ers obtained nearly 250 hours of ob- 
servations with apertures of various 
sizes. The amplitudes of the well- 
known 300-second oscillations are 
about 25% weaker in regions where 
the magnetic field is greater than 80 
gauss as compared to regions where 
the field is less than 10 gauss. No dif- 
ference in the frequency spectrum 



of the oscillations could be found 
between strong-field and field-free re- 
gions, thus indicating that the oscilla- 
tions are not basically of hydromag- 
netic origin. Howard suggests that 
the oscillating motion occurs only 
where the magnetic field is effectively 
absent, and that the reduction in 
amplitude in a strong-field region is 
the effect seen when a certain frac- 
tion of the area within the aperture 
does not contribute to the oscillation. 
There is evidence to support the view 
that at least a large part of the mag- 
netic flux observed on the solar sur- 
face exists as small isolated bundles 
of lines of force within which the field 
is high — perhaps 500 gauss. 

One of the features of the new 
magnetograph exit-slit assembly is 
provision for observing two spectrum 
lines simultaneously and for obtain- 
ing magnetic and velocity signals 
from each. Although the correspond- 
ing portion of the data system has 
not yet been built, it is possible to re- 
cord in an analogue fashion the ve- 
locity signals from two lines simul- 
taneously. This was done as a part of 
the oscillation program, and signifi- 
cant phase differences amounting at 
times to as much as 30 seconds were 
found upon comparison of the signals 
from two lines formed at depths dif- 
fering by perhaps 100 km in the solar 
atmosphere. The higher-level line of- 
ten showed oscillations of a slightly 
higher frequency. 

A short-period oscillation is found 
superposed on the 300-second oscilla- 
tion. These "SPOs" come in bursts 
that last for a minute or two and have 
peak-to-peak amplitudes of about 
0.05-0.10 km/sec. Their periods fall 
in the range 1-5 seconds. All at- 
tempts to explain them as seeing ef- 
fects or instrument effects have 
failed. The horizontal scale of the 
SPOs is smaller than that of the 300- 
second oscillations, and they are more 
nearly isotropic. They do not repre- 

sent a high-frequency tail of the 300- 
second oscillation, but appear to be a 
completely separate phenomenon and 
are not markedly dependent in their 
occurrence or strength upon phase or 
amplitude of the longer-period oscil- 
lation. These observations were re- 
corded with a digitizing interval too 
long to provide power spectra of the 
SPOs, but it is clear from their ap- 
pearance that the SPOs have a much 
wider range of frequency than do the 
slower (300-second) oscillations. The 
amplitude of the SPOs in the solar 
atmosphere must be large, and they 
must contribute greatly to the non- 
radiative energy flux. It is suggested 
that they represent a large "micro- 
turbulence" line-broadening effect. 

In an attempt to establish that the 
SPOs were a solar phenomenon and 
not some subtle effect of daytime see- 
ing, Dr. William C. Livingston of the 
Kitt Peak National Observatory col- 
laborated with Howard in a joint ef- 
fort to observe the bursts simultane- 
ously at Kitt Peak and Mount Wilson. 
On June 2 both observers set the 
aperture of their magnetographs on 
the same spot in the solar atmosphere 
and observed the variation in line-of- 
sight velocity for several hours. A 
comparison of the two observations 
indicated that the SPO bursts came 
always at the same time at the two 
observatories. Even some individual 
SPOs could be identified on both 
records. This is strong evidence that 
the oscillations are solar in origin. 

Dr. J. M. Wilcox of the Space Sci- 
ences Laboratory of the University 
of California at Berkeley has been 
cooperating with Howard in a study 
of velocity and magnetic fields on the 
solar surface, using the new scanning 
and data system of the 150-foot tower 
telescope. The observations consisted 
of repeated scans over a straight line 
on the solar disk. The data have been 
analyzed to reconstruct the structure 
and time dependence of the velocity 



fields, magnetic fields, and brightness 
at many points along the scan line. 
First results indicate that the dis- 
tances over which 5-minute oscilla- 
tions are in phase may be as great as 
30.000 km, contrary to previous esti- 
mates of oOOO- or 4000-km cell sizes 
for such oscillations. 

Wilcox and Howard are also co- 
operating on a study of the differen- 
tial solar rotation as determined from 
autocorrelation analysis of magnetic- 
field data from 1959 to the present. 
When the magnetic data are digitized 
and an autocorrelation is run for one 
latitude,, the correlation peak due to 
rotation is. in general, easily identi- 
fied. Thus the latitude dependence of 
the rotation of magnetic features may 
be studied. The long interval over 
which data are available will provide 
a clue to possible time dependence 
of solar differential rotation. It is 
planned that the autocorrelation tech- 
nique applied in this way will also 
be used in a study of the large-scale 
features of solar activity and of the 
solar magnetic-field distribution. 

High-Resolution Magnetograms 

Rust has developed computer pro- 
grams for analysis of high-resolution 
(2-10 arc-second) magnetograms of 
small regions on the solar disk. These 
"fine-scan" programs extend the 
rapid data-gathering capacity of the 
magnetograph, guider, and tape-re- 
cording system in the 150-foot tower 
to a wider range of solar studies than 
was possible with programs designed 
for whole-disk scanning only. De- 
tailed maps of the velocity and mag- 
netic fields near Ha filaments, and in 
active regions, will be produced en- 
tirely by computer techniques from 
data obtained during the summer of 
1967. Some sequential magnetic scans 
have already been obtained in an ac- 
tive region during a small flare. The 
scans have a 10-minute time resolu- 
tion and a 10-arc-sec spatial resolu- 

tion. Rust hopes to coordinate obser- 
vations of this kind with flare patrol 
sequences made by Lockheed observ- 
ers in order to examine the connec- 
tion between changes in the photo- 
spheric magnetic fields and the 
sudden energy releases occurring in 
flares. In another application of the 
fine-scan programs, Rust and Howard 
are studying the magnetic fields in 
the sun's polar regions. 


A new Ha- filter was installed at the 
150-foot tower and used in periods of 
good seeing. Assisted by Lawrence 
Anderson and Terence Allen (Caltech 
undergraduates), Zirin obtained a 
number of interesting films of various 
solar phenomena. One of the principal 
goals was to study oscillatory motion 
in Ha. Another goal has been to un- 
derstand the nature of the chromo- 
sphere as disclosed by limb observa- 
tions. By using a narrow-band 
suppressor plate, Zirin showed that 
he had been wrong in the interpreta- 
tion of the "second limb" on filter- 
grams, and that Simon and White had 
been correct in their contention that 
the inner limb is the photosphere seen 
in filter side bands. But he was able 
to confirm another "inner limb" — 
that of the low chromosphere — ex- 
tending from zero to 1500 km above 
the photosphere. This appears as a 
dark band crossing active regions at 
the limb. Because the chromospheric 
emission profile broadens with height, 
the dark band is most apparent in 
pictures in the wing of Ha. The 
broadening of Ha in the upper chro- 
mosphere may be connected with the 
high velocities of the oscillating ele- 

In addition to the cinematography 
of quiet sun features, a number of 
films of solar flares were obtained. 
Outstanding among these were high- 
resolution records of the class 3 flare 
of August 28, 1966. One of the inter- 



esting features is the upward erup- 
tion of a filament preceding the flare, 
a type of phenomenon reported in 
other instances by the Lockheed Solar 
Observatory. The X-ray and radio 
emissions do not begin until the flare 
brightening occurs, some time after 
the prominence erupts. 

Eclipse Observations 

The predicted infrared coronal 
lines Al.43 p [Si X] and A3.03 /x [Mg 
VIII] were successfully observed by 
Munch, Neugebauer, and McCammon 
during the total solar eclipse of No- 
vember 12, 1966, from the airplane 
"Galileo" of the National Aeronautics 
and Space Administration. For this 
purpose, the 50-cm Ebert spectrom- 
eter used in the past for observa- 
tions of infrared stellar spectra was 
fitted with two PbS detectors at suit- 
able exit slits and attached to a fixed 
F5 parabolic reflector. Solar light was 
delivered to the instrument by a gyro- 
stabilized heliostat through a special 
5.5-inch CaF 2 window. On the basis 
of the measured wavelengths and in- 
tensities of the lines, it appears that 
they should be observable from high- 
altitude ground stations. The absolute 
intensity of the [Si X] line is about 
one tenth that of the strongest coro- 
nal lines in the photographic region 
and does not correspond to the 
strength of the permitted silicon spec- 
trum in the rocket ultraviolet. 

Coronal Physics 

Zirin worked on a new interpreta- 
tion of the excitation of coronal lines. 
It has been known that coronal lines 
may be excited by photospheric radi- 
ation or by electronic collisions, but 
the significance of the variation in 
this process from line to line has not 
been clear. There are a number of 
coronal lines (for example, Fe XV 
A7060, Fe XIII A10798) for which 
photospheric excitation is impossible 
or very small, either because the 

transition from the ground state is 
completely forbidden or because the 
required energy is too great. These 
lines will therefore be excited by col- 
lisions only, and their intensity will 
vary as the square of the electron 
density N e . They will thus have very 
low intensity at low coronal density. 
Ordinary coronal lines may be excited 
by photospheric radiation and will 
vary as N e at low densities, so they 
will be relatively stronger. The effect 
is marked for the ratio of Fe XIV 
A5303 to Fe XV A7060. The latter 
has always appeared anomalously 
weak compared to the strong Fe XV 
resonance line at 283 A; we see now 
that it can have appreciable density 
only for N e > 2 X 10 9 . 

Continuous Observations 

For continuous study of the chro- 
mospheric network and of active re- 
gions, Leighton and Zirin proposed 
some years ago the establishment of 
one or two temporary observing sta- 
tions well spaced in longitude around 
the world. With a grant from the Na- 
tional Science Foundation, Zirin has 
now set up such a station at Tel 
Aviv University, Israel, with a small 
telescopic pulse camera and Ha filter. 
Its operation will be the responsibility 
of the Tel Aviv staff. 

Spectroscopic Study of Sunspots 
Makita analyzed the green system 
of a TiO band in the spectrograms of 
a large sunspot which Dr. Olson of 
the University of Illinois obtained 
with the 150-foot tower telescope in 
September 1966. Only approximate 
equivalent widths were measured, 
since the lines were weak and most 
of them were disturbed by blending. 
Coarse analysis, plotting the line 
strength as a function of the rota- 
tional quantum number, gave the 
loci of the maximum and minimum 
widths. The most probable rota- 
tional temperature thus obtained was 



Ma rs 

For the 19G7 opposition of Mars 
an extensive series of spectroscopic 
observations had been planned. Un- 
usually poor weather conditions, 
however, permitted the completion of 
only about one quarter of the pro- 
gram. Using the 100-inch coude scan- 
ner and a Fabry-Perot interferom- 
eter with a free spectral range of 
1.96 A and a finesse of 75, Munch has 
measured photoelectrical^ the inten- 
sities of lines in the weak C0 2 band 
5r 3 at AS690 A with a precision higher 
than has been previously available by 
photographic means. A complete re- 
duction of the data is now in progress. 

In the spectral range between 2.0 
and 2.4 a, Munch and Neugebauer at- 
tempted to obtain spectra of different 
areas on the surface of the planet. 
For this purpose there was con- 
structed a chopping system that com- 
pares the spectra of two areas within 
a scanning spectral channel and also 
in a broadband monitor channel. 
Since the defining apertures are 
fixed at the entrance plane of the 
100-inch coude spectrograph, an all- 
reflecting field rotator had to be made 
also. The spectral resolution used was 
about 40 A, and the angular resolu- 
tion could be as small as 1 arc-sec, 
although poor seeing conditions did 
not allow full advantage to be taken 
of it. The data obtained have not yet 
been processed, but preliminary ex- 
amination shows the existence of con- 
spicuous broadband differences in 
planetary surface brightness. Using 
the same chopping scheme, difference 
spectra between the center and the 
limb of the planet were attempted; 
these should effectively cancel the ef- 
fects of telluric absorption. 

Beginning in January 1967, inter- 
ference-filter observations in the re- 
gion of the 2-ji bands of C0 2 have 

been made periodically by T. Hilge- 
man, a graduate student at the Cali- 
fornia Institute of Technology, with 
the 24-inch telescope at Mount Wil- 
son. The effects of telluric absorption 
are calibrated by lunar observations 
at various air masses. The purpose of 
the observations is to determine the 
degree of constancy of the C0 2 con- 
tent of the Martian atmosphere. The 
work will be continued as late in the 
year as possible. 

Infrared Photometry and Colorimetry 
of Planets 

Murray observed Mars successfully 
in the evening twilight with the 200- 
inch on May 2 and 3, 1967. Other 
scheduled attempts in April and May 
were unsuccessful due to weather 
conditions. A maximum spatial-reso- 
lution map of the 8-13-/X brightness 
temperature was obtained on May 2, 
and similar maps in the 9-, 11-, and 
13-/X intervals were acquired the fol- 
lowing night. Venus was observed at 
similar air mass for comparison in 
each case. These results constitute the 
most detailed information on the 
surface temperatures of Mars yet 

Murray observed the 8-13-/* emis- 
sion of the shadow of Jupiter's satel- 
lite, Io, as it crossed the planet's disk 
during the evening twilight of May 
2. No anomalous emission was ob- 
served, the upper limit being esti- 
mated at a flux not greater than twice 
that of the disk itself over an area 
equal to the geometric size of the 
shadow. The origin of the earlier ob- 
served shadow enhancements remains 

A maximum nighttime surface 
temperature of Mercury of 155°K 
was estimated by Murray from his 
earlier observations. 



Differential Infrared Spectrometry 
of the Moon 

A. F. H. Goetz completed his doc- 
toral thesis in which he achieved 
greater than an order-of-magnitude 
improvement in the detection of 
spectral variations of emissivity in 
the 8-13-/A region between different 
localities on the lunar surface. The 
24-inch telescope was used. A detec- 
tivity approaching 0.2% in relative 
emissivity was obtained over most of 
the wavelength interval and two new 
anomalous areas were discovered. 
Previously reported anomalies were 
not confirmed. Such emissivity varia- 
tions probably are indicative of dif- 
ferences in age and possibly also in 
composition of lunar surface mate- 
rials; they constitute an important 
new source of information about 

An attempt by Goetz to extend the 
method of differential infrared spec- 
trometry to the observations of Mars 
with the 200-inch telescope was un- 
successful because of the strong sur- 
face variations of intensity charac- 
teristic of planetary images. 

Daytime Infrared Colorimetry 

Because the terrestrial planets are 

such strong infrared emitters, and be- 
cause whole-disk colorimetry at 9, 11, 
and 13 ji potentially can provide im- 
portant information not otherwise 
available about these objects, an ef- 
fort has been under way for over a 
year by Murray and students to de- 
velop a suitable colorimetric tech- 
nique involving direct comparison 
with the sun. One major problem has 
been solved — that of rapid intercom- 
parisons with the sun in order to min- 
imize the effects of rapidly varying 
daytime infrared transmission. The 
solution involves the use of a 1-inch 
aperture coelostat system mounted on 
the 24-inch telescope. With this sys- 
tem, the solar image can be rapidly 
brought into the photometer field of 
view for signal comparison without 
moving the telescope off the planetary 
object being observed. 


A comet, 1966e, was discovered 
with the 48-inch schmidt telescope by 
Rudnicki. The periodic comet Grigg- 
Skjellerup 1966f was rediscovered, 
also with the 48-inch schmidt by 
Rudnicki, upon the request of Dr. G. 
Sitarski of the Polish Academy of 


White Dwarfs 

A third list of confirmed white 
dwarfs, with colors and spectra, has 
been prepared by Eggen and Green- 
stein, bringing to 202 the number of 
those with relatively complete data. 
The space motions show that relics of 
all population types are present. They 
have U, V, W components for 132 
stars; one half of the stars have 
tangential velocities less than 33 km/ 
sec. The tangential velocities of 22% 
are less than 20 km/sec, and 10% 

are greater than 100 km/sec. In spite 
of the selection effect arising from 
proper motion, halo stars make up 
only 10%. The peculiar spectra occur 
very often in the higher-velocity 

Special efforts have been made to 
find red degenerate stars, types DG, 
DK, or DM, or red DC stars. The 
success of this effort is still small; 
only 7 very red stars are known with 
confirmed peculiar spectra, of which 
three have known parallax. The anom-, 



alously high luminosities of this 

group suggest that our discovery 
technique is still poor; the faintest 
new object, LP 658-2, has about the 
same luminosity as do Wolf 489 and 
van Biesbroeck 11 (about +15.5 = 
My). If the colors are interpreted as 
from a blackbody continuum, the 
radii derived are large, ranging up- 
ward from 0.01 Rq for G38-27 and 
W489, to 0.10 Kg for G5-28 and 
H67-138. A possibility exists that 
these large red stars are objects of 
small mass, contracting past the main 
sequence (although no spectral anom- 
alies, such as emission lines, are 
noted). The paucity of red degener- 
ate stars might become understand- 
able if the theory of freezing of the 
nucleons into a solid lattice (Vila, 
Van Horn) should predict a low spe- 
cific heat at the low interior tempera- 
ture of these highly evolved stars. 

