m CffiA FOUNDATION COLLOQUIA ON AGEING Vol. 5. The Lifespan of Animals Leaflets giving details of available earlier volumes in this series^ and also of the Ciba Foundation General Symposia, Colloquia on Endocrinology t and Study Groups, are available from the Publishers. CIBA FOUNDATION COLLOQUIA ON AGEING VOLUME 5 The Lifespan of Animals Editors for the Ctha "Foundation G. E. W. WOLSTENHOLME, O.B.E., M.A., M.B., M.R.C.P. and MAEVE O'CONNOR, BjV. ■<(. C^ With 58 niustratioiis and Cumulative Index to Volumes 1-5 LITTLE, BROWN AND COMPANY BOSTON This book is protected under the Berne Convention. It may not be reproduced by any means, in whole or in part, without permission. Application with regard to reproduction should be addressed to the Publishers. J. & A. CHURCHILL LTD. 1959 The Ciba Foundation, a unique international institution, owes its inception to the generosity of CIBA Limited, Basle. However, being established under British trust law, it enjoys complete independence in practice and policy. Under the guidance of its distinguished Trustees, the Foundation offers accommodation to scientists from all over the world at its home in Portland Place. Foremost in its acti- vities is the organization of small conferences, the proceedings of which are published in book form in the manner of the present volume. The Foundation convenes many other in- formal discussions between research workers of different dis- ciplines and different nationalities and each year invites an outstanding authority to deliver a special lecture. An exchange programme between French and British postgraduates is con- ducted and a library service is available. Furthermore, the Ciba Foundation attempts in every other way possible to aid scientists, whether they be Nobel Laureates or young grad- uates making their first original contribution to research. The purpose of the Ciba Foundation, which is to promote international co-operation in medical and chemical research, is symbolized in the armorial bearings by five interlaced rings representing the continents, a black sacrificial cock (emblem of Aesculapius) holding a medical caduceus, and three regular hexagons for chemistry. Its domicile in London is indicated by the red sword of St. Paul and the British lion; the wyvern and the crozier, symbols associated with Basle, refer to the sponsoring firm located in this ancient Swiss town. THE CIBA FOUNDATION Jor the Promotion of International Co-operation in Medical and Chemical Research 41 Portland Place, London, W.l. Trustees The Right Hon. Lord Adrian, O.M., F.R.S. The Rt. Hon. Lord Beveridge, K.C.B., F.B.A. Sir Russell Brain, Bt. The Hon. Sir George Lloyd-Jacob Sir Raymond Needham, Q.C, F.S.A. Executive Council Sir Russell Brain, Bt., Chairman Professor Dr. Dr. h.c. R. Meier Professor A. Haddow, F.R.S. Sir George Lloyd-Jacob Sir Arthur Vere Harvey, C.B.E., M.P. Professor F. G. Young, F.R.S. Director and Secretary to the Executive Council Dr. G. E. W. Wolstenholme, O.B.E. Deputy Director Dr. H. N. H. Genese Assistant Secretary Miss N. Bland Scientific Assistant Editorial Assistant Miss Cecilia M. O'Connor, B.Sc. Miss Maeve O'Connor, B.A. Librarian Miss Lyliane A. R. Treuil vi PREFACE Early in 1954 the Trustees of the Ciba Foundation desig- nated funds over a period of five years for the encouragement of basic research relevant to the problems of Ageing. One feature of this special programme was to be the organization of smaU international conferences, on lines familiarly known at the Foundation in regard to other subjects in medical research, to assemble and consider such information as could be gathered in the youthful work of ageing research. This volume contains the proceedings of the fifth Colloquium on Ageing, and also a combined index to this and the previous four volumes. The Trustees now suspend the Foundation's special stimulation of work in this field. The Director believes that considerable attention has been drawn to the many large gaps in knowledge which must be filled before the postpone- ment and amelioration of senescence can be realized. Research workers should now be left to their task, the size and importance of which cannot be exaggerated. When exciting progress has been made, the Director will, no doubt, be anxious to arrange for further conferences and discussions as part of the general programme of the Foundation. The Director and his co-Editor wish to place on record with these proceedings the indebtedness of the Foundation and themselves to Professor Danielli and Dr. Comfort for giving life to this meeting; to Dr. Genese and Miss Chater for its adminis- tration; to Mr. William Hill for his skill and speed in indexing; and to Messrs. J. & A. Churchill Ltd. and Messrs. Spottiswoode, Ballantyne & Co. Ltd. for minimizing, without skimping or harassment, the delay in publication in this difficult year. vu CONTENTS PAGE Chairman's opening remarks J. F. Danielli . Actuarial aspects of human lifespans by B. Benjamin ....... 