Greenstein and Miss Trimble have 
measured the Einstein gravitational 
redshift of a group of 53 white 
dwarfs. Lack of information on the 
true radial velocity makes the result 
statistical, as does the low accuracy 
of the velocities. Thirty-seven stars 
with hydrogen lines gave a mean 
systematic redshift (K-term) of 
— 66 km/sec. The helium-rich stars 
have a lower K-term and systematic 
differences from line to line, caused 
by pressure shifts. If an approximate 
temperature scale is derived from 
UBV colors, the median radius is 
0.0107 Ro. The mass derived is about 
0.98 uto; this surprisingly high value 
is reduced to 0.86 n\o for a helium 
core and 0.71 ttl© for an iron core. 
These masses are appreciably higher 
than those derived heretofore. 

Dwarf K and M Stars 

Spectroscopic observation of nu- 
merous dwarf K and M stars, mostly 
from the catalogs of Vyssotsky et al., 
has been completed by O. C. Wilson. 
The plates have all been measured 

for radial velocity and this material is 
now in press. It turns out that the 
probable error of a single observa- 
tion, defined as one plate measured 
by two individuals, is ±1.8 km/sec, 
and the overall agreement with the 
system of the General Catalogue is 

Careful eye estimates of the in- 
tensity of the chromospheric Ca II 
emissions have been made with a 
comparator and certain standard 
stars. Internal consistency checks in- 
dicate that these estimates are suffi- 
ciently accurate to yield a reliable 
frequency distribution of the line 
strengths. This work is not yet com- 
pleted, however, and is being con- 

Photoelectric Flux Measurements at 
the Center of Stellar H and K Lines 

The procedure involved in this 
work with the coude scanner of the 
100 -inch telescope was described 
briefly in Year Book 65 (p. 141). As- 
sessment of the first year's accumula- 
tion of data is now nearly complete. 
During this period, observation was 
restricted almost entirely to a sample 
of about 140 F2-G2 stars from the 
catalog of Stromgren and Perry, in- 
cluding all of those objects in which 
bright H and K components have been 
seen on spectrograms of 10 A/mm. 

The average probable error of a 
single measurement of flux in either 
the H or the K line is found to be 
±1.8%. This is fairly satisfactory 
but is still more than twice as large 
as the probable error expected on the 
basis of statistics alone. It is hoped 
that the observational accuracy can 
be improved. In any event, the ac- 
curacy is quite adequate to show the 
general characteristics of chromo- 
spheric emission throughout the 
main-sequence band in the portion of 
the IT-R diagram under investigation. 
One result of considerable interest 
is the ratio of the chromospheric H 


and K components. It turns out that, period. It is tentatively concluded, 

for the stronger emissions, the K therefore, that (1) periodic or quasi- 

intensity is about 25% greater than periodic light variations are found in 

that of H. Presumably this means some dMe but not dM stars, and (2) 

that the gas masses in the stellar these variations are not related to 

chromospheres are not optically thick binary characteristics. Rejecting vol- 

in the Ca II lines. ume pulsation and eclipses as mech- 

The first year's observations pro- anisms to explain the light variabil- 

vide first-epoch values of chromo- ity, it seems that the most promising 

spheric flux for the emission-line model appears to be the rotational 

objects. It is proposed to continue modulation of a star with a nonuni- 

observing these in order to see form distribution of surface bright- 

whether long-term variations (ana- ness. 

logues of the solar cycle) can be de- ^ 7 - „ , ,. A , 

tected. In addition, it is hoped to ex- Dependence of Rotation on Age for 

tend this investigation soon to giants btars of boiar l ype 

and supergiants. The rotational velocities of a large 

sample of field stars with spectral 

Dwarf Me Stars types between F2 IV, V, and G3 IV, 

Following P. F. Chugainov's dis- V have been obtained by Kraft from 
covery of periodic light variability in coude spectrograms of resolution 
Popper's flare star, HDE 234677, about 6 km/sec. It is shown that the 
Krzeminski and Kraft have searched average rotational velocity is higher 
for similar variations in brightness among those with Ca II emission than 
among emission-line and nonemis- among those without. Since there is 
sion-line stars in the range of spectral strong evidence from the work of 
type dK7 to dM3.5. Six emission-line Wilson that stars with Ca II emission 
and three nonemission-line objects are younger on the average than most 
were observed extensively. It was solar-type stars, a picture is advanced 
found that: (1) Three emission-line in which rotation declines with ad- 
objects are certainly variable; Chu- vancing age. Kraft proposes that 
gainov's period (P = 3<?83) for HDE magnetically coupled winds, similar 
234677 is confirmed but the light am- to the solar wind, are responsible for 
plitude is now smaller. The other two this deceleration even after the star 
objects are probably periodic or takes up a position on the main se- 
quasi-periodic variables; BD + 34° 106 quence. The winds are presumably 
has a period of 3^92 and AC + 31° generated only in stars that have sub- 
70565 has a period of 4 < ?65. The re- surface hydrogen convection zones 
maining emission-line stars showed and concomitant chromospheres. 
no variability in excess of 0^03. (2) Estimates of rotational velocity 
Nonemission-line stars had no varia- among solar-type stars in the Hyades 
bility in excess of the same limit. and Pleiades that confirm this general 

Spectroscopic observations indicate picture were also obtained. The mean 

that BD+34°106 and HDE 234677 rotation in the Pleiades turns out to 

are both double-line spectroscopic bi- be the highest of all groups considered 

naries; AC + 31°70565 shows no ra- and averages about 40 km/sec for 

dial velocity variation. The orbital stars of mass 1.20 nio. The time scale 

period of BD + 34°106 has not been for the reduction by a factor of 2 of 

found, but the orbital period of the rotational velocities of Pleiades 

HDE 234677 is 5<?981, a number in- stars is equal to the age of the Hyades, 

commensurate with the photometric about 4 X 10 s years. 



The results do not bear directly on 
Dicke's suggestion that the solar wind 

decelerates only the outer convection 
zone of the sun and, thus, that the in- 
terior retains a large fraction of its 
"primordial" angular momentum. 
This is because the derived time scale 
refers to the conditions of wind den- 
sity and magnetic-held strength that 
were presumably appropriate when 
the sun was as young as the Pleiades 
stars, not to the present values of 
these quantities. The results do indi- 
cate, however, that solar-type stars 
rotate slowly, in comparison to more 
massive main-sequence stars, as a 
result of the deceleration produced 
by winds. 

Stellar Evolution and the Conserva- 
tion of Angular Momentum 

Kraft has given observational evi- 
dence that whenever an evolving star 
encounters regions of the H-R dia- 
gram where subsurface convection 
zones play a role in the internal struc- 
ture, rotational deceleration takes 
place as a result of magnetically 
coupled winds. These findings apply 
to contracting T Tauri variables, 
main-sequence stars below F5 V, and 
for post-main-sequence red giants be- 
tween about GO III and G5 III. The 
rate of this deceleration, dJ/dt, can 
be estimated from the difference in 
rotational velocity of two points on 
an appropriate evolutionary track, 
and the time elapsed between occupa- 
tion of the two points. Using the re- 
cent evolutionary calculations of 
Iben, Kraft is able to show that, very 
roughly, dJ/dt = k (L K ) 17 , where L K 
is the absolute rate of emission of 
energy in the K2 emission. Thus the 
rate of deceleration increases ap- 
proximately as the 1.7 power of the 
strength of the chromospheric emis- 
sion. Qualitatively this means that 
the torque exerted by the wind in- 
creases as the activity of the chro- 
mospheric regions increases ; this, in- 

deed, is almost self-evident when we 
consider the sun's magnetic activity 
and the solar wind. 

Structure and Evolution of W Ursae 
Majoris Stars 

Current ideas on the formation, 
structure, and evolution of the main- 
sequence contact-binaries of type W 
UMa were reviewed by Kraft. It was 
shown from relative star counts that 
W UMa stars cannot now be trans- 
ferring mass from one component to 
the other, and cannot, therefore, have 
evolved from close but not contact bi- 
naries satisfying the mass-luminosity 
relation. The possible formation and 
evolution of Lucy's contact configura- 
tion were considered. Kraft suggests 
that such a configuration evolves on 
the nuclear time scale of the primary 
into a U Geminorum binary; losses 
of angular momentum in magneti- 
cally coupled winds are probably too 
small to be important on this time 
scale. Some consequences of this evo- 
lutionary identification were explored 
for U Gem variables. In particular, if 
W UMa stars last on the average 
about 3 X 10° years, then star counts 
indicate that U Gem stars last about 
1 X 10° years, and it is not easy to 
reconcile this result with the lifetime 
of 10 6 years appropriate to the blue 
components in these systems. Accre- 
tion heating of the blue star is sug- 
gested as a possible way out of this 
dilemma, and some mass-loss mech- 
anisms appropriate to the red com- 
panion are considered. 

Faint Blue Stars 

Krzeminski carried out UBV pho- 
tometry of about 30 faint blue stars 
in the regions of the north and south 
galactic poles, chosen from the To- 
nantzintla catalogs. One of the ob- 
jects, Ton S 120, was found to be an 
eclipsing binary with a period of 
3 h 28 m . Its light curve resembles that 
of Nova DQ Herculis, but with much 


deeper eclipses. The eclipse is partial, In particular, the "manganese" stars 

its depth is 2.5 mag, and its duration occupy a very small range in this 

is about 30 minutes. Secondary eclipse diagram. 

has not been detected. The continu- Deutsch, with Abt, Conti, and Wal- 

ously trailed spectra obtained by lerstein, has discussed the double-line 

Kraft showed large radial-velocity spectroscopic binary HD 98088. The 

variations. primary star is known to be a periodic 

In order to improve the period and magnetic variable star. It is also a 
features of the light curve, Krze- spectrum variable, with the lines of 
minski observed photoelectrically HZ Sr II exhibiting two intensity maxima 
29, a peculiar subdwarf with very per cycle. As has been known for 
wide and complex He I lines recently some years, the periods of orbital mo- 
discovered by J. Smak to be a vari- tion, magnetic reversal, and spectrum 
able star with period 17.5 minutes; variation are identical, indicating 
it is supposed that this object is a that the primary star is a rigid ro- 
close binary system. tator. The spectroscopic and photo- 

. metric observations are satisfied by a 

X Boons btars model in which the primary has nor- 

Oke has used photoelectric spectro- mal Balmer lines and mass (2.2 ttl©) 

photometric observations and Hy pro- for an A3 V star. The secondary star 

files to determine effective tempera- lies near A8 V, but its spectrum is 

tures and gravities of the known X probably abnormal at the K line and 

Boo stars. These stars are near spec- at D. 

tral type A0, are rapid rotators, and Dr. Stephen Strom of the Harvard 

have abnormally weak metal lines, and Smithsonian Observatories is 

Two of the stars, y Aquarii and 9 Hy- collaborating with Sargent on an 

drae, have energy distributions indi- abundance analysis of the sharp-lined 

eating that they are binaries. All peculiar A star, HD 204411. Equiva- 

stars, apart from 2 Andromedae, are lent widths were measured for about 

near or slightly above the initial main 1300 lines. The star appears to have 

sequence, and the values of log g are a normal composition, 
similar to those of stars on the initial 

main sequence or the brightest stars Stellar Envelopes 
in the Hyades. This class of star is Mrs. Boesgaard has made a study 
therefore unevolved or only slightly of the line profiles of Ha and the so- 
evolved. The star 2 And has a low dium D lines on direct-intensity 
value of log g and also appears to be microphotometer tracings of 4 - 
below the main sequence. 8-A/mm spectrograms of over 40 late- 
type stars. These stars may be divided 
Peculiar A Stars m ^- three groups: (1) Both the 

Jugaku and Sargent completed an sodium D lines and the Ha- line 
analysis of photoelectric scans of the are asymmetric, with the short- 
continuous energy distributions of 17 wavelength side being the steeper; 
peculiar A stars and 5 A- and B-type (2) the Ha line only, and not the so- 
normal stars. Temperatures and dium lines, shows asymmetry; and 
gravities were derived for all the (3) both the hydrogen and sodium 
stars. It was found that most of the lines are completely symmetric. 
Ap stars have atmospheres character- (There are no cases so far of a star 
istic of dwarf B stars. The Ap stars showing a symmetric Ha profile but 
are segregated in (log g, T e ff) dia- asymmetric sodium lines.) The three 
gram according to their peculiarities, groups are about equally populated 



with stars in this random sample. On 
the basis o£ measurements made of 
half widths, equivalent widths, the 
velocity difference between the core 
and the center of the line (defined at 
the half-intensity point), the stellar 
absolute magnitude measured by the 
Wilson-Bappu method, and previously 
known spectral types and R-I color 
indices, the groups can be interpreted 
in the following way. The asymmetry 
of the almost double-line structure of 
the sodium D lines probably arises 
from the presence of a circumstellar 
envelope, and the asymmetry in Ha 
may be due to the chromospheric be- 
ginnings of this envelope. The second 
group of stars may be in an earlier, 
more rudimentary stage of mass loss 
and development of a circumstellar 
envelope. The stars in group 2 and 
the "normal" stars in group 3 seem 
similar in almost all respects except 
for the Ha asymmetry. The stars in 
group 1 tend to have higher absolute 
magnitudes and later spectral types. 
These stars also have larger equiva- 
lent widths for a given half-width 
than the stars in the other two 
groups. The velocity differences be- 
tween the core and the line center 
range from —5 to —15 km/sec, com- 
mensurate with ejection velocities as- 
sociated with circumstellar envelopes. 
The doubling of absorption lines 
noted in the photographic infrared 
spectra of S stars by Merrill and 
Greenstein is being extensively stud- 
ied by Tsuji. Available are Palomar 
spectrograms of R Andromedae 
(postmaximum in 1956 and 1957) or 
R Cygni (postmaximum, 1964), and 
Okayama 74-inch telescope spectra of 
R Cyg (maximum, 1964) and R And 
(1965 maximum and postmaximum). 
All postmaximum spectra clearly 
show line doubling, while the violet- 
displaced component is absent or very 
weak at maximum, except in Rb I and 
K I lines, which seem to be produced 
in an outermost, very low-tempera- 

ture expanding shell. In other ele- 
ments, the violet lines are at about 

— 35 km/sec. High excitation lines 
have the strongest violet components, 
with an indicated temperature near 
3000°K, as against 2500°K for the 
red component. The excitation tem- 
perature of the red component de- 
creases after maximum, so that the 
decrease of H - opacity makes this 
shell transparent, revealing the hotter 
layers exemplified by the violet com- 
ponent. The strong emission of the 
infrared Ca II triplet is located at 

— 35 km/sec, the same velocity of ex- 
pansion as the violet absorption com- 
ponent. As this expansion is highly 
supersonic, it will produce a shock 
wave, after whose passage the kinetic 
temperature rises to 20,000-30,000° 
K, sufficient to ionize Ca II and H. 
The emission lines of H and Ca II, 
therefore, can be produced in an ex- 
panding shell. A tentative model of an 
S-type long-period variable consists 
of (1) an outermost, very low-tem- 
perature expanding shell (violet-dis- 
placed resonance lines), (2) a normal 
extended atmosphere (red-displaced 
absorption component) , (3) high- tem- 
perature, expanding emitting-layer 
immediately behind the shock front 
(emission lines), and (4) an expand- 
ing layer behind this emitting-layer 
(violet-displaced absorption compo- 
nent) . 


Heintze has obtained 20 A/mm 
spectra of the early-type eclipsing 
variables U Coronae Borealis and 
V 448 Cygni. In cooperating with J. 
Grygar of the Ondrejov Observatory, 
he found the limb-darkening coeffi- 
cient of the 09.5 component of SZ 
Camelopardalis to be near 0.45, and 
that of the B2(?) component to be 
between 0.3 and 0.6. 

Conti has completed his study of 
HR5110 (HD 118216). This F star, 
which has emission lines from the sec- 


ondary at H and K and H« filled in consistent with a spectral type of B3, 
by emission, is interpreted as a semi- (2) that the Si/H abundance ratio is 
detached close binary, viewed at low normal, and (3) that the He/C 
inclination. From an anomalous in- abundance ratio is lower than normal, 
f rared color, the secondary is deduced An attempt was made by Stephen 
to be a subgiant K star. This binary Strom to determine the helium con- 
is much like the well-known eclipsing tent of subdwarfs by using the loca- 
systems, Z Herculis and S Velorum. tions of 62 field subdwarfs in the 
There should be about five times as theoretical H-R diagram. The posi- 
many semidetached systems in exist- tions of the subdwarfs in the H-R 
ence as are known through eclipses, diagram relative to the Hyades main 
It may be possible to discover more sequence permits the estimation of a 
of these from infrared color anoma- range of possible helium abundances. 
lies. These estimates are based upon ho- 
mology relations that follow from re- 

Helium Abundance cent interior calculations. Accurate 

positions of the subdwarfs on the 

Sargent has obtained coude spec- H-R diagram were determined 
trograms (18 A/mm or 9 A/mm) and through a combination of the best 
continuum scans of four blue halo available parallax data and a calibra- 
stars (Feige 56 86, and 92, and tion of the relation between corrected 
AC + 9°6-12) which earlier lower- (B — V) and effective temperature, 
dispersion studies have shown to be The direct determination of the effec- 
deficient in helium and heavier ele- tive temperature for the subdwarfs 
ments. This material, together with was accomplished by comparing spec- 
spectrograms of similar southern trum scans covering the wavelength 
stars, is being analyzed in coopera- region 5000-7500 A, with fluxes pre- 
tion with Dr. Leonard Searle of the dieted from model atmospheres. 
Mount Stromlo Observatory. The These scans were obtained with Oke's 
main aim is to set an upper limit on scanner at the Cassegrain focus of 
the helium abundance with which the the Mount Wilson 60-inch reflector. 
Galaxy originated. This has impor- Values of effective temperature were 
tant cosmological implications. determined for stars having a wide 

Sargent obtained 4.5 A/mm spec- range of ultraviolet excesses, and 

trograms of 38 Draconis B9p, which these values allowed a calibration of 

Eggen has described as one of the the effective temperature-corrected 

brightest horizontal-branch stars in (B — V) relation. A discussion by 

the old disk population. The star has Strom and Strom (Astrophys. J., in 

a very low v sin i; an analysis of its press) of the permissible values of 

spectrum is being made with the helium abundance all but rules out 

assistance of Mr. S. Adelman. zero as a choice for the initial galactic 

Dr. Peter Strittmatter of the Insti- helium content, 

tute for Theoretical Astrophysics, Spectral scans at selected 50 A 

Cambridge, England, is collaborating bands from AA3300-10800 A have 

with Sargent on a study of the spec- been obtained by Visvanathan for the 

tra and compositions of the five B study of the continuum of helium 

stars with weak helium lines in the stars BD + 10°2179, HD 160641, and 

Orion I Association. Preliminary re- BD + 13°3224. The construction of 

suits for i Ori B are (1) that the Si models of pure helium atmosphere in 

Ill/Si II lines ratios and the Hy the temperature range of the ob- 

profiles indicate a temperature that is served stars is in progress. 



Stellar Composition 

Conti has completed his studies of 
the lithium-beryllium ratio in main- 
sequence stars. The new observations 
of the Be line essentially doubled the 
number of stars discussed previously. 
The image tube on the 73-inch camera 
was employed for about 7 of the 
faintest stars in the program. Using 
the parameter Li Be rather than the 
abundance of either element, one can 
atangle the depletion mechanism 
from the effect of different initial 
light-element content. The change of 
lithium-beryllium among main-se- 
quence stars can be interpreted as an 
age effect due to depletion of the 
lithium. The depletion mechanism for 
lithium is more effective in G stars 
than in F stars, i.e., the rate is more 
rapid. The variations in the beryllium 
abundance from star to star can be 
interpreted as different initial 
amounts of this element (and lith- 
ium). Some otherwise normal stars 
contain no detectable Be or Li; pre- 
sumably this is the result of initial 
stellar conditions unfavorable to the 
production of light elements. 

Mrs. Boesgaard has found a very 
strong line due to Li I on a 6.8 A/mm 
spectrogram of T Sagittarii taken by 
Greenstein with the 200-inch coude 
shortly after maximum light in Au- 
1966. This was independently 
discovered by P. C. Keenan on a 13.5 
a mm Palomar spectrogram. The 
lithium line A6707 has a measured 
equivalent width of 1.46 A; a rough 
curve-of-growth analysis yields a 
Li/Ca abundance ratio that is about 
two orders of magnitude larger than 
the Greenstein-Richardson solar 
value. Several other spectrograms of 
M-S stars and S stars are being an- 
alyzed for lithium abundance. 

Equivalent-width measures and 
model-atmosphere studies on 8 sharp- 
lined Pleiades A stars have been com- 
pleted by Conti and Strom. Two turn 

out to be normal in abundance, like 
Vega, and have microturbulence 
amounting to a few kilometers per 
second. The other 6 stars all have a 
higher turbulence, up to 7 km/sec. 
These stars are also all overabundant 
in the s-process elements (Sr, Zr, Y, 
and Ba) by about a factor of 4 with 
respect to Vega. Three of the stars 
are additionally deficient in Sc, Ca 
(and C, with less certainty) and, in 
their composition, appear to be the 
early-type analogues of Am stars. 
It is difficult to understand these 
anomalous abundances in a young 
cluster in terms of current astrophys- 
ical ideas. 

The isotopic abundances of mag- 
nesium can be studied by the MgH 
molecular bands. Mrs. Boesgaard has 
taken spectrograms of 10 bright late- 
type stars at 3 A/mm in the region 
of the 1,0 (A4844) and 0,0 (X5211) 
bands for this purpose. The terres- 
trial isotope ratios are Mg 24 /Mg 25 / 
Mg 26 : 78.6/10.1/11.3. The stellar 
amounts of Mg 25 and Mg 2G are at 
least 10%, and in many cases it ap- 
pears that 20% for both Mg J25 and 
Mg 26 fits the observations better. 

In the photographic infrared the 
crowding of lines in late-type stars is 
not serious and metallic lines can be 
found that are free of blends. The 
relative abundance of zirconium- 
titanium can then be determined by a 
method that should be insensitive to 
the model. Tsuji finds that Zr/Ti in 
R Cygni is 30 times larger than in 
the sun, and 20 times larger in R 
Andromedae. These are lower limits, 
since zirconium is more strongly de- 
pleted by molecule formation than is 
titanium. For example, at 2500 °K 
and a gas pressure of 10 3 dynes/cm 2 , 
with solar composition, the fraction 
of Zr in the form of neutral atoms is 
only 5 X 10" 8 , the rest being ZrO. For 
titanium, nearly half of the atoms are 
neutral. Thus even when Zr I/Ti I is 
10, the Zr/Ti ratio may be 10 3 . Mrs. 


Boesgaard has collected spectra of K, variable S star, HR 1105. A Palomar 

M, M-S, S, and Ba II stars to deter- spectrogram at 2A/mm, obtained by 

mine the Zr/Ti ratio, and finds the Greenstein. has been elaborately 

expected increase from M to M-S to studied. After improvement of the 

S-type stars. She is also studying HR molecular constants, machine compu- 

2028 (classified as M2 + IV by Yama- tation of the detailed band structure 

shita), which is a high-velocity, over about 20 A was made for a wide 

lithium-rich, and apparently metal- variety of isotopic abundances of 

weak star. zirconium. Limits to the reliability of 

On a well-exposed postmaximum the isotopic composition arise from 

spectrogram of R And, taken in 1956, plate noise and from blending of ZrO 

the red CN system is fairly strong in rotational lines with weak TiO fea- 

the violet-displaced velocity system, tures, lines of CaH, and a very few 

Tsuji finds the lines to be similar to atomic features. A few definite con- 

those of cool carbon stars, and that elusions may be drawn: (1) The 

the C 13 N 14 features suggested by abundance of O 18 is not large, in that 

Wyller can be used. The C 12 /C 13 ratio no trace of the displaced ZrO 18 band 

of R And is smaller than the terres- heads can be found. (2) It seems un- 

trial value, but not as low as the likely that the terrestrial ratios of 

equilibrium value suggested by the Zr isotopes hold. (3) It is impossible 

CNO cycle. Thus incomplete proc- to exclude a wide variety of Zr iso- 

essing of the C 12 suggests that the tope ratios, but mixtures in which 

C 13 («, n) O 16 neutron source is oper- either Zr 93 (an unstable isotope) or 

ating in the S stars. those in which the heavier, even iso- 

Coude spectra of M stars have been topes exist seem most likely to be 

obtained in the photographic infra- present in HR 1105. It will be neces- 

red by Tsuji; HD 95735, a high- sary to provide even greater spectral 

velocity dwarf, is of particular inter- resolution to settle this important 

est. Its lines are much weaker than question; either pure (and fainter) 

those of giants. The only strong mul- S-type variables should be observed, 

tiplets are K I (l),NaI (4),theCa or more precise spectrophotometry 

II triplet, Fe I (60), and Ti I (33, profiles of HR 1105 will be needed. 

68). Tsuji plans to compute realistic There are no wavelength shifts to be 

M-star model atmospheres, and is expected in ZrO rotational lines, but 

estimating the line opacity from rather changes of details in the line 

molecular bands, such as TiO and profiles of individual, unblended lines. 

MgH. The isotope splitting is small com- 

In collaboration with L. H. Aller pared to the line width, 
and K. Hunger, Jugaku is continuing Greenstein has obtained high- 
the study of the early-type helium- dispersion spectra of several K stars, 
rich star HD 160641. Between A3270 following a discovery by Spinrad at 
and A4920, most of the lines identified the Lick Observatory that line 
are of He I, He II, C II, C III, N II, strengths measured photoelectrical^ 
N III, O II O III, Ne II, Si III, Si were variable from star to star 
IV, S II, S III, and Mg II. Equivalent among both giants and dwarfs. Spec- 
widths of about 190 lines have been tra confirmed variable strength of the 
measured and will be used for coarse Na I D lines and, most interestingly, 
and fine analyses of this star. of the blue CN band. Two K dwarfs 

Schadee and Mrs. Locanthi have have CN as strong as a giant. The 

attempted to measure the zirconium other conspicuous peculiarity is the 

isotope abundances in the bright non- increased strength of the strong 


pressure-broadened wings of Cr I, gram on G- and K-type stars by tak- 

Mg I, and Xa I lines in both sub- ing spectra at 2 and 7 A/mm with 

giants and dwarfs. The possible ex- the coude spectrograph of the 100- 

planations are increased abundance inch telescope. The approximate 

of metals, a high gas-to-electron pres- wavelength region is AA4300-6700. 

sure, and. for CX, a high value of the The abundance analysis will be exe- 

C ratio. The high C/O ratio is most cured at the Astronomical Institute 

surprising in unevolved K dwarfs. of the University of Amsterdam. He 

The hydrogen-to-helium ratio in .is also studying spectrograms of S 

horizontal-branch blue stars, in globu- Cephei at nine different phases, taken 

lar clusters, and in halo stars re- at his request by Greenstein and 

mains a major puzzle. Greenstein has Kraft. The plates cover the region 

obtained long-exposure 90 A/mm AA3900-4900 and have a dispersion of 

spectrograms of blue stars in M 13, 4.5 A/mm. 

M 15, and M 92, for which Oke has After Dr. Bodo Baschek pointed 
spectrophotometry scans. Except in out an error in their original formu- 
M 13, the He I lines are very weak lation, Conti and Deutsch corrected 
or absent in stars whose UBV colors their discussion of last year concern- 
would demand their presence. The low ing the significance of photometric 
He H ratio found by Sargent and line-strength indices in solar-type 
Searle, by Sargent, and by Green- dwarfs of the disk population. The 
stein and Munch seems to be con- new results still indicate that these 
firmed. This direct contradiction of indices are highly sensitive to micro- 
the big-bang prediction of a He/H turbulence and that the weak-line 
ratio near 30 % remains as a serious characteristic may not always be 
anomaly. One puzzling feature of the taken to indicate metal deficiency. 
globular-cluster spectra was the ap- The authors now agree, however, that 
pearance of the K line of Ca II in abundance-blanketing can have an 
stars of negative B — V color. Radial- appreciable or dominant effect on 
velocity measurements suggested that line-strength indices. In an individual 
the K line was interstellar rather case, spectroscopic analysis may be 
than stellar, and that interstellar Ca required to establish the relative im- 
II lines probably are produced far out portance of turbulence-blanketing and 
in the halo of our Galaxy, above the abundance-blanketing, 
thin-layer gas model normally as- Conti and Deutsch have completed 
sumed. such analyses for two pairs of stars: 

If the globular-cluster blue stars A Serpentis and ft Canum Venatico- 

are as hot as their colors indicate, rum at GO V, and r Bootis and y Ser- 

therefore, the weakness of the helium pentis at F6 IV. In both pairs the 

lines suggests a low ratio for He/H second star differs from the first 

at the surface. A suggestion has been chiefly by showing weaker lines, as 

made by G. Greenstein, Truran, and indicated by the Stromgren indices 

Cameron that in nonrotating, non- m r and the ultraviolet excesses 

convective B stars the helium sepa- S(U — B) . From spectrograms at 

rates by diffusion downward from the 6.7 A/mm in the yellow, over 200 lines 

hydrogen, and that the surface com- of common metals were measured for 

position is not a reliable indicator of equivalent width. The curves of 

the true He/H ratio in the gas out of growth for the stars in each pair were 

which the globular clusters were directly compared. It was found that 

formed. the stars of these pairs differ little in 

In July, Koelbloed finished his pro- microturbulence but are distinctly 


different in metal content. In these vestigated appear to have signifi- 

cases, then, the differences in turbu- cantly smaller masses than those to 

lence-blanketing are very small and be expected for stars in this region of 

the differences in abundance-blanket- the color-magnitude diagram, which 

ing are appreciable. On the basis of would be evolving from left to right, 

the revised calculations of Conti and Values of the pulsation constant com- 

Deutsch, a quantitative agreement puted for these stars indicate that 

that is reasonable is found between they are probably pulsating in a har- 

the predicted index differences and monic mode higher than the funda- 

the observed ones. The evidence no mental. 

longer appears to support the earlier A member of a wide binary system, 

conclusion of Conti and Deutsch that ADS 2849A = HD 24550, was found 

microturbulence decays in solar-type by Dickens to be a variable with an 

dwarfs on the same time scale as extremely small amplitude of 0.02- 

chromospheric activity. 0.03 mag and a period of about 110 

min. The absolute magnitude is de- 

Variable Stars rived from that of the companion 

Danziger and Oke have completed stars; the gravity and temperature 

a study of the short-period variable inferred from the UBV colors show 

VZ Cancri. Line-blanketing correc- that the star has an extremely small 

tions were applied to the scan obser- pulsation constant, suggesting a high 

vations and effective temperatures harmonic mode of pulsation, 

and gravities were derived around A number of stars in the galactic 

the cycle. The minimum temperature cluster NGC 6940 have been investi- 

and the phase relations between light, gated photoelectrically by Dickens in 

temperature, and velocity are similar the search for short-period variabil- 

to those of cluster-type variables such ity. Four of the stars investigated so 

as RR Lyrae, SU Draconis, and X far show some variability, and, in 

Arietis. The surface gravity and the particular, VR 77 varies with an am- 

period-density law suggest an ab- plitude of about 0.05 mag and a period 

normally low value of Q and a low of about 0<?2. This star has a spectral 

mass. VZ Cnc is slightly metal deft- type of F0, an absolute magnitude of 

cient, but lithium is overabundant by +0.5, and lies near the blue end of 

a factor of 25 relative to the sun. the Hertzsprung gap. It could well be 

Danziger and Dickens have derived an example of a 8 Scuti variable, a 

effective temperatures, surface gravi- type which heretofore has not been 

ties, and rotational velocities for a identified in clusters, 

number of recently discovered short- Dickens is obtaining spectrum 

period variables of the 8 Scuti type, scans in the wavelength region 3500- 

using photoelectric spectrum scans 8000 A for a number of bright stars 

and coude spectra. Some of the new between A3 and F7 in order to inves- 

variables have higher temperatures tigate the statistics of gravity and 

and gravities than the previously temperature in the region of the 

known variables of this type. Larger color-magnitude diagram occupied by 

values of the projected rotational the 8 Scuti stars. Spectra of the same 

velocity, v sin i, also occur among the stars are being obtained at the Royal 

new variables and indicate that, in Greenwich Observatory in order to 

the region of the color-magnitude provide accurate rotational velocities 

diagram occupied by these stars, pul- and line profiles for the stars, 

sation is not incompatible with rapid R. F. Garrison has been taking 

rotation. About 50% of the stars in- classification dispersion spectra of 


most of the long-period variable stars lines. Among early- type stars (B8- 
north of declination —30° with max- F), A10830 is found only in super- 
ima brighter than 9th magnitude, giants so far, in particular /? Orionis 
Spectrograms have been obtained cov- (absorption) and e Orionis (emis- 
ering as much of their cycles as prac- sion) . The image tube that was used 
ticable, with preference for the pre- formerly at the 144-inch camera has 
maximum and maximum phases. This now been adapted to the 72-inch cam- 
program is being carried out in col- era of the 200-inch coude spectro- 
laboration with Deutsch and Keenan graph. This has made possible a sub- 
who are continuing to obtain spectro- stantial increase in speed, so that 
grams at 20 A mm in blue and violet observation of fainter stars is now 
for studying systematic anomalies in proceeding. 

various absorption lines and bands. 