2 )/ Discussion: Benjamin, Berg, Comfort, Danielli, Gerking, Gruneberg, Jalavisto, Rockstein, Rotblat, Sacher, Smith ......... 15 Parental age effects on man . by Eeva Jalavisto ....... 21 ./ Discussion : Benjamin, Berg, Comfort, Danielli, Hinton, Jalavisto, Muhlbock, Rockstein, Sacher, Smith . . 31 Studies on the longevity and mortality of English thorough- bred horses by A. Comfort ........ 35 ^ Discussion: Comfort, Hartwig, Kershaw, Rockstein, Smith ......... 55 Lifespan of cattle and horses under various climatic con- ditions and the reasons for premature culling by W. Hartwig ....... 57 Discussion: Benjamin, Bourliere, Comfort, Danielli, Hartwig, Kershaw, Muhlbock, Wolstenholme . . 70 Onset of disease and the longevity of rat and man by H. S. Simms, B. N. Berg, and D. F. Davies . . 72 Discussion: Berg, Bourliere, Comfort, Gerking, Grune- berg Holt, Jalavisto, Muhlbock, Perks, Rockstein, Rotblat, Sacher, Smith, Tanner, Verzar, Wigglesworth 79 Lifespans of mammalian and bird populations in nature by F. Bourliere ....... 90 Discussion: Bourliere, Comfort, Chitty, Danielli, Ker- shaw, Sacher, Scheidegger, Smith, Rotblat . . . 103 Arteriosclerosis in birds by S. Scheidegger ....... 106 Discussion: Berg, Bourliere, Comfort, Danielli, Hinton, Kershaw, Jalavisto, Lindop, Nigrelli, Rockstein, Scheidegger, Verzar . . . . . . .112 ix 76551 X Contents PAGE Relation of lifespan to brain weight and body weight in mammals / by G. A. Sacher 115 ^ Discussion : Berg, Bourliere, Comfort, Danielli, Holt, Lin- DOP, NiGRELLI, ROTBLAT, SaCHER, SmITH, VeRZAR, WiGGLES- WORTH ......... 133 A review of the lifespans and mortality rates offish in nature and their relation to growth and other physiological charac- teristics by R. J. H. Beverton and S. J. Holt .... 142 Discussion: Beverton, Comfort, Danielli, Gerking, Holt, NiGRELLI, ROCKSTEIN, ROTBLAT ..... 177 Physiological changes accompanying ageing in fishes / by S. D. Gerking 181 ^ Discussion: Comfort, Danielli, Gerking, Holt, Nigrelli, RocKSTEiN, Rotblat ....... 208 Longevity of fishes in captivity with special reference to those kept in the New York Aquarium by R. F. NiGRELLI 212 Discussion: Beverton, Bourliere, Comfort, Danielli, Gerking, Holt, Nigrelli, Rockstein, Scheidegger . 226 Factors influencing the lifespan of bees by Anna Maurizio ....... 231 Discussion : Bourliere, Comfort, Hinton, Holt, Kershaw, Maurizio, Nigrelli, Rockstein, Sacher, Wigglesworth 243 The biology of ageing in insects by M. Rockstein ....... 247 Discussion : Berg, Comfort, Gerking, Jalavisto, Kershaw, Rockstein, Sacher, Smith, Tanner, Wigglesworth . 265 The rate of ageing in Drosophila subobscura / by J. Maynard Smith ...... 269 ^ Discussion : Berg, Danielli, Gerking, Gruneberg, Hinton, Kershaw, Rockstein, Rotblat, Sacher, Smith, Wiggles- worth ......... 281 Group Discussion Benjamin, Chitty, Comfort, Gerking, Holt, Muhl- BOCK, Perks, Rockstein, Rotblat, Sacher, Smith, Tanner, VerzAr ....... 286 Chairman's closing remarks J. F. Danielli 298 Appendix Notes on some mathematical mortality models by R. E. Beard 302 List of those participating in or attending the Colloquium on " The Lifespan of Animals ", 14th-16th April, 1959 B. Benjamin B.N. Berg R. J. H. Beverton F. BOURLIERE D. H. Chitty A. Comfort J. F. Danielu . S. D. Gerking . H. Gruneberg , W. Hartwig H. E. Hinton S. J. Holt . Eeva Jalavisto W. E. Kershaw Patricia J. Lindop Anna Maurizio O. MUHLBOCK R. F. NlGREUil W. Perks M. Rockstein J. Rotblat Dept. of Statistics, General Register OflBce, London Dept. of Pathology, Columbia University College of Physicians and Surgeons, New York Min. of Agriculture, Fisheries and Food, Fisher- ies Laboratory, Lowestoft, Suffolk Dept. of Physiology, Faculty de M^decine, Paris Bureau of Animal Population, University Oxford Dept. of Zoology, University College London Dept. of Zoology, King's College, London Dept. of Zoology, Indiana University, Bloom- ington, Indiana Medical Research Council Group for Experi- mental Research in Inherited Diseases, Uni- versity College, London Stade Herdbuchgesellschaft, Stade, Germany Dept. of Zoology, University of Bristol Research Programmes Section, Fisheries Bio- logy Branch, F.A.O.U.N., Rome Inst, of Physiology, Helsinki Dept. of Parasitology and Entomology, Uni- versity of Liverpool Dept. of Physiology, St. Bartholomew's Hospi- tal Medical College, London Bienenabteilung, Eidg. Milchwirtschaftliche Versuchsanstalt, Liebefeld-Bem Netherlands Cancer Inst., Amsterdam New York Aquarium, New York Pearl Assurance Co. Ltd., London Dept. of Physiology, Bellevue Medical Center, New York University Dept. of Physics, St. Bartholomew's Hospital Medical College, London xi Xll List of Participants G. A. Sacher S. SCHEIDEGGER . J. Maynard Smith J. M. Tanner F. Verzar V. B. WiGGLESWORTH Division of Biological and Medical Research, Argonne National Laboratory, Lemont, Il- linois Inst, of Pathology, University of Basle Dept. of Zoology, University College London Inst, of Child Health, Hospital for Sick Children, London Gerontology Laboratory, Inst, of Anatomy, University of Basle Dept. of Zoology, University of Cambridge CHAIRMAN^S OPENING REMARKS J. F. Danielli The investigation of problems of ageing is still in its early stages. One major reason for this is that observations must be made on old animals, which are not readily come by. A research worker in this field must have patience above average, or he will not wait until his animals are sufficiently old. And he must have money above average, or he will not be able to afford to keep his animals sufficiently long. Even where patience is available, the money is usually not. In view of this, we must remember that, where data pre- sented in this colloquium seem inadequate, it is usually the cost of getting better information which is the main restricting factor. This colloquium arose largely as a result of the enthusiasm and initiative of Alex Comfort, and despite the seeming paucity of data the contributors appear to have succeeded in producing a fascinating group of papers. AGEING — ^V — 1 ACTUARIAL ASPECTS OF HUMAN LIFESPANS B. Benjamin General Register Office^ London Actuaries have always been profoundly interested in the variation in the incidence of mortality with age since many of their calculations of contingencies depend upon this variation. The study of this age variation in mortality has given rise to a number of hypothetical "laws" of mortality based on theories about the exertion on the human body of deleterious influences or about the wearing out of components of the body and the exhaustion of living resources. These theories date from Gompertz (1825) who argued on physiological grounds that the intensity of mortality (in his terms the average exhaustion of man's power to avoid death) gained equal pro- portions in equal intervals of age and Makeham (1867) who introduced a constant component as well as a logarithmically increasing component of the force* of mortality as a reflection * It is necessary to define certain functions of the life-table : (1) Ix, the number still living at exact age out of an original generation of Iq births (Iq is called the radix of the table). (2) dx, the number dying between exact ages x and x + 1 {= Ix — Iz+i) (3) px, the chance of surviving from exact age x to exact x + 1 (4) qx, the chance of dying between exact age x and exact x + 1 d-i i-t) Note: p^ + q^ = 1 (5) [ix, the force of mortality. The concept is of an "instantaneous" rate. It is approached by expressing the average rate of mortality at age x (nix) over a finite interval of time as the ratio of (deaths at age X in the interval) to (average population at age x in the interval) 2 Actuarial Aspects of Human Lifespans 3 of the division of causes of death into two kinds, those due to chance and those due to deterioration. It was soon evident that such a relatively simple law would not represent mortality experience throughout life and sub- sequent developments led to the proposal of more complex mathematical relationships between age and the force of mortality and even of different relationships over different parts of the age range. Thiele (1871) for example proposed pL^ = a^ e-^»^ -f a^ ^-^^(^-^>' + ^3 e^^'^ in which the last term is a Gompertz curve to represent