Infrared Photometry 

Infrared Stellar Spectroscopy The search for 10-/* circumstellar 

Using the 100-inch coude spectro- emission near a Ori has been contin- 

graph and the detecting system em- ued by Westphal. Observations on 

ployed for observing Mars, Munch three nights have shown no emission 

and Xeugebauer have obtained addi- with a brightness temperature 

tional spectra of a Bootis, a Herculis, greater than 100°K within 2 arc-min 

p Pegasi, and x Cygni in the 2.0-2.4 of the star. 
fi range. These spectra are being 

studied for the purpose of identifying Temperature Scale 

absorption features suspected in Heintze has emphasized a problem 

lower-resolution spectra obtained already recognized by Oke and Searle. 

earlier with the Cassegrain spectrom- The B — V colors of bright, nearly 

eter. unreddened B stars and spectral 

Vaughan and Zirin have continued scans suggest a temperature anomaly 

their study of the He I A10830 absorp- between BO and AO. Heintze feels 

tion in late-type stars as a key to the that the temperature scale must be 

nature of chromospheres in these lowered, and that the surface gravi- 

stars. The line has now been found in ties also must be lowered. He suggests 

emission in 5 stars (e Geminorum, 12 that at BO, B3, and B5 the effective 

Pegasi, c Herculis, y Draconis, and i temperatures are near 27,000, 16,500, 

Aurigae), and temporal changes have and 14,500°K, respectively. The 

been found in 2 stars (a Aquarii and y sharp-lined B stars that he has 

Dra) . The data suggest that the ab- scanned suggest that for the main 

sorption or emission originates in sequence log g < +4. This lower 

discrete clouds or streams. The prin- temperature scale disagrees with that 

cipal conclusions of the study remain given by the Balmer discontinuity 

as stated last year. It has now been and the Hy profiles. There has been a 

noted that a considerable fraction of similar discrepancy, known for some 

the stars showing strong A10830 are time, in the sense that the surface 

close binaries. The A10830 absorption gravity as deduced from observations 

also appears relatively weaker in using even the most advanced model 

dwarf stars. In many of the stars atmospheres has been too low for the 

with A10830, this line shows a sizable mass of the unevolved main-sequence 

velocity shift with respect to other stars. 



A program of absolute spectro- guiding platform. A two-channel 
photometry that will include a re- pulse-counting system has been con- 
calibration of the flux of a Lyrae and structed for the program. This system 
other bright standards has been ini- will continuously monitor the sky 
tiated by Oke and Schild. The absolute background and make possible obser- 
calibration is needed to make pos- vations in moonlight, when the 18- 
sible the comparison of continuum inch schmidt telescope cannot be used 
fluxes predicted by stellar models with effectively for photography, 
spectrophotometric measurements of Standard sources to be compared 
stellar fluxes. The need for an im- with stars include a standard lamp 
proved calibration arises in the su- whose calibration has been referred to 
perior accuracy with which stellar fundamental standards maintained 
fluxes can be measured for compari- by the National Bureau of Standards 
son with increasingly sophisticated and to a blackbody furnace operating 
and accurate stellar models. at the melting point of platinum. The 

During the report year, instrumen- furnace, still under construction, will 
tation for the recalibration has pro- employ as a radiation source a cylin- 
gressed under the direction of Schild. drical cavity 25 mm long with an 
A 4-inch F3.3 Newtonian telescope aperture of 1 mm. The cavity will be 
has been constructed and mated to the surrounded by platinum heated to its 
existing prime-focus scanner of the melting point in an induction furnace. 
200-inch telescope. Because of the For observations at Palomar, the fur- 
short focus of the calibration tele- nace will be installed on the roof of 
scope, lamp bulbs and other standard the machine shop where it will be 
sources located as close as a half mile visible from the dome of the 18-inch 
are readily compared with stars. To telescope. 

ensure stability of optical properties Visvanathan and Oke have taken 
and to minimize scattering by the ac- up a study of 25 stars in the Pleiades 
cumulation of foreign matter on alu- cluster. The absolute fluxes in selected 
minized optical surfaces, the calibra- 50-A bands from 3450 to 8000 A, and 
tion telescope is hermetically sealed Hy line strengths, have been obtained 
in a nitrogen-filled tube having quartz for 16 stars. These are being corn- 
entrance and exit windows. The cali- pared with fluxes computed from 
bration telescope and scanner can be model atmospheres to determine ef- 
mounted on the Palomar 18-inch fective temperatures, 
schmidt telescope, which serves as a 


Color-Magnitude Diagrams V ~ 17.5 mag to And the character of 

The photometric survey of the the density distribution along the 

bright parts of color-magnitude (C- horizontal branch and the steepness 

M) diagrams of northern globular of the giant branch. Both features 

clusters was continued by Sandage are valuable in the morphological 

with the assistance of Katem, using description of clusters because they 

photoelectric standards and photo- have previously appeared to be cor- 

graphic interpolation techniques. Re- related with metal abundance, 

connaissance studies of NGC 5897 The C-M diagram of NGC 5897 re- 

and NGC 5053 were completed to sembles M 3, M 92, and w Centauri, 



with most of the stars along- the 
horizontal branch confined blueward 

of the RR Lyrae gap. The gap ap- 
pears at V = 16.2, which indicates a 
distance modulus of m — M ~ 15.7 if 
M v = =±0.5 for the RR Lyrae stars. 
The reddening- appears to be very 
small and is here neglected. The 
cluster appears to be moderately 
metal poor because AV of the giant 
branch, read at B—V = +1.4, is 2.7 
mag brighter than the horizontal 

XGC 5053 is a sparse cluster in 
which photometry can be carried 
close to the center. Early work in col- 
laboration with Dr. H. L. Johnson of 
the Lunar and Planetary Laboratory, 
University of Arizona, combined with 
recent data, shows that the horizontal 
branch occurs at V=16.9, which gives 
m—M ~ 16.4, again assuming M v = 
— 0.5 for the RR Lyrae stars. The 
morphological characteristics are sim- 
ilar to NGC 5897. 

Helium Deficiency 

Sandage and Wildey completed a 
study of the remote globular cluster 
XGC 7006. The horizontal branch 
occurs at V = 18.8, which indicates 
the cluster is at the large distance of 
45 kpc from the sun. Its distance from 
the galactic plane is 15 kpc. The C-M 
diagram is unique. The density dis- 
tribution along the horizontal branch 
would indicate high metal abundance 
if the correlations, heretofore ac- 
cepted, were to hold, because the 
branch is very heavily concentrated 
toward the red end, as in NGC 6171, 
NGC 6356, NGC 6712, and 47 
Tucanae. However, the giant branch 
has &V = 2.6, which is 0.5 mag 
larger than the value for strong-lined 
clusters. Furthermore, spectrographic 
data by Morgan show NGC 7006 to 
be very metal poor. 

The foregoing is a clear-cut viola- 
tion of the correlation between metal 
abundance and density gradient along 

the horizontal branch, and shows that 
a second parameter, in addition to 
metal abundance, controls the charac- 
ter of this branch. Theoretical calcu- 
lations by Faulkner suggest that this 
parameter is the variation of the he- 
lium abundance. The sense of the 
change is that NGC 7006 is deficient 
in helium compared to the majority 
of globular clusters in the halo. 

Evidence from other remote clus- 
ters, such as the Draco system with 
data by Miss Swope, the Leo II sys- 
tem by the same author, and all clus- 
ters in the Small Magellanic Cloud by 
Gascoigne, suggests that the NGC 
7006 anomaly occurs from many ag- 
gregates that have not participated 
in the chemical evolution of the galac- 
tic system. In this regard, NGC 7006 
may not be gravitationally bound to 
the galaxy, as evidenced by its ex- 
tremely high radial velocity. If the 
result is true that the helium abun- 
dance differs inside and outside the 
galaxy, this may be a clue to some of 
the early events of the nucleosyn- 

Other clusters are known which 
are certainly not bound gravitation- 
ally to the galactic system, as shown 
by an excess of kinetic energy in the 
radial coordinate over the potential 
energy required for negative total 
energy via the viral theorem. A test 
of the hypothesis of different helium 
abundance inside and outside the gal- 
axy can be made by determining 
whether or not the NGC 7006 anom- 
aly of the horizontal branch is present 
in these aggregates. 

Completing work begun at the Lick 
Observatory, Sargent has discussed 
low dispersion spectra of eleven blue 
horizontal-branch stars in M 13, M 
15, and M 92. The main conclusions 
were: (1) In each cluster, helium is 
deficient by a factor greater than 
5-10, and (2) there are abnormally 
strong interstellar K lines of Ca II 
in the spectrum of stars in these 


clusters, which are all relatively un- its present rotation against the torque 
reddened. The observation of helium exerted by the solar wind. The argu- 
weakness in a star in M 15 is in con- ment tends to support Dicke's con- 
flict with the observation by O'Dell, tention that the sun's interior rotates 
Peimbert, and Kinman (Astrophys. much more rapidly than does the sur- 
/., HO, 119) that the He/H ratio is face, 
normal in a planetary nebula in the 
cluster. NGC 6171 

"Blue Stragglers" Th, e R , R , L 7 r f e ™riablra in the 

_ , , , . , , , metal-rich globular cluster NGG 6171 

Sargent has continued worn, begun are being studied by Dicke ns in the 

at the Lick Observatory on stars in UBV col usi direct Mes ot> 

M 67, which Deutsch described as tained at the Newtonian focus of the 

blue stragglers in Year Book 65 100 _ inch te ie S cope. A total of 60 V, 

(pp. 148-150). Photoelectric meas- 56 £, and 22 ?7 plates were obtained 

urements were made of the continuous in 1966 and further lateg are being 

energy distribution for ten of these taken in 1%7> Measurement of the 

stars These were combined with Hy lateg hag - ugt commenced using the 

profiles measured on Lick spectra in Sartorius i r i s _ dia phragm photometer 

order to obtain gravities and tern- that hag recent] been modified to 

peratures and hence masses for these ide digitized output of the X and 

stars Although there are difficulties y coordinates an d the iris-diaphragm 

in fitting the observed energy distn- readi on punched cards . Compu t er 

butions to models, it appears that the programs have been writt en to proc- 

stars have masses near 1 mo, and egs the punche d-card output, 

hence that the blue stragglers in this The reddening in the field of NGC 

cluster should be described as hori- 1(m hag been determ ined from photo- 

zontai-branch stars. electric UBV observations of nearby 

Deutsch has continued to observe Md gtars> A value of E = Q 29 

the so-called blue stragglers— the wag obtained) wh i c h confirms the 

anomalous A stars that occur in sev- previous value f ound by Sandage but 

eral old open clusters. In developing which ig somewhat less than the 

the hypothesis that these objects are value of E = Q 37 ma derived by 

the metamorphs of solar-type stars sturch baged on UBV observat ions of 

that have evolved through the giant four variables near minimum light. 
stage of evolution, he now proposes 

to explain their positions relative to ^, 7 ^ 7 , 7 ^ 7 , 

the main sequence by taking them to 0ther Globular Clusters 

be helium - core horizontal - branch Clusters being studied by Dickens 

stars of the kind for which Dr. John are NGC 5466, NGC 6402, NGC 6981, 

Faulkner has given models appropri- and NGC 7099. A number of direct 

ate to Population II objects. He has plates, together with preliminary 

also shown that if there now resides photoelectric sequences, have been ob- 

in the solar interior as much angular tained with the 100-inch and 60-inch 

momentum as was contained in the telescopes. Most attention has been 

protomorphs of the blue stragglers given to the metal-rich cluster NGC 

when they were solar-type dwarfs, 6981, which contains about 40 RR 

then the radiation entering the base Lyrae variables. The observations of 

of the solar hydrogen convection this cluster will be combined with 

zone will supply angular momentum those of NGC 6171 in order to provide 

at a rate that is sufficient to maintain more statistical weight to the analy- 



ses of the relatively few variables 
that occur in that cluster. 

Praesepe Cluster 

Rotational velocities have been de- 
termined by McGee, Khogali, Baum, 
and Kraft using- an electronographic 
image tube as well as conventional 
spectroscopy for virtually all the 
stars in Praesepe brighter than V — 
9.6 mag. Xew UBV photoelectric 
observations have been made by 
Dickens and Krzeminski for all the 
stars for which rotations are avail- 
able. The photometry is combined 
with that obtained much earlier by 
Johnson. The brighter stars in Prae- 
sepe appear to rotate on the average 
about 15 r c faster than those in the 
Hyades at the same absolute magni- 

The effects of rotation on the tem- 
perature and surface gravity have 
been investigated in the U — B, B — 
V diagram. A weak correlation of 
ultraviolet excess with departure 
from mean rotation (at a given color) 
is obtained, although it is rather less 
pronounced than that found earlier 
in the Hyades by Kraft and M. Wru- 
bel. The effect of rotation on absolute 
magnitude has been investigated in 
the color-magnitude diagram in a 
manner similar to that followed pre- 

viously by Strittmatter; he analyzed 
the same cluster but used less accu- 
rate rotations by Treanor and in- 
cluded fewer stars in the analysis. A 
much weaker correlation of magni- 
tude excess with rotation excess (from 
the mean at a given color) is found 
by means of the new data. This result 
is more nearly in accord with the 
smaller effects predicted by models in 
uniform rotation than those in non- 
uniform rotation. 

Cluster Stars with Anomalous 

In his classification of the Upper 
Scorpius stars, Garrison noted a num- 
ber of stars which have the colors of 
B3-B5 stars but the spectra of B7- 
B9 giants, with some peculiarities. 
He is now measuring 20 A/mm plates 
of these stars taken by Kraft and 
Deutsch. In a survey of the Orion 
belt region several more stars have 
been found with these characteristics. 
Since one of them is only a few min- 
utes of arc away from o- Orionis E, a 
"helium-rich" star, it is interesting to 
examine their fit on an H-R diagram. 
For this reason, spectra have been 
obtained and classified for most of 
the stars in the cluster surrounding 
a Ori. 


Interstellar Absorption Lines 

Yaughan and Munch continued 
their study of interstellar sodium and 
calcium lines with interferometric 
equipment attached to the coude 
spectrograph of the 100-inch tele- 
scope. The newly completed coude 
scanner of the 200-inch telescope, 
equipped with a Fabry-Perot inter- 
ferometer, is also being employed. 
Both installations give a velocity res- 
olution of the order of 1 km/sec, 
comparable with the resolution now 

reached on radio-absorption observa- 
tions at 21 cm. 

Present work confirms that the 
line-of-sight velocities of distinct in- 
terstellar sodium (Na I) and calcium 
(Ca II) components are identical to 
within a measuring accuracy of ±0.5 
km/sec (p.e.). The relative strengths 
of many sodium and calcium com- 
ponents differ considerably from one 
interstellar "cloud" to another, prob- 
ably implying local excitation dif- 
ferences. Not all such differences are 



consistent with an earlier correlation 
by Routly and Spitzer (Astrophys. 
J., 115, 227, 1952) between Na I/Ca 
II and the peculiar velocities of the 
clouds. A detailed study of this prob- 
lem is in progress. 

In the Orion region, the Na I spec- 
tra obtained by Vaughan and Munch 
contain some 35 real and reasonably 
distinct absorption features in the 
directions of five bright stars within 
11° of NGC 1976 (the stars i, f, 9 \ 
e, o) Orionis) . A similar distribution 
of Ca II features is found. In the best 
spectra the weakest features whose 
reality is certain show equivalent 
widths of 1.5 ml, although a few 
slightly weaker features seem to be 
detected. Component velocities range 
from —5 to +40 km/sec heliocentric, 
or from —22 to +23 km/sec in the 
local standard of rest (the velocity 
associated with differential galactic 
rotation is about +7.5 km/sec at 
500 pc). Intercomparison of velocities 
within the line complexes in dif- 
ferent directions indicates that of 
the 35 features two or three are 
surely common to four directions at 
least, and two are common to three 
directions. It appears that the region 
is crossed by at least seven individ- 
ually coherent "clouds," "wisps," 
or "sheets" of interstellar gas, some 
of which subtend angles of 11° or 
more. Since a feature associated with 
NGC 1976 (the Orion nebula) is one 
of these, its transverse size can be 
taken to exceed 100 pc. 