old- age mortality, the first a decreasing Gompertz curve to represent the mortality in childhood, and the middle term a normal curve. Perks (1932) introduced a new family of curves in the general form A +5c^ Kc"^ + 1 + Dc' and rationalized this procedure with some interesting specu- lations on the theory of mortality. He found an analogy between the "inability to withstand destruction" of Gom- pertz and the then current physical concept of entropy change — the measure of the time progression of a statistical group from organization to disorganization. Perks also referred to the previous work of Karl Pearson who fitted overlapping curves not to the force of mortality but to the curve of deaths, the curves being intended to represent the mortality of old age, middle life, youth, childhood, and infancy, the causes of death being different in these different periods of life. Perks pointed out that this search for homogeneity in the pattern of causes of death might lead to endless subdivision and then considering what happens to rrix when the interval of time becomes infinitely small. Clearly then the deaths between ages x^ Xi x+1 and X2 in the life-table = J y.x.lx-dx- and dx = j [ix-lx-dx- Xi X The continuous curve of {[ix-lx) is called the curve of deaths. 4 B. Benjamin of the curve of deaths into component curves and that it was questionable whether such subdivision could be theoretically validated. Nevertheless he agreed that one important case of sudden change of composition was the rapid transition between the period of physical growth and the adult period. " In the adult period we appear to be subject to a continually increas- ing disorganisation or 'inability to withstand destruction' while in childhood we appear on balance to gain organisation or 'ability to withstand destruction'." While further work continued on the fitting of mathematical functions to the force of mortality, [i^., attention was focused on the curve of deaths (i.e. of {i^-h) ^s an alternative operand by a paper by Phillips (1935), though this was not yet to be taken up. Meanwhile, taking a closer look at the concept of deterioration. Rich (1940) evolved a theory of mortality based on an analogy between degrees of health and degrees of temperature. This gave rise to a "health frequency distri- bution" (a concept close to the present view of biological normality as a dispersion of characteristics within broadly separated limits), and changes of rating within this distribu- tion could be regarded as forces of deterioration or of recupera- tion. A "natural law of mortality" emerged from which Makeham's and one of Perks's functions could be derived as special cases. Clarke (1950) took up again the analysis of the curve of deaths. He argued that mortality improvements had not extended the natural lifespan but had only allowed more to achieve it. He distinguished between "anticipated" and "senescent" deaths; the ages at death in the latter group were measures of natural lifespans and had a frequency dis- tribution like other animal characteristics. His paper formed the basis of the present work and his hypothesis will be re- ferred to again presently. Beard (1950) also paid attention to the curve of deaths, using the incomplete gamma function as the basic analytical function. Actuarial Aspects of Human Lifespans 5 Later, in a discussion of another paper on fitting a mathe- matical law to l^ of the mortality table (Ogborn, 1953), Perks suggested that reference to the probability models of the biological field and the data of simple populations was the only way in which an advance would be made in the develop- ment of a satisfactory theory of a life-table. It is a pity that this suggestion has not yet been exploited as, if linked with the idea of loss of biological organization, it seems to the present author to open up an important line of approach. But we are straying from the objective of this historical introduc- tion. In 1954 Phillips returned to consideration of the curve of deaths and hypothesized the existence of a basic curve of deaths "to which all curves of deaths are, as it were, striving to attain". Clarke's division of deaths into "anticipated" and "sene- scent" has been further developed by Barnett (1955 and 1958) but applied to the force of mortality, not the curve of deaths. On the basis of cause of death grouping and the actual shape of the curve of observed age