Nova Envelopes 

With the purpose of studying the 
interaction of matter ejected in nova 
explosions with surrounding inter- 
stellar material, a number of old 
novae are being photographed anew 
with the 200-inch by Munch and com- 
pared with earlier images obtained by 
Baade and others. 

The envelope of Nova Herculis 1934 
has been found to be definitely decel- 

erating. Between 1956 and 1966 the 
expansion rate of the semi major 
axis was 0.19 arc sec/yr, compared 
with the mean rates 0.30, 0.25, 0.23, 
and 0.22 arc-sec/yr derived between 
the time of explosion and 1940, 1942, 
1956, and 1966, respectively. For the 
semi major axis, the rate between 
1956 and 1966 was 0.10 arc-sec/yr, 
while the mean rates from zero age 
are 0.22 arc-sec/yr for 1940; 0.19 for 
1942; 0.17 for 1956; and 0.15 for 
1966. The most remarkable change 
observed in 1966 in the envelope of 
Nova Her 1934 is that the bar run- 
ning across the central star in the 
direction of the minor axis, which 
in 1966 appeared brighter than the 
elliptical ring, was relatively much 

The envelope of Nova Persei 1901 
also shows indications of deceleration. 
Between 1949 and 1964, the expan- 
sion rate of the fastest filaments was 
0.35 arc-sec/yr, while the mean rate 
between initial explosion and those 
dates are 0.55 and 0.50 arc-sec/yr, 
respectively. Not all the filaments in 
this complex envelope, however, show 
the same behavior, and indeed a few 
of them appear to be accelerating. 

Crab Nebula 

For a doctoral thesis, Virginia 
Trimble has undertaken the measure- 
ment of a series of Ha -[N II] plates 
of the Crab Nebula taken with the 
100-inch and 200-inch telescopes by 
Baade and Munch, with the purpose 
of determining the proper motions of 
the filaments. On the whole, the fila- 
mentary mass is in a state of expan- 
sion, with an apparent center that is 
8" east and 15" south of the suspected 
central star. The deviation is several 
times the probable error of the deter- 
mination of the convergence point. 
The date of convergence, assuming 
uniform velocities, is about 90 years 
after zero age (1054 A.D.), indicat- 
ing some acceleration, but somewhat 


smaller than found earlier by Duncan. The flux density from the nebula in 

On the basis of this determination the two wavelength bands 1.5-1.8 /x 

and published radial velocities, the and 2.0-2.4 fx has been measured with 

distance to the nebula is 1.1 or 1.7 an accuracy of 10%. The results, 

kpc, depending on whether it is a when combined with O'Dell's (1963) 

prolate or an oblate spheroid. Spec- near-infrared photometry, show that 

trograms of various portions of the the origin of the infrared radiation 

nebula taken by Munch have been re- is nonthermal with a spectral index 

measured for radial velocities, which, of about —0.8. 
combined with the proper motions, 

will provide a three-dimensional Extinction in Emission Nebulae 

picture of the structure and kinemat- Spectrophotometric data for 25 
ics oi the nebula, as well as a unique early-type stars imbedded in 14 emis- 
determination of its size and distance, sion nebulae have been obtained by 
J. Scargle has completed a photo- Christopher M. Anderson utilizing the 
electric and photographic study of the photoelectric scanner at the 60-inch 
continuum radiation of the Crab and 100-inch telescopes. The apparent 
Nebula. The plates taken with the spectral energy distributions have 
100-inch and 60-inch Mount Wilson been compared with the predictions 
reflectors and the 48-inch schmidt at of model-atmosphere calculations in 
Palomar, are being used in connec- order to obtain the wavelength de- 
tion with 200-inch direct plates by pendence of the interstellar extinc- 
Munch, Baade, and Arp to study tion. It has been found that the stars 
rapid motions in the ''amorphous" HD 164492 in NGC 6514 and Her- 
mans. The spectral energy distribu- schel 36 in NGC 6523 show the same 
tion was measured through inter- reddening anomaly as does the star 
ference filters at the Newtonian focus 6 1 Orionis. A theoretical study of the 
of the 100-inch. In agreement with anomaly is under way by Anderson, 
earlier work by Oke, the characteris- since the previously accepted expla- 
tics of the radiation in the ultraviolet nation, based on the removal of small 
appear to indicate departures from a particles by radiation pressure, can- 
single power law in the relativistic not operate because of viscous drag, 
electron-energy spectrum, or the ex- as Krishna Swamy and C. R. O'Dell 
istence of bremsstrahlung from a hot have recently shown, 
gas also producing the observed X 
rays. A second result is that the slope High-Latitude Planetary Nebula 

of the continuum is the same all over A blue star imbedded in very faint 

the nebula — a fact that places restric- nebulosity has been discovered by Arp 

tions on possible models for injection and investigated by Arp and Scargle. 

of fresh relativistic electrons by the Spectra of the nebulosity and central 

stellar remnant of the supernova, star show the object to be an old 

Scargle has also studied the physics planetary nebula, but to have a higher 

of a relativistic gas in a magnetic galactic latitude (& 11 — —47°) than 

field in order to understand the prop- any previously listed by Abel (Astro- 

agation of disturbances in a gas such phys. J., lUU y 259, 1966). The central 

as the high-energy component of the star is identified as No. 932 in the 

Crab Nebula. Haro-Luyten Palomar Catalogue 

Eric Becklin, a graduate student in (Bol. Obs. Tonantzintla y Tacubaya, 

physics, has made infrared observa- 3, 37, 1962). A distance of 150 pc has 

tions of the Crab Nebula using the been computed for the star, with the 

60-inch telescope at Mount Wilson, use of an accurate proper motion 


furnished by Luyten together with parent diameter is about 4". A spec- 

the measured radial velocity. Inter- trum at 400 A/mm on 103a-D emul- 

ference-filter photographs of the neb- sion shows a typical planetary nebula 

ulosity in the light of [0 II] and Ha emission-line pattern and a faint cen- 

show an unexpected comet shape for tral-star continuum. Since the appar- 

the outer material and elongation ent brightness of the object is only 

along a different axis for the inner V — 15 mag, B = 16 mag, it is ap- 

material. parently at a great distance above the 

An object near the north galactic galactic plane. Further analysis is in 

pole, at R.A. — 12 h 57^5 and Dec progress. 

= +27° 50', was noticed by Haro Several 200-inch direct photographs 

(Publ. Astron. Soc. Pacific, 63, 144, and spectra of Sharpless 71, a galac- 

1951) to have emission features tic nebulosity, and its central star 

characteristic of a planetary nebula, were obtained by Arp. Joseph Miller 

Although the object is stellar in ap- of the University of Wisconsin is 

pearance on Sky Survey plates, a using these observations in his study 

photograph in good seeing by Arp of the object, 
with the 200-inch shows that its ap- 


The survey of the sky for objects data for the first coverage and for 

emitting in the spectral ranges 0.68- about five sixths of the second cover- 

0.92 \l and 2.01-2.41 p, sponsored by age have gone through preliminary 

the National Aeronautics and Space processing. Approximately 6000 

Administration, has been continued sources have been found brighter than 

at Mount Wilson by Leighton and K = 3, of which about 450 have / — 

Neugebauer and several graduate stu- K > 5. The sources show a marked 

dents. A description of the survey concentration along the galactic 

instrumentation was given in Year plane ; the mean I — K of sources ob- 

Book 6Jf> (pp. 25-26) . served in the galactic plane is larger 

Although only about one third of by about one magnitude than that 
the nights during the last year were for sources out of the plane. A statis- 
used, a second coverage of the sky tical study of the data is now in prog- 
has been essentially completed. The ress. 


In August 1966, E. Becklin ob- 110 arc-sec resolution. There is one 

served radiation from the vicinity of main source of radiation which is 

the center of the Galaxy in the wave- about 3-4 arc-min (10 pc) in diameter 

length region 2-2.5 p. Further obser- and which is extended along the ga- 

vations were subsequently made at lactic plane. Within observational 

0.9, 1.65, 2.2, and 3.4 /x by Becklin error, this main source agrees both in 

and Neugebauer using the 24-inch, position and size with the radio source 

60-inch, and 200-inch telescopes. Sagittarius A. Scans of high spatial 

The general infrared radiating resolution show that the source is non- 
structure of the central region of the gaussian in shape. Instead, its radia- 
Galaxy is shown on the contour map tion is highly concentrated toward 
in Fig. 2. This represents data ob- the center, where there appears to be 
tained with the 24-inch telescope with a core of radiation 30 arc-sec in diam- 



■28' 40' - 





-I -29° 00 - 



-29° IC' - 


Fig. 2. A contour map of the galactic center region at 2.2 fx taken with an aperture of 1.8 arc- 
min diameter is presented. Contour lines are separated by 8.5 X 1 0~ 20 W m~ 2 Hz -1 ster -1 (5.2 
X 10~ 10 W cm -2 jut 1 ster -1 ) and are uncertain by about the same amount. The X represents a 

pointlike source. 

eter (1.5 pc linear size). There is 
also a pointlike source just slightly 
offset from the position of maximum 
brightness. Besides this main source, 
Fig. 2 shows that there is general 
background radiation lying predom- 
inantly along the galactic plane and 
that there are also several weaker 
extended sources of radiation. None 
of the weaker infrared sources is 

coincident with a known radio source. 
A comparison of the infrared 
structure of the galactic center has 
also been made with the infrared 
structure of the nucleus of M 31 as 
determined from observations with 
the 200-inch telescope. When a resolu- 
tion of 14 pc is used for comparison, 
the two galactic centers look similar 
in both shape and brightness. 


Distance Modules of NGC 2/+03 

As a first step for the redetermina- 
tion of the Hubble constant, Tam- 
mann and Sandage have completed 
measurements and analysis of the 
variable .stars in NGC 2403. This late 
Sc galaxy closely resembles M 33 in 
stellar content and in the character 

ing member of the M 81 group and is 
the first galaxy beyond the Local 
Group in which cepheid variables 
were found soon after the 200-inch 
telescope went into operation in 1949. 
Hubble began the work in 1950. 
Plates continued to be taken by San- 
dage until 1963 and a photoelectric 

of the spiral pattern. It is an outly- sequence was determined in four ob- 


serving seasons from 1962 to 1965 types in M 31 and M 33; and (5) the 
ranging from B = 10 to B = 22.8. angular size of the largest and the 
Altogether, 160 blue plates and 22 average of the five largest H II re- 
yellow plates are available covering gions, calibrated from linear sizes 
the time interval from 1910 to 1963. determined in M 33 and the Magel- 
By blinking the complete material, 56 lanic Clouds. 

variable stars have been located, of The five methods agree remarkably 

which 17 are cepheids with periods well (±0.2 mag), and although the 

from 20 to 87 days, 8 are very bright final analysis is not quite complete, 

blue variables with irregular light an interim value of (m — M) = 

curves, 14 are supergiant red varia- 27.56 ± 0.2 is suggested as of June 

bles of the SRc or Ic class, 1 is an 1967. Even after final reduction is 

eclipsing binary, and 16 are faint un- complete, an improvement of this 

classified variables among which are value can eventually be made from a 

cepheids of undetermined period. recalibration of photometric data in 

The 17 cepheids are all extremely the LMC and SMC for the bright red 

faint. None becomes brighter than variables that appear near F = 11. 

B = 21.2, and most are fainter than A start on this program is expected 

B = 22.0 at maximum light. Light next year during a visit by Sandage 

curves have been determined for all to Australia. 

17 variables in both B and V wave- The distance to NGC 2403 is im- 

lengths. All but one of the cepheids portant because it permits the cali- 

disappear below the plate limit (B ~ bration study of linear sizes of H II 

23.5) at minimum, and all become regions to be expanded to the M 81 

fainter than the limit of the magni- and M 101 groups. With distances to 

tude sequence (B = 23.0) at some these groups available, the number of 

phase of their light curve. calibration galaxies of Sc and Irr 

The distance modulus of NGC 2403 types is enlarged to nearly 20 sys- 

has been determined by five methods, terns, and the mean value of the linear 

which depend on (1) the cepheids at diameters, D (largest) and D (5), 

maximum light in both B and V, to- w [\\ have high weight. To obtain the 

gether with a new composite period- Hubble constant the angular diame- 

luminosity relation at maximum, ob- ters f h II regions and the magni- 

tained by combining recent data for tudes of the bright red stars can then 

cepheids in the LMC, SMC, M 31, and ^ e use d m the range 31 > (m - M) 

NGC 6822, calibrated by nine ceph- > 2 7 for determining distances to 

eids of known M B (max) and M v Sc and Irr galaxies that have meas- 

(max) from the galactic system; (2) ure( j re dshifts. Kristian has begun 

the apparent magnitude of the red the measurement of angular sizes of 

supergiant variables, assuming M v the H II regions of such galaxies for 

(max) = -8.10 for the first-ranked w hich 200-inch H« interference-filter 

member of the class as derived from p i a t es are available, and a first value 

similar stars in NGC 6822 and the of H is expected next year. 
Magellanic Clouds; (3) the apparent 

magnitude of the brightest resolved Variables in Divarf Galaxies 

stars as obtained by counts across the Miss Swope has now found periods 

face of the galaxy, and calibrated as- photographically for 76 RR Lyrae 

suming M B = —9.5 as determined variables in the Leo II system. Their 

from other members of the Local mean period of 0.589 day is shorter 

Group; (4) the bright irregular blue than the 0.611 day found for the 

variables calibrated from their proto- Draco-system variables and the am- 


plitudes appear to be smaller than pletely dominated by a nonthermal 

those found in Draco. The magnitudes component that, conjecturally, arises 

are not yet based on a photoelectric from the center of the nucleus. Oke 

sequence. However, the two variables and Sargent's model is similar in its 

with period longer than 1 day corre- essentials to one that I. S. Shklovsky 

spond in absolute magnitude to those has derived from considerations of 

found in Draco and Sculptor and are the radio-frequency spectra of Sey- 

brighter than the variables of like fert galaxies. 

period found in galactic globular A spectroscopic comparison has 

clusters. The Leo II system also has been made by Arp of the compact 

six irregular variables that lie on the galaxy III Zw 2 [R.A. = h 8^0, Dec 

bright end of the giant branch in the = + 10°42' (1950)] and the radio 

color-magnitude diagram. There are galaxy 3C 120, also designated Parkes 

also six with a blue color index that 0439 + 05. Both have typical Seyfert 

vary slowly and that are bluer and spectra. The radio source in 3C 120 

slightly brighter than RR Lyrae var- has a small angular diameter and a 

iables. Xo such stars were found in flat radio spectrum. 

Draco. The nuclei of ten Seyfert galaxies 

have been studied further by Kurt S. 
Seyfert Galaxies Anderson as a thesis project. Photo- 
Oke and Sargent have completed electric-scanner observations of the 
their work on the nucleus of the Sey- continua and emission features have 
fert galaxy NGC 4151. Their model been obtained in the region AX3400- 
of the nucleus, deduced from photo- 5800, and are currently being ex- 
electric scans and photographic spec- tended to approximately A8400. Spec- 
trograms, consists of an object 50 pc trograms are also being obtained to 
in diameter, filled with 10 5 ttlo of assist in the determination of line 
gas at 10 6O K and with an elec- strengths and profiles to permit the 
tron density N e = 100 cm- 3 . This gas estimation of electron densities and 
gives rise to the coronal-line spectrum temperatures. Theoretical calculations 
first observed by O. C. Wilson. In this have been undertaken in order to as- 
are embedded cooler clouds or fila- certain the nature of the processes 
ments of total mass 3 X 10 4 tit© responsible for the observed optical 
that have random motions of about and radio continuous spectra. 
1000 km /sec, giving rise to the main 

forbidden-line spectrum of the object. Compact Galaxies 

The cool component fills only 0.004 of Sargent began observations with 

the volume of the nucleus and is sup- the 200-inch prime-focus scanner of 

posed to be in pressure equilibrium the continuous energy distributions 

with the hot component, having T e = and emission-line strengths in eleven 

20,000 °K and N e = 5000 cm- 3 . It is compact galaxies discovered by 

conjectured that the broad wings in Zwicky (four lists privately circu- 

the H lines and He II A4686 are pro- lated) and by Arp (Astrophys. J., 

duced in a dense gas having N e > 14,2, 402). The resolution was 50 A 

10 cm- 3 and T e <~> 20,000°K. This and the wavelength range covered 

last component contains only a few was AA3300-6000. Four of the galax- 

tens of solar masses of gas and oc- ies studied (I Zw 0246 — 03, I Zw 

cupies a volume, perhaps at the center 0955 + 51, II Zw 0431 — 02, and III Zw 

of the nucleus, only about 0.01 pc in 0824 + 21) have energy distributions 

diameter. It is concluded that the op- that span the small range exhibited 

tical continuum of NGC 4151 is com- by normal dwarf and giant ellipticals. 


Four of the remaining objects (I Zw nucleus of NGC 5934) has a continu- 

0930 + 55, I Zw 1439 + 53, II Zw ous spectrum intermediate between 

0430 + 05, and II Zw 1622+41) have the compacts with and without emis- 

spectra with broad emission and sion lines. Greenstein has found that 

much flatter continua than the non- the spectrum of this object is dom- 

emission-line objects. There is an in- mated by A-F stars in the blue and 

dication that there are two compo- by a cooler population near A.5000. 

nents in the continuous spectra of the The two Arp compacts have fairly flat 

emission-line compacts ; namely, a spectra with emission lines similar to 

flat component that dominates the those of the Zwicky emission-line 

blue end of the spectrum and a com- compacts. Sargent intends to observe 

ponent rising toward longer wave- many of these objects in the near in- 

lengths (although less steeply than frared during the next season. He is 

the energy distributions of normal also analyzing the emission-line data 

ellipticals) that dominates the red in an attempt to understand the phys- 

end of the spectrum. It is not likely ical conditions in the compact gal- 

that either of these components is axies. 

primarily due to recombination radi- A large family of compact galaxies 
ation from a hot gas because Oke has with a great variety of spectra con- 
found that I Zw 1727 + 50 has a two- taining very blue, normal visual, and 
component continuous spectrum simi- infrared objects, some with weak 
lar to that of the emission-line com- radio emission, can now be regarded 
pacts; however, this object shows no as firmly established. Zwicky has corn- 
emission lines either on the scans or piled a list containing 446 such ob- 
in spectrograms obtained by Zwicky jects. From observations of differen- 
(Astrophys. /., 1^3, 192). tial redshifts, the percentage of sky 

At the telescope the compact gal- coverage by clusters of galaxies, and 
axies in Zwicky's lists appear to be the theory of characteristic limiting 
sharply bounded and 5-15 seconds of bodies, Zwicky inclines to the view 
arc in diameter. So far, any pro- that the quasi-stellar radio sources 
nounced central condensation (which belong to the family of compact gal- 
is always in the form of a star like axies and that Einstein shifts are 
nucleus) is shown only by the emis- responsible for a large part of the 
sion-line objects. observed redshifts. 

Spectrograms of three of the emis- Compact galaxies remarkably often 
sion-line compacts have been de- occur in small groups, some members 
scribed by Zwicky (op. cit.). The of which may be interconnected by 
additional galaxy II Zw 0430 + 05 was luminous intergalactic formations, 
observed spectroscopically by B. Among the latter, the widely sepa- 
Peterson and by Sargent with the rated but interconnected compact gal- 
Newtonian spectrograph of the 100- axies at R.A. = 19 h 26 m , Dec = +61° 
inch telescope. It has broad emission 50' (1950) may be mentioned, 
lines of H and [O III] and appears Only about six clusters of compact 
to be a Seyfert galaxy. The redshift galaxies among 8000 clusters have 
is z = 0.0323 and the absolute mag- been found so far. Examples are the 
nitude of the nucleus is M v = —20.8. medium-distant cluster 0658 + 6321 
On this basis, II Zw 0430 + 05 is in- with a population of 189 semicompact 
trinsically the most luminous known galaxies of very uniform types and 
Seyfert nucleus. It is also a radio the distant cluster 0152.0 + 3337 with 
source. a population of 144 medium-compact 

The object I Zw 1356 + 37 (the to compact galaxies. Observations of 


differential redshifts among: these and show that it is an almost pure 

may. in Zwicky 's opinion, make it emission feature, observable in the 

possible to decide if some fraction of light of both [0 II] and H«. Forbid- 

these shifts is due to the Einstein den emission lines of [0 II], [N II], 

gravitational effect, and thus throw and [S II] have been found in the 

new light on the distances of the nucleus of M 87. Other lines are seen 

quasi-stellar objects. which seem to be asymmetric with 

In 1965 Arp reported the discovery respect to the nucleus. The possible 

of a small double galaxy not much connection of the latter lines with the 

larger than a globular cluster (Astro- jet and counter-jet will be studied in 

phys. /., 142, 402, 1965). Since then the coming season and a detailed 

he has obtained a spectrogram at 190 analysis of the redshifts of all these 

A mm in very good seeing on which lines will be attempted. All the ellip- 

it is possible to measure the differen- tical (E) galaxies in this region of 

tial velocity of the emission lines of the Virgo cluster are distributed in 

the two components. O'Connell and a narrow band on either side of M 87. 

Arp now report a derived mass of Arp considers it significant that the 

10 s m,©, on the assumption that the axis of this distribution of companion 

two components are mutually bound galaxies coincides with the position 

in orbit about each other. The mass- angle of the jet and counter-jet. 
to-luminosity ratio agrees well with 

ratios previously found by Zwicky for Ultrafamt Blue Stars 

blue compact galaxies. The mass den- In continuation of earlier investi- 

sity in the present object is about one gations, a survey of the location and 

hundred times that in the average the magnitudes of individual blue 

galaxy. Triple-image studies by Arp stars on 200-inch blue and yellow 

and O'Connell with the 48-inch plates has been undertaken by 

schmidt demonstrate that this object Zwicky in collaboration with Luyten. 

has a large ultraviolet excess. Spec- Fifteen pairs of 5" X 7" plates (with 

trum scans by Sargent confirm the a scale of 11"2 arc/mm) were used. 

ultraviolet excess. Arp and O'Connell The blue exposures, generally of 27 

have analyzed other compact galaxies minutes, were on unfiltered 103a-O 

with ultraviolet excesses and have plates; the yellow exposures, gener- 

pointed out some of their connections ally of 45 minutes, were made on a 

with Haro blue galaxies, compact 103a-D plate through a GGII filter. 

galaxies, and quasi-stellar galaxies. The observations, unfortunately, were 

mostly made under inferior conditions 

Counter-Jet in M 87 f see ing. The plates were blinked by 

It has long been known that the Luyten, who also determined the star 

bright elliptical galaxy in the Virgo positions — usually relative to a single 

cluster, M 87, has a jet with a con- AGK-2 or BD star on the plate. Luy- 

tinuous spectrum extending about 22" ten is also responsible for the magni- 

from the nucleus of the galaxy. Arp tude and color estimates. The former 

has now discovered luminous streaks were derived by comparing the blue 

extending in the opposite direction, plates with the blue Sky Survey 

The markings are brightest between plates taken with the 48-inch tele- 

30" and 40" from the nucleus, but scope, on the assumption that their 

extend out to more than 120" from limiting magnitude is 21.2. Colors 

the center of the galaxy. Spectro- were estimated relative to some six 

grams and interference-filter direct comparison stars of about the same 

photographs register the counter-jet magnitude, and subsequently cor- 



rected for the expected average color 
of these stars. It is hoped that the 
accidental errors in the magnitudes 
and colors do not exceed r "5 and 
0^25, respectively 

Data for 217 ultrafaint blue ob- 
jects will be published as a separate 
list. These objects are generally 
fainter than m p — 19.0 and are stellar 
in appearance. The number of stars 
is as yet too small and the material 
not sufficiently homogeneous to allow 
one to draw significant conclusions 
concerning the rate of increase with 
fainter magnitudes. The intention is 
to determine proper motions for ob- 
jects of 17 mag and brighter from 
the 48-inch Survey plates, and then 
to combine these measurements with 
more accurate photoelectric colors in 
order to derive the relative frequency 
of white dwarfs, subdwarfs, horizon- 
tal-branch stars, and compact galax- 
ies or even quasars among these 

Catalogue of Galaxies and of 
Clusters of Galaxies 

Volumes I, II, and III, which cover 
the unobscured areas from Dec —3° 
to +57°, and Vol. V, Dec -3° to 
+ 21°, on the southern galactic cap, 
have now been published. These four 
volumes contain, respectively, data on 
9500, 6700, 5300, and 4200 galaxies 
as well as on 1300, 2350, 2700, and 
1200 clusters of galaxies. 

About half of the work necessary 
on the clusters of galaxies and on the 
magnitudes and positions of the gal- 
axies brighter than rn p = 15.7 to be 
included in Vol. IV, covering the un- 
obscured areas north of Dec +57°, 
has been completed by Zwicky and E. 

Work has been completed on the 
clusters of galaxies that are to be in- 
cluded in Vol. VI, which covers the 
unobscured areas between Dec +21° 
and +51° of the southern galactic 
cap. Positions and magnitudes for 

about two thirds of the galaxies to 
be included in this volume have been 
determined by C. Kowal. 

This work has been supported in 
part by a grant from the National 
Science Foundation. 

Statistical Analysis 

The distribution over the least ob- 
scured parts of the sky of 2350 clus- 
ters of galaxies listed in Vol. II of the 
Catalogue has been analyzed by M. 
Karpowicz and Zwicky by the method 
of dispersion-subdivision curves. It 
was found, as from the previous anal- 
ysis of data in Vol. I, that this dis- 
tribution is random except for some 
slight deviations that are possibly 
ascribable to the absorption of light 
by intergalactic dust clouds in the 
clusters in Virgo, Coma, Hercules, 
and Leo A and B. 

The distribution of the 1200 clus- 
ters of galaxies listed in Vol. V of the 
Catalog has also been analyzed by 
Karpowicz and Zwicky by means of 
the dispersion-subdivision method. 
They report that this distribution is 
the most random found so far in any 
region of the sky and that, as for the 
distribution of the clusters in Vols. I 
and II, no clustering of clusters of 
galaxies is found. 

The analysis by Zwicky of the dis- 
tribution of clusters of galaxies over 
the northern galactic cap, in right 
ascensions from Milky Way to Milky 
Way and declinations —3° to +57°, 
shows the remarkable feature that the 
total numbers of clusters of galaxies 
in the half areas east and west of 
R.A. 13 h m are, respectively, in the 
ratio of about 2 to 3, while the nearer 
clusters are more evenly distributed. 
The relatively high depletion of the 
numbers of medium distant (MD), 
distant (D), very distant (VD), and 
exceedingly distant (ED) clusters in 
the eastern half of the area mentioned 
is ascribed to the greater depth of 
interstellar dust on that side of the 



Milky Way, and, conjecturally, to 
intergalactic dust in the Virgo, Coma, 

Hercules, and other near clusters, 
and in the Ursa Major and the Shane 
Clouds. A total relative excess of 
about 0.2-0.3 mag in dimming over 
the eastern half of the area could 
account for the difference in the total 
counts of clusters. The total area of 
the sky covered by the clusters in the 
two halves, significantly, is much less 
affected. The distribution over the 
part of the southern galactic cap an- 
alyzed in Vol. V does not show any 
analo go u s a symmet r y . 

Using the contours as they are 
plotted in the Catalogue, Zwicky es- 
timates that rich clusters with sym- 
bolic velocities of recession in the 
range 5000 < V s < 100,000 km/sec 
in projection cover three fifths of the 
whole celestial sphere. He calculates 
that, to redshifts of the order of z = 
2.0, the sky is therefore covered 
three times by the central parts of 
clusters, even if we assume that the 
number density of clusters is uni- 

Classification of Clusters of Galaxies 

In the first four volumes of the 
Catalogue of Galaxies and of Clusters 
of Galaxies, 7505 rich clusters are 
listed. They have been classified as 
compact, medium compact, and open. 
A first survey concerning the types 
of galaxies to be found in clusters 
has now also been made. Zwicky 
points out that clusters are of differ- 
ent morphological types and that total 
populations of galaxies in clusters 
vary widely. The material content 
of the clusters varies inasmuch as 
the following preponderances of 
galaxies among their brightest mem- 
bers have been found: (1) Clusters 
containing only irregular galaxies. 
These are very rare. (2) Clusters con- 
taining a great variety of spirals and 
irregular systems. (3) Clusters of all 
types of galaxies, except that few 

very compact systems are in evidence. 
(4) Clusters like type 3 that also con- 
tain many compact objects. (5) Clus- 
ters composed mainly of elliptical 
galaxies but with few compact ob- 
jects. (6) Clusters like Type 5 that 
also contain many compact galaxies 
as well as ellipticals with giant com- 
pact cores. (7) Clusters of compact 
galaxies. Although numerous groups 
of compact galaxies have been found, 
rich clusters consisting of only com- 
pact galaxies are rare, at least at dis- 
tances at which the symbolic velocity 
or recession is less than 50,000 km/ 

The Local Group, although it is not 
a rich cluster, would be described as 
of type 3. 

From the data given in the Cata- 
logue, the volume r, within which, on 
the average, one rich cluster of gal- 
axies may be expected, has been re- 
calculated assuming a Hubble con- 
stant of 100 km/sec per megapc. It 
was found that t is equivalent to a 
cube with an edge of 40 megapc. This 
closely parallels the value derived by 
Zwicky in 1938 from the study of only 
about one hundred of the nearest clus- 
ters of galaxies, which had been pho- 
tographed with the 18-inch schmidt 

Special Cluster 

As a curiosity, Zwicky mentions 
the exceedingly distant cluster CI 
1101 + 7042, which shows a ringlike 
arrangement of 39 galaxies on an 
ellipse about 5' in diameter, and 
which might possibly be influenced by 
a gravitational lens effect. 


On the search program directed by 
Zwicky, 15 supernovae were discov- 
ered at Palomar: 7 by Kowal, 7 by 
Rudnicki, and 1 by Reaves. 

Extended statistical studies on the 
frequency and distribution of super- 



novae were started by Zwicky and 
A. P. Fairall (working at the Mc- 
Donald Observatory of the University 
of Texas). 

Light curves have been constructed 
and spectral characteristics have been 
studied of the supernovae SN 1962e 
(by Zwicky and Rudnicki) and SN 
1966b (by Zwicky, Gates, Bertola, 
and Ciatti). The light curve of SN 
1966e in NGC 4189 is being deter- 
mined by Kowal. 

Zwicky, as chairman of Committee 
28 of the International Astronomical 

Union, has continued to coordinate 
the search for supernovae at 14 ob- 
servatories in both hemispheres and 
has, in particular, proposed the con- 
sideration of new instrumentation 
and procedures to accelerate the rate 
of discovery and subsequent detailed 
observations and evaluations. Special 
attention is being given to procedures 
of searching for supernovae within 
the bright cores and disks of ellipti- 
cal So, Sa, and Sb galaxies, many of 
which have presumably been missed 
by the methods used in the past. 


Cosmic-Ray Sources 

The origin of cosmic rays is usually 
associated with supernovae in our 
Galaxy. The alternate hypothesis that 
cosmic rays are universal has been 
based on estimates of the contribu- 
tion made by radio galaxies (Bur- 
bidge and Hoyle) . The long harmonic- 
mean life (10 9 years) for radio 
galaxies derived by Schmidt (see 
Year Book 65) reduces their contri- 
bution to the cosmic rays by a factor 
of 10 3 . As a consequence, most prob- 
ably only a small part of the locally 
observed cosmic rays can be attrib- 
uted to radio galaxies. 

Redshifts of Radio Galaxies 

As time permitted, Sandage con- 
tinued the program of redshift deter- 
mination of radio galaxies identified 
from the 3CR catalog. The project 
is coordinated with parallel work by 
Schmidt with the aim of eventually 
determining redshifts of all identified 
radio galaxies from the homogeneous 
Third Cambridge Catalogue, The re- 
sults will eventually be combined with 
existing photoelectric magnitudes in 
the BVR system which Sandage has 
been assembling for several years. 

During the report year, redshifts 
of 9 new radio galaxies, corrected for 

galactic rotation, were obtained as 

follows: 3C 29 (AA/A = z = 

0.0450), 3C 76.1 (0.0328), 3C 135 

(0.1270), 3C 227 (0.0855), 3C 236 

(0.0988), 3C 285 (0.0797), 3C 381 

(0.1614), 3C 449 (0.0181), and 3C 
455 (0.0331). 

Energy Distribution 

Oke has begun a study of absolute 
energy distributions in the continua 
of radio galaxies and related objects. 
Radio source 3C 371 has an energy 
distribution that is very similar to 
that of Seyf ert galaxies and to some 
of the Zwicky compact galaxies that 
are being studied by Sargent. 