rates of mortality Barnett distinguished several different groups of anticipated deaths. We may now return to Clarke's paper. The earlier approach The objective of Clarke's approach was the forecasting of the rates of mortality which would operate in the future. His hypothesis was that every individual carried with him from birth a genetically endowed term of life beyond which it was impossible for him to survive, and that if we knew these terms for every member of the population we could form a frequency distribution similar to that of any other biometric quantity. This distribution would represent a limiting form of the curve of deaths. Clarke went further and supposed that this distribution would not shift as a whole toward later 6 B. Benjamin ages, i.e. that the modal span of hfe was invariant; he re- jected the possibihty as "intangible speculation" outside the practical framework of his study, namely the relatively short- term trend of mortality. Deaths were then divided into two categories, namely those which happened because the term of the lifespan had run out — senescent deaths — , and all others, whether from accident or disease, which were in fact a cutting short of the lifespan — anticipated deaths. He first obtained, therefore, a limiting curve of deaths by constructing a table of mortality in which the q^ (probability of death between age X and age x -[- 1) column would consist of values (g|) appro- priate only to causes of death which could be regarded as senescent. The next stage was to predict the pace at which actual rates of mortality would in a given future period approach those of the limiting table. It is not necessary to discuss this aspect here. Clarke originally intended to define "senescent" deaths by choosing certain degenerative diseases (e.g. cerebral vascular lesions, myocardial diseases, angina pectoris, arteriosclerosis, other diseases of the circulatory system, bronchitis, nephritis), but he naturally found it difficult to select disease groups with sufficiently specific reference to degeneration. Ultimately therefore he arbitrarily assumed ratios of qljq^ rising from 0-05 at age 20 to 0-10 at age 40, 0-20 at age 50, 0-70 at age 70, 1-00 at ages 80 and above. His limiting curve of deaths was not symmetrical. There was a sharp peak at age 80 with a tailing off rapidly on one side to age 100 or so and on the other side a rapid decline to about age 60 and then a much slower tailing off to age 20. The present approach The present author's approach has been even more arbitrary and pragmatic. In a particular life-table the values of d^ (deaths between age x and x -\- 1) have been plotted for every value of x in the table (Figs. 1-3), thus producing an Actuarial Aspects of Human Lifespans approximation to the curve of deaths {d^ is of course discon- tinuous while the "curve of deaths" is continuous. It has been assumed that d^ = [i^^:^.!^^^). The curve has then been treated from its later mode (e.g. the peak at age 76 in Fig. 3) to the upper limit of age as the right-hand side of the distribution of "senescent" deaths, i.e. of normal lifespans, and the left-hand Fig. 1. Curve of deaths. English Life-Table No. 1. 1841. Males. total deaths senescent deaths — — — — — — . — . anticipated deaths side of this distribution has been drawn in (broken line in the figures) to exactly mirror the right-hand side. It is thus assumed (unlike Clarke) that the biometric distribution of lifespans is symmetrical. When the deaths of this left-hand side of the distribution are subtracted from the main curve of deaths the residual (of "anticipated" deaths) tails off (by a broken line in Fig. 3) to zero at the peak of the senescent deaths. In effect it is assumed in Fig. 3 that no deaths before 8 B. Benjamin age 55 and all deaths after age 76 are senescent. It is hoped that this brash sweeping aside of the honest testimonies of millions of medical practitioners which are recorded in the vaults of Somerset House wdll not give offence. It is of course a moot point as to whether any death after age 76, even Fig. 2. Curve of deaths. English Life-Table No. 8. 