Direct Photography 

Direct photography of radio 
sources with the 200-inch telescope is 
being continued by Arp. The fields 
include quasi-stellar radio sources 
and some that are blank upon visual 
inspection, but most of the objects be- 
ing investigated are radio galaxies. 
The more important radio-galaxy 
identifications with 3C radio sources 
have been and are being photo- 
graphed, and some 4C sources are 
also being inspected. John Bolton has 
furnished a list of the most important 
Parkes sources for which optical in- 
formation is needed. One preliminary 


result is that a number of radio gal- giant characteristics) turn out, under 

axies that have been classified "D" on high resolution, to be various kinds of 

small-scale photographs (implying spirals. 


Scorpius X-I began a series of photoelectric meas- 

A cooperative effort was begun urements of the object with the 200- 
with the American Science and Engi- inch telescope immediately after 
neering - Massachusetts Institute of Tokyo identification results were re- 
Technology X-ray astronomy group ceived by telephone from Giacconi. 
for the optical identification of X-ray The object was found to vary in opti- 
sources. These workers, under the cal brightness by ±0.03 mag in time 
leadership of R. Giacconi, performed intervals as short as 1 minute. 
a series of spectacular rocket experi- Sandage then obtained spectro- 
ments in 1966 and early 1967 from grams on six nights in July at the 
which relatively precise celestial co- same time that Patrick Osmer, a stu- 
ordinates were obtained for the dent at the California Institute of 
brightest point X-ray source Sco X-l, Technology, monitored the object 
and for the fainter sources Cyg X-l, photoelectrically with the Palomar 
Cyg X-2, and Cyg X-3. 20-inch reflector. Variations of 0.8 

The first flight concentrated on mag were observed from night to 

Sco X-l. By use of an X-ray modu- night, and on these the very short- 

lation-collimator, together with op- period fluctuations were superposed, 

tical photographs of the sky taken at Large night-to-night variations in 

1-sec time intervals through the col- the spectral-line intensities were also 

limator on board the rocket, the ASE- found during this period. 

MIT team succeeded in obtaining a The spectra show all hydrogen 

set of positions that were accurate to lines in emission down to at least 

= 1 minute of arc in both coordinates, H 10. The high excitation He II line 

with two equally probable positions at A4686 is present. A complex blend 

differing by 5 minutes of arc. of high excitation C III, N III, and 

Predictions of the nature of the O II is present from AA4630 to 4655, 

optical counterpart of the X-ray but the intensity of this feature 

source are based on the hypotheses changes radically from night to night, 

that the X rays are produced either as does the absolute intensity of the 

by free-free transitions of hot gas hydrogen lines. 

(bremsstrahlung), or by synchrotron The interstellar H and K lines of 

emission. These predictions had sug- Ca II are seen in absorption with 

gested that the optical object should equivalent widths of 0.2 A. Analysis 

be very blue with B — V ~ +0.1, by Munch suggests that the path 

U — B ~ -0.9, and should appear length of Ca II is N Ca n ' L = 5 X 10 12 

at V ~ 12 mag. atoms cm -2 . This, combined with esti- 

Communication of the X-ray posi- mates of average ionization equilib- 

tion by Giacconi to both the Tokyo rium conditions and abundance ratios 

Observatory and to Mount Wilson and of carbon and calcium to hydrogen, 

Palomar permitted the identification gives N 2 H • L — 2 X 10' 20 atoms cnr 5 . 

and confirmation in June 1966 of a By combining the N 2 • L values with 

starlike object of the correct color the measured X-ray absorption due to 

and magnitude in one of the error galactic hydrogen at 1 kev, Sandage 

boxes of the X-ray positions. Sandage estimates that the distance is about 


400 psc. The X-ray absorption data each of six nights. Jugaku suspects 

were obtained from rocket flights by that radial-velocity changes also oc- 

Steward, Mark, Chodel, and Grader cur on a short time scale of about 60 

at the University of California. This minutes, but the measurements are 

distance, combined with the observed exceedingly difficult and the results 

X-ray flux, shows that the source are marginal. Westphal and Kristian 

emits about 10 37 ergs/sec in X-rays in are continuing the work by remeasur- 

the range 1-10 A. ing the plates with digital techniques, 

In April 1967, Westphal and San- using a microphotometer connected 

dage began an experimental monitor- to the Caltech 7094 computer. Analy- 

ing of the optical flux of Sco X-l sis is planned with the use of West- 

with time resolutions as small as 2 phaFs cross-correlation method that 

seconds. Observations with the 60-, is capable of very high precision for 

100-, and 200-inch telescopes, using wavelength determination, 

the pulse-counting equipment operat- Neugebauer is making a series of 

ing in an automatic mode as designed measurements of X-l Sco at a wave- 

and built by Dennison's group, length of 2.2 /*. 
showed that the source was highly 

variable in time intervals as short as Sources in Cygnus 

1 minute. Continuous monitoring of A second rocket flight by the ASE 

the source for intervals of 2 to 4 group produced positions for three 

hours on five separate nights shows X-ray sources in Cygnus with an ac- 

evidence for nonrandom variations curacy of about ±11 minutes of arc 

with amplitudes of 0.04 mag in a time in both coordinates. Three-color im- 

interval of 4 minutes, superposed on age plates of the Cygnus region were 

longer-term variations with periods obtained by Osmer, and several ultra- 

of about 30 and 60 minutes. Fourier violet candidate objects were marked 

analysis of the records gives partial before the X-ray positions were 

evidence that some of the character- available. One of these objects was 

istic periods repeat from night to subsequently found to lie within the 

night, but the work is still in a pre- X-ray error box and is considered as 

liminary state. a tentative identification with Cyg 

Jugaku has measured Sandage's X-2. Position measurements by Peach 

spectrographic plates for radial ve- and photoelectric photometry by San- 

locity changes and finds evidence for dage are consistent with this possi- 

small, but apparently significant, var- bility on the basis of the X-ray po- 

iations from night to night. The sition and the predictions of the color 

plates were taken by the single-trail and magnitude of the optical object 

technique with a long slit over time from the bremsstrahlung hypothesis. 
intervals of from 2 to 3 hours on 


Spectroscopy 454.3 (0.860), 4C -04.6 (1.445), AO 

Spectroscopic observations of 0952 + 18 (1.472), 3C 205 (1.533), 

quasi-stellar sources are being con- 4C 29.50 (1.927), and PKS 0229 + 13 

tinued by Schmidt. Further redshifts (2.067) . 

have been determined for the QSSs Sargent and Schmidt have collab- 

4C 01.4 (z = 0.261), 4C 15.1 (0.451), orated with Dr. J. N. Bahcall of the 

3C 57 (0.670), 4C -00.6 (0.720), 3C California Institute of Technology 



in an analysis of the absorption spec- 
trum of 3C 191 using a 200 A/mm 
spectrogram obtained by Schmidt. 
The main part of the analysis was a 
study of the relative populations of 
the ground-state fine-structure levels 
of various ions, which Bahcall has 
shown may be used to obtain limits 
on the electron density and photon 
flux in the region where the absorp- 
tion lines are formed. Fourteen ab- 
sorption lines were studied, typically 
having equivalent widths of about 10 
A. A study of the central intensities 
of different absorption lines showed 
that most of them are unsaturated. 
The main conclusions of the analysis 
are: (1) Most of the absorption lines 
are resolved and have half -widths of 
the order of 3 A in the rest frame of 
the source. (2) Either the electron 
density is of order 10 +3 cm -3 or the 
distance between the continuum 
source and the absorbing regions is 
of order 10 +2 sl pc. (3) The value of 
the fine structure constant at z = 2 
equals the laboratory value to within 
measuring errors (about 5%). (4) 
The C/Si abundance ratio by number 
is 2.5/1, with a certainty that is prob- 
ably less than a factor of 3. Thus 
the composition of 3C 191 is signifi- 
cantly different from Population I 
stars, where the C/Si ratio is about 
13. (5) There is no evidence on this 
spectrogram for the dependence of 
Zabs on ionization potential, which 
Bahcall found earlier in the data of 
Burbidge, Lynds and Burbidge, and 
Stockton and Lynds. (6) There is no 
evidence for absorption lines arising 
from metastable states of S II and C 
III, as other observers have claimed. 
The emission-line spectrum of 3C 
273 B was obtained by Greenstein at 
38 A/mm. No trace of any sharp lines 
was found, and no evidence of A3727 
of [0 II] . The latter is suppressed at 
high density. Its absence indicates 
that no cool, low-density gas is pres- 
ent, for example, in a cool shell or in 

the cool interfilamentary spaces. The 
relatively cool shells seen in absorp- 
tion in more highly redshifted quasi- 
stellar sources have small internal 
velocity dispersion and electron den- 
sities near 10 4 cm- 3 . Such shells 
might have produced sharp forbidden 
emission lines of [O II] if present in 
3C 273B. Apparently the dense fila- 
ments that produce the H and [0 III] 
lines of 3C 273B are imbedded in a 
hotter region. 

Parkes Source 0237-23 

In the fall of 1966, John Bolton, 
director of the Australian Radio 
Observatory at Parkes and, at that 
time, guest investigator at Mount 
Wilson and Palomar Observatories, 
pointed out the unusually flat radio 
spectrum and sharp, synchrotron self- 
absorption cutoff of the radio source 
Parkes 0237-23. Two spectra at 400 
A/mm and one at 190 A/mm were 
obtained with the 200-inch by Arp. 
Later, spectrograms further into the 
red, and photometry were obtained by 
Kinman of the Lick Observatory. 
This relatively bright quasi-stellar 
source has the largest redshift meas- 
ured to date. The emission-line red- 
shift is z = 2.22. Arp, Bolton, and 
Kinman reported only a slightly 
smaller redshift, z = 2.21, for the 
numerous sharp absorption lines that 
are present in the spectrum. Later, 
E. M. Burbidge reported an absorp- 
tion-line redshift of z = 1.95. This 
led to an attempt by Greenstein to 
observe it at high dispersion. Two 
spectrograms at 90 A/mm permitted 
a resolution near 1.5 A on the better 
plate. Although a larger number of 
probably real absorption lines were 
measured, a selected list of about 50 
was published by Greenstein and 
Schmidt. The existence of two sepa- 
rate sets of lines with shifts of 2.202 
and 1.956 was established. The lines 
are sharp and deep. There is little 
doubt that lines from excited states 



exist, suggesting relatively high den- 
sity and some collisional excitation. 
In addition to the usual strong ab- 
sorption lines such as Lyman a, C IV, 
Si III, and Si IV, the ionized metals 
are very strong, notably Ti III, Fe III, 
and Mn II. Continued observations of 
this extraordinary double shell are 

Spectral Energy Distribution 

Oke has continued to measure ab- 
solute spectral energy distributions 
of the continua of quasi-stellar 
sources. Wavelengths are now chosen 
in such a way that some of the 
stronger emission lines also are meas- 
ured. Data now exist for several 
sources in addition to the seven for 
which energy distributions have al- 
ready been published. For 3C 345 
several sets of observations exist, in- 
cluding detailed observations of the 
Mg II line at A2798. No obvious vari- 
ations have been seen. The object 3C 
208 has an energy distribution simi- 
lar to 3C 245, and the A1550 line of C 
IV must be very strong. Source 3C 
334 has a spectrum in which the flux, 
F v , increases only slowly as v de- 
creases; the Mg II line A2798 is 
strong. Source 3C 249.1 has a nearly 
flat spectrum in terms of F v , and a 
strong Mg II line. The radio-quiet 
quasi-stellar object PHL 938 has an 
energy distribution and line intensi- 
ties almost identical with those in 

The two highly variable quasi- 
stellar sources 3C 279 and 3C 446 
have been studied in detail. Changes 
in magnitude in 3C 279 of 0.25 mag 
per day have been observed. The vis- 
ual magnitude has varied by nearly 2 
mag during the past year. Following 
the outburst of 3C 446 discovered by 
Sandage in June 1966, scanner obser- 
vations showed a decrease in bright- 
ness of the continuum at the rate of 
0.1 mag per day for 5 days. In both 
objects the continuum is somewhat 

redder when the object is fainter. The 
absolute strengths of A2798 of Mg II 
in 3C 279 and A1550 of C IV in 3C 
446 have remained constant, inde- 
pendent of any change in the con- 
tinuum. This confirms an earlier re- 
port by Sandage, Westphal, and 
Strittmatter. The rapid light varia- 
tions of the continuum imply that 
most of the continuum comes from a 
region with a diameter of the order 
of 3 X 10 16 cm. The region produc- 
ing the nonvariable emission lines 
must be much larger, and conven- 
tional models with dimensions of 1 
pc or more apply. 

Wampler and Oke have studied 3C 
273 in great detail using the Lick 
and Palomar photoelectric scanners. 
They have found several broad com- 
plex emission features; the evidence 
is strong that these are lines of Fe II. 
A considerable contribution to the 
[0 III] line, A5007, is in fact due to 
Fe II. One feature is probably the 
sodium D lines. No evidence for He 
I A4471 or He II A4686 is found. The 
presence of permitted Fe II lines and 
the absence or extreme weakness of 
[Fe II] lines indicates that the elec- 
tron density in the emitting region is 
at least 10 6 cm -3 . 

Optical Survey for Radio-Quiet 
Quasars or Quasi-Stellar Objects 

Work has continued on the nature 
of the many faint blue objects found 
in high galactic latitudes. Sandage 
and Dr. W. J. Luyten of the Univer- 
sity of Minnesota completed the first 
phase of their investigation of the 
Haro-Luyten blue star field centered 
at l h 36 m , +6° (1855). As reported 
last year, UBV photometry of 69 ob- 
jects, divided equally among the three 
Haro-Luyten color classes, showed 
that a large number of the whiter 
stars of color class III (PHL catalog, 
Table IV), are F and G subdwarfs 


of the globular-cluster main-sequence density to B = 18.1 is at least 1 per 

population. These stars populate the square degree. 

galactic halo and occur in all magni- In a separate program, UBV pho- 

tude intervals of the PHL catalog to tometry of 28 objects in the blue-star 

its limit (B ~ 18.1). field of Richter near the globular 

For the bluer stars of color classes cluster M 3 indicated that the surface 
I and II (PHL catalog, Tables II density of radio-quiet QSOs is be- 
am! Ill), objects of two types dom- tween 1 and 3 per square degree to 
inate the list. Sandage obtained spec- B ~ 19.7, and, therefore, that the 
trograms of 28 objects in the l h 36 m , number of QSOs increases rapidly 
+6° held with the prime-focus spec- with increasing limiting magnitude. 
trograph of the 200-inch. These were In view of these encouraging re- 
selected from a candidate list of pos- suits, Luyten and Sandage began a 
sible radio-quiet quasi-stellar galaxies more extensive search and analysis 
in this field. The list was composed of program for blue objects in seven 
objects from the 69-object photo- fields for which early first-epoch Sky 
metric sample which had negligible Survey plates existed. These fields 
proper motions as measured by Luy- are: 8 h 28 m , +18°; 9 h 32 m , +24°; 
ten on 48-inch schmidt plates with 10 h 24 m , +18°; ll h 16 m , +30°; 12 h 34 m , 
an 11-year interval. Additional ob- +30°; 14 h 18 m , +24°; and 15 h 10 m , 
jects were included from the bluest +24°. Three-color image plates were 
PHL color class, again on the basis obtained with the 48-inch schmidt and 
of negligible proper motion, even catalogs of the 8 h 48 m , +18° and 
though no UBV photometry existed. 15 h 10 m , +24° fields have been pre- 

The spectra showed that the can- pared and published. At the end of 

didate list was dominated by radio- the report year, Sandage had obtained 

quiet quasars (formerly called QSG UBV photometry for 73 faint objects 

but now referred to as QSO, denoting in the 8 h field and 15 objects in the 

"quasi-stellar objects") and white 15 h field with the 200-inch, and Luy- 

dwarfs. Eleven definite QSOs were ten had initiated a good start on the 

found, three of which showed several proper-motion measurements. 
emission lines so that unambiguous Preliminary analysis of the photo- 

redshifts could be determined. The metric data shows that the color dis- 

redshifts were z = 1.990 for PHL tribution is of the same character as 

1127, 0.298 for PHL 1194, and 1.847 in the l h 36 m field. The separation into 

for PHL 3424. Although the remain- two types of distributions in the U — 

ing eight objects had only one broad B, B — V diagram at V ~ 15 mag is 

emission line somewhere in the region striking. Brighter than this limit, 

AA3150-5000 A, these lines were of most objects follow the standard U — 

such a character that the QSO iden- B,B — V line for main-sequence OB 

tification was positive. stars, nuclei of planetary nebulae, hot 

Sandage and Luyten estimate that, subdwarfs, and blue horizontal- 

to B = 18.1, the surface density of branch stars of globular clusters. 

QSOs in this particular field is at Fainter than this limit, few, if any, 

least 0.3 per square degree, and prob- such stars are present, but rather the 

ably 0.5 per square degree. This is distribution is almost entirely con- 

100 times larger than the surface fined to the neighborhood of the 

density of radio quasars in the SCR blackbody line, which is the domain 

catalog and agrees with the esti- of the QSOs, white dwarfs, U Gemi- 

mate given last year before spectra norum variables, and composite stars, 

had been obtained. The white-dwarf These data suggest that the density 



gradient perpendicular to the galac- tion of QSSs must be increasing with 
tic plane for normal OB stars, hot redshift, the excess density factor 
subdwarfs, etc., is so steep that few being about 15 at redshift 0.5, and 
exist at halo distances corresponding about 100 at redshift 1. These esti- 
to V c=l 15. On the other hand, white mates depend on the cosmological 
dwarfs with M v > +10 do exist to model; those given are for a value of 
distances of at least z = 500 pc and q between and 1. 
contribute to the sample fainter than This density increase at earlier 
V = 15. There are also many objects cosmic times is of the same order of 
in the photometric sample that fall magnitude as that found by Cam- 
above the blackbody line and are, bridge radio astronomers from source 
therefore, good QSO candidates. On counts. An alternate interpretation of 
the basis of the data now available, the source counts considered by the 
Sandage and Luyten estimate that radio astronomers is that the mean 
there are at least 100,000 QSOs to an luminosity rather than the density of 
optical limit of B ~ 19.7 over the the sources was higher at earlier 
entire sky, and that this number is cosmic times. The present investiga- 
an increasing function of limiting tion of 3C sources shows that the 
magnitude. density increase is independent of 
Braccesi has proposed and tested a luminosity, and thus excludes a varia- 
technique for discovering QSOs photo- tion of mean luminosity with cosmic 
graphically through their infrared time. 

excess, which in three-color photome- The distribution of absolute optical 

try should permit them to be distin- magnitudes per unit volume is fairly 

guished from stars. In trials using constant over the observed range of 

1-N plates at the 48-inch schimdt some four or five magnitudes. The 

telescope, he produced a list of four radio luminosity function shows a 

QSO candidates that were later considerable gradient, the number 

confirmed to have large redshifts. per magnitude interval increasing by 

Three of the spectrograms were ob- a factor of 3 or 4 per magnitude. If 

tained by R. Lynds of the Kitt Peak the quasi-stellar objects found by 

National Observatory and one by Sandage are relatively weak radio 

Schmidt. sources, then their surface density on 

... the sky can be explained by extend- 

Space Distribution mg . j-j^ f a i r ]y steep radio-luminosity 

Schmidt has studied the space dis- function only a few magnitudes be- 
tribution of QSSs. For this purpose yond the observed range of radio 
he used the 33 sources in the re- absolute magnitudes. If this is cor- 
vised 3C catalog that are optically rect, then the local (2 = 0) space 
brighter than visual magnitude 18.4 density of all quasi-stellar objects, in- 
and for which the redshift has been eluding those observed as radio 
determined. For each source a sources, is about 10 s Mpc -3 . 
"maximum" redshift, z m , can be com- 
puted at which it would still be con- Optical Variability 
tained in the revised 3C catalog Peach has started a program to 
(which has a lower limit of 9 flux monitor the optical variability of the 
units at 178 Mc/s) and would be known quasi-stellar sources. Using 
optically not fainter than magnitude the 48-inch schmidt, blue plates are 
18.4. A comparison of the actually taken monthly of all objects observ- 
observed redshift with z m for each able at the time. Despite the poor ob- 
source shows that the space distribu- serving season, over 300 plates of 83 


objects have been obtained within the of the brighter QSSs has been de- 
interval from November 1966 to June layed by poor weather. 
1967. This program is intended both 

to provide information on optical Infrared Emission 

variability throughout this class of Neugebauer and Becklin have 

object and to detect sudden changes measured the infrared emission from 

of brightness that may be correlated eight quasi-stellar sources at 2.2 /x 

with spectroscopic or other changes, with the 200-inch telescope; (V — K) 

A further program to observe is about 3. Further observations are 

shorter-term variations among some now in progress. 


Heating of the Solar Atmosphere QSO Spectra 

Stein proposed a model for the Dr. J. N. Bahcall of the California 

heating of the solar atmosphere in Institute of Technology has collabo- 

which the magnetic field controls the rated with Sargent on a study of the 

heating through its effect on wave emission lines to be expected in the 

propagation. The solar atmosphere is spectra of quasi-stellar objects be- 

heated mainly where the magnetic tween Lyman a (A 1215) and the Ly- 

field is large. Where the field is weak, man limit (A 912). Estimates were 

the acoustic wave flux, produced by made of the strength of emission lines 

the turbulence associated with the of the abundant elements, based on 

convective zone, is dissipated in the the behavior of these elements in 

form of shocks in low-lying high- QSO spectra that have already been 

density layers, and produces little observed at longer wavelengths. This 

heating. On the other hand, where work was intended partly to facili- 

the field is large and acoustic waves tate the discovery of QSOs having 

become fast-mode waves, the dissipa- redshifts greater than z = 2. It was 

tion decreases at great heights, where concluded that only one very strong 

the ratio of the Alfven speed to the line, A 977 of C III, is to be expected 

sound speed becomes large. Thus, in the wavelength region studied. 

where the magnetic field is large, the Bahcall and Sargent also pointed out 

wave energy can be carried up to that two effects, the presence of Ly- 

great heights. To pursue the model man a in the B filter and the existence 

in detail, the propagation properties of a large Lyman discontinuity in 

of magnetohydrodynamic waves were absorption, will cause QSOs having 

investigated. A study of the reflec- redshifts greater than z = 2.2 to be 

tion and refraction of fast-mode and redder in U — B than objects having 

slow-mode waves from density and z < 2.2. Bahcall and Sargent thus 

magnetic-field jumps was nearly com- suggest that the search for optical 

pleted, and an investigation of their identifications of radio sources should 

propagation in a stratified atmos- not be confined to objects having a 

phere in the Wentzel-Kramers-Bril- pronounced ultraviolet excess, 
louin approximation was begun. A 

crude model of the magnetic field in Observational Cosmology 

the network was also considered, and A series of computer programs has 

it was found that only along lines of been written by Peach for the analy- 

force reasonably close to the vertical sis of the redshift-magnitude obser- 

does the Alfven velocity increase up- vations by Sandage of first-ranked 

ward. cluster galaxies and radio galaxies. 


The data are analyzed in terms of the dent, has completed calculations of 

homogeneous isotropic cosmological the dynamics of a collapsing proto- 

models of general relativity, both star for various initial conditions. In 

with and without the inclusion of a all cases the collapse is found to be 

cosmological constant in the redshift- extremely nonhomologous, in such a 

magnitude relation. These programs way that a very small part of the 

determine the best values of the den- mass at the center reaches stellar 

sity and deceleration parameters in a densities and stops collapsing before 

least-squares sense; the errors of most of the cloud has had time to 

these quantities are investigated by collapse very far. The stellar core 

Monte Carlo techniques. The neces- thus formed subsequently grows in 

sary evolutionary corrections to the mass as material falls into it. During 

apparent magnitudes of distant ob- most of this time, the forming star is 

jects can be included in the compu- completely obscured by dust in the 

tations. As the observational data are inf ailing cloud, the absorbed radia- 

not yet completely reduced, a final tion reappearing in the infrared as 

value of the deceleration parameter thermal emission from the dust. The 

is not available, but preliminary cal- resulting star is almost a conven- 

culations indicate a higher value than tional Hayashi pre -main -sequence 

those previously obtained. These cal- model, but it appears rather low on 

culations further indicate that the the Hayashi track. It seems that some 

cosmological constant is zero to the of the observed properties of T Tauri 

precision of the determination. stars may be explained on the basis 

of the infall of material and the in- 
Star Formation f rared emission from dust, as found 
Richard B. Larson, a graduate stu- in this work. 


Under the guest investigator pro- object have been submitted for publi- 

gram, which has been in effect at the cation while others are in prepara- 

Observatories for many years, 25 as- tion. Photographic and photoelectric 

tronomers from other institutions spectrophotometry observations of 

made use of the observing facilities planetaries are complementary in the 

at Mount Wilson or Palomar. sense that strong lines can be meas- 

Dr. Lawrence H. Aller of the Uni- ured best by the photoelectric tech- 

versity of California at Los Angeles nique, and weak lines by the 

and Dr. Stanley J. Czyzak of The photographic method. However, the 

Ohio State University observed plane- photographic measurements must be 

tary nebulae with the 100-inch and calibrated photoelectrically. Advances 

60-inch telescopes at Mount Wilson, in the theory of atomic line strengths 

At the 100-inch coude spectrograph and collisional cross sections permit 

they secured plates of NGC 6818, a degree of precision in the studies of 

6886, and IC 4997. With the photo- the physical processes in gaseous 

electric spectrum scanner of the 60- nebulae that puts severe requirements 

inch they observed NGC 40, 650, on the observations. It is now evident 

1514, 2440, 4361, 6537, 6751, 6790, that much more attention will have to 

6818, 6826, 7026, 6905, 7354; IC 1747, be paid to photoelectric intensity 

4846, 5217; and VV 286. Particular measurements. A fundamental limi- 

emphasis was placed on NGC 2440, tation in photoelectric work is set by 

and results on the spectrum of this the spectral resolution, which re- 


quires in all instances a judicious South Wales, Australia, worked as a 

combination of photographic and guest investigator at the Mount Wil- 

photoelectric photometry. son and Palomar Observatories. He 

With the 48-inch schmidt telescope obtained approximately one hundred 

Dr. G. A. Tammann of the University 5 X 7-inch plates with the 48-inch 

of Basel, working on Professor W. schmidt telescope to determine the 

Becker's program, obtained R, G, U- precise optical positions of the coun- 

plates for two fields centered on the terparts of identified radio sources of 

globular clusters M 13 and NGC 4147. small diameter. The plates were sub- 

The availability of faint photoelectric sequently measured on the new 

sequences in these helds will permit proper-motion machine at the Lick 

the ready determination of the lumi- Observatory. All reductions have been 

nosity and density function for field completed and the material is being 

stars down to the plate limit, while prepared for publication. The posi- 

the delay continues in the investiga- tions will be of particular value in 

tion of some other fields taken earlier future calibration of position meas- 

with the same telescope because urements on other radio sources and 

faint photoelectric standards are lack- were made primarily for use at the 

ing. Parkes 210-foot telescope. 

Dr. Roberto Barbon of Asiago About 20 of the foregoing plates 

Astrophysical Observatory recorded were taken in two colors to confirm 

spectra of five bright meteors with suspected quasi-stellar objects. Fur- 

the 18-inch schmidt telescope and ob- ther photoelectric or spectroscopic 

jective prism while working on a observations were made by Kinman 

spectroscopic survey of regions near and E. M. Burbidge at the Lick Ob- 

the galactic pole. All of the spectra servatory and by Arp, Sandage, and 

are of low excitation with the usual Schmidt at Palomar. Included is the 

Xa I, Mg I, and Fe I emission lines. QSO with the largest known redshift 

The interpretation of some bandlike (PKS 0237-22, z = 2.223) and two 

features in the red portion of the others with the second and fifth larg- 

spectrum of three of the meteors is est redshifts (PKS 0056-17, z = 

rather uncertain; explanation in 2.125, and PKS 0229 + 13, z = 2.065). 

terms of N 2 and 2 bands is tenta- These results have been published 

tively suggested. The work of inter- (Arp, H. C., J. G. Bolton, and T. D. 

pretation is in collaboration with Dr. Kinman, Astrophys. J., 1U7, 840, 

J. H. Russell of the University of 1967; and Kinman, T. D., J. G. Bolton, 

Southern California. R. W. Clarke, and A. Sandage, Astro- 

On a 48-inch schmidt plate, taken phys. J., 47, 848, 1967). 

on August 15 to search for faint Twelve 14 X 14-inch plates were 

blue, high proper-motion stars, Bar- taken by Bolton in two colors (GG-13 

bon found a new comet, 1966c. filter, 8-min exposure; UG-1 filter, 

With the collaboration of Dr. A. 60-min exposure) on centers of the 

Braccesi, who provided the radio Palomar Sky Survey in the declina- 

data, and by inspection of three-color tion 0° zone. Five of these plates are 

plates, Barbon identified the follow- of excellent quality. They are being 

ing 4C radio sources with blue stellar examined for objects with marked 

objects: + 03.02, +02.04, +01.02, UV excess and a deep survey at 11- 

-01.06, -01.61, -02.04, and -04.89 cm is being made with the Parkes 

Between August 1966 and Febru- 210-foot telescope in the areas cov- 

ary 1967, John G. Bolton, director of ered by these plates for radio-optical 

the Parkes Radio Observatory, New comparison. 


Danziger of the Harvard College A/mm; the wavelength interval cov- 

Observatory has obtained as guest ered was from 3400 to 7400 A. Special 

investigator with the 200-inch tele- efforts were made to ensure the 

scope lithium abundances in main- highest possible photometric accu- 

sequence stars down to spectral types racy; the interagreement of overlap- 

K0 in the Pleiades and Ursa Major ping spectrograms suggests acci- 

clusters. The results for the Pleiades dental errors of 1% to 2% of the 

support previous, less secure, obser- continuum. The plates were traced 

vations that suggested less depletion with one of the Sinclair Smith micro- 

in the Pleiades stars than in Hyades photometers of the California Insti- 

stars of the same type. This would be tute of Technology, a linear scale of 

consistent with depletion of lithium ordinates being obtained with an ana- 

in stars on the main sequence. logue output system brought from 

Danziger and Dickens are studying Cambridge. An investigation of the 
spectral characteristics of new short- instrumental profile at A6328 in the 
period variable stars and are accumu- second order was made, using a 
lating rotational velocities, gravities, helium-neon laser source, 
and effective temperatures. An at- Dr. Hugh M. Johnson of the Lock- 
tempt is being made to explore the heed Palo Alto Research Laboratory 
variable star gap in the region of observed with the 48-inch schmidt 
short-period variability. telescope at Palomar. Exposures were 

A program is under way to investi- made on the fields of X-ray sources 

gate the physical properties of the en- Sco X-l, Cyg XR-1, Leo XR-1, Lyr 

velopes of shell stars and Be stars — XR-1, and EX Hya. Plates of Cas A 

using low-resolution scans and spec- were exposed in an attempt to detect 

tra obtained at Mount Wilson by the yellow continuum that is to be 

Danziger. expected between the radio-frequency 

In a study of the hydrogen-poor and X-ray extremes of the spectrum, 
star HD 30353, Danziger, Waller- and which may be revealed by high- 
stein, and E. Bohm-Vitense have contrast techniques. Tests for polari- 
used interstellar polarization meas- zation were made of NGC 6302, a 
urements, reddening of neighboring candidate for identification as Sco 
B stars, continuum scans, and models XR-2. One of the 14 t/V-excess stars 
of hydrogen-poor stellar atmospheres in the field of Sco X-l was independ- 
to estimate the most probable values ently found by C. B. Stephenson to 
of reddening, E (B — V) = 0.35, show a peculiar O-type spectrum on 
effective temperature (10,000°K) and Case objective-prism plates, and John- 
absolute magnitude (—3.2). son obtained a slit spectrogram of it 

Dr. and Mrs. Roger Griffin of the at the Kitt Peak National Observa- 

Cambridge Observatories, England, tory, identifying it as the X-ray 

spent four months in Pasadena to ob- source. The Palomar plates gave the 

tain material for a high-resolution first evidence of null polarization for 

high-accuracy photometric atlas of Sco X-l. 

the spectrum of Procyon. The 100- With the 200-inch Hale telescope, 
inch coude spectrograph (114-inch Dr. Philip C. Keenan of Perkins Ob- 
camera and 133B grating) was used servatory obtained 22 coude spectro- 
on 10 nights, and a total of 17 very grams of variable stars and two 
wide spectrograms was secured, using plates of comparison stars on 12 
the second, third, and fourth orders nights during the summer of 1966. 
of the grating to obtain reciprocal This program was undertaken in co- 
dispersions ranging from 0.7 to 1.6 operation with Deutsch and Garrison. 



One of the Mira variables, T Sgr, 
observed on September 3, showed the 
strongest absorption of the line Li 
I A 670$ yet observed in an S-type 
star. The measured total absorption 
of 1.2 A for X6708 in T Sgr was about 
three times as great as that observed 
by Miss Bretz in HD 212667 or in 
AA Cyg, which have been considered 
as the S-type stars in which lithium 
was most abu