1910-1912. Males. total deaths senescent deaths _ — ._._ — — ._. anticipated deaths allegedly due to some accident, is other than of senile origin, but some deaths before age 55 are of degenerative origin and at age 55 Clarke assumed that 20 per cent of them were "senescent". Whether this degeneration is senescent in the sense of the completion of a genetically endowed lifespan or whether it is the cutting short of the span by departure from optimum environmental conditions and behaviour is at least Actuarial Aspects of Human Lifespans 9 arguable and it is proposed to adopt the latter hypothesis here. The data used The analysis already described was applied to three national life-tables (for both males and females, though Figs. 1-3 relate to males only): Fig. 3. Curve of deaths. English Life-Table No. 11. 1950-1952. Males. — — total deaths senescent deaths — — . — — . — anticipated deaths English Life-Table No. 1. The first to be compiled by Dr. William Farr and based on the deaths of 1841. It differs from the other two tables which in the case of both males and females start with a radix of 100,000 births, in that Farr to facilitate the product of a "persons" table chose radices of 51,274 male births and 48,726 female births (the proportions 10 B. Benjamin of males and females in the births of that period). All the ordinates of Fig. 1 should therefore be almost doubled to render it comparable with Figs. 2 and 3, though the general shape is correct. English Life-Table No. 8. The first short period table prepared by an actuary, George King, for the Registrar- General and based on the deaths of 1910-12. English Life-Table No. 11. Based on the deaths of 1950-52 and prepared by the Government Actuary. The results The basic results of this analysis are : Senescent deaths Period of deaths Males Females Peak age Standard deviation of distri- bution Propor- tion of total deaths Peak age Standard deviation of distri- bution Propor- tion of total deaths 1841 1910-12 1950-52 years 72 0 73-5 75-7 years 9-38 8-70 7-89 % 39-9 51-5 69-4 years 73-5 760 80-3 years 919 8-51 712 /o 41 0 55-3 70-3 It is clear that from a practical point of view Clarke was right. The main change in the hundred years or so has been the increase in the proportion of people attaining their allotted term of lifespan, from about 40 per cent to 70 per cent, while the shift in the peak has been a mere three or four years for men and seven years for women. Judged by the standard deviation the spread of the distribution has narrowed a little but this is not a substantial change. If — and it is a large question mark — this arbitrary distribution of "sene- scent" deaths can be used as an indicator of the natural Actuarial Aspects of Human Lifespans 11 distribution of lifespan in humans then it appears possible that over and above the large increase in ability to attain the allotted span, the modal span itself is slowly increasing. Even in 1841 there had been a "break through" of the barrier of three score and ten. Women can now talk modestly of "four scores". Comparison between men and women If the proportion of all deaths which are senescent can be used as a measure of ability to survive the allotted term of life, then the improvement in the mortality of women as compared with that of men appears more as a greater shift in the optimum than as a greater improvement in attainment. The advantage of women over men (or conversely the dis- advantage of men as compared with women) is thus a very general one and calls for intense examination. Other aspects for examination It would be possible to forecast mortality not only by examining the medical and environmental possibilities for the reduction of anticipated deaths as Clarke suggested but also by predicting the further change in the parameters shown in the table. It is tempting for example to suggest that the national life-table in 1970-72 will show for men a modal span of almost 77 and an attainment of natural senescence by 78 per cent, but this kind of temptation will be pursued else- where. Another possibility is that the "attainment" proportion (i.e. ratio of senescent deaths to all deaths) might be used as a mortality index for international comparison. This also is outside the present discussion. The limitations of the life -table The life-table as normally computed is based on the rates of mortality experienced by the population of all ages as they 12 B. Benjamin pass through those ages within a short period of term. The population observed is therefore a combination of a large number of generations born over a very long period, as long indeed as the extent of the table. It does not follow that those who passed through age 40 in 1950-52 will experience at 60 (in 1970-72) the mortality rates given at age 60 by English Life-Table No. 11. For actuaries the Life-Table is an experience which they know will never be reproduced but is nevertheless a model which serves to guide estimation of the future expectations. It should therefore be borne in mind that the senescent deaths are of people born long before those whose deaths fall in the "anticipated" category, so that it is not strictly correct to compute the ratio which one group bears to the other as in the "attainment" proportions referred to above. Similarly the peak age of senescent deaths refers to generations born 70 years or so ago and does not indicate the natural lifespan of those who are now in their infancy. However, the indications of secular trend are acceptable. Strictly we should calculate "generation life-tables", i.e. tables each based on the observed mortality of a single genera- tion of births (for practical purposes those born in a narrow interval of years, say five) but the recorded mortality of this country does not permit the calculation of complete tables for generations separated by more than ten to 20 years. Accuracy of age at death In all the kinds of analysis of death rates or of the curve of deaths which are undertaken by actuaries and have been considered above, use is made of a life-table model rather than the actual deaths of a particular year in order to base the analysis on a population with a fixed birth entry, thereby avoiding the irregularities in the run of deaths from one age to another which affect the published deaths statistics of any one period and arise from birth fluctuations or other Actuarial Aspects of Human Lifespans 13 population disturbances. It is sometimes thought that the hfe-table may be inaccurate at very advanced ages because of errors in the stated age of the population and deaths involved. In order to settle this issue a check was made of the ages of alleged centenarians. The Home Office supplied a list of persons (males and spinsters) reaching the age of 100 years to whom a message of congratulation had been sent from Buckingham Palace between April 1956 and June 1958 (married women were excluded because a maiden name would be needed for checking the age and would not be known). This group is somewhat selected but their age accuracy is probably not seriously affected. For the 114 persons (53 males, 61 females) on the list, my colleagues at Somerset House searched the birth registers for the relevant entries of a 100 years or so earlier (no small undertaking). Of these seven could not be found and identification was doubtful in four other cases. This left 103 identified and of these there were 92 cases where the birth entry agreed exactly with the alleged date of birth while 11 showed errors as follows: 1 day younger than stated 4 1 day older ,, ,, 2 ^ 2 days older ,, ,, 1 2 years older ,, ,, 2 5 years older ,, ,, 2 11 Serious errors amounted therefore to only 4 per cent of the total. If anything these figures suggest a slight understate- ment of lifespan. Discussion For the purpose of indicating the tendency for natural lifespans to become longer, or for there to be a more general 14 B. Benjamin approach to some as yet unknown natural lifespan (and it is difficult to separate the two concepts) the present over- simplified analysis may serve well enough. It is indeed argu- able whether the present knowledge of ageing processes justifies a more recondite approach. We may, however, discern the possible lines of future development. If it be accepted that ageing is a process of disorganization — the introduction of the random element — then we may apply, as Perks has suggested (1932), stochastic processes to study first the distribution of ages at which organization gives place to disorganization and, second, the distribution of subsequent lengths of life of the group of lives, w^hich at each age are subject to progressive disorganization. A prerequisite is co-operative study by biologists and statisti- cians of available evidence of the age incidence of the dis- continuity between organization and disorganization in animals. This means shifting attention from death to early signs of degeneration in healthy lives under continuous observation. Summary In the past hundred years or so the peak in the age distri- bution of deaths in the general population has moved to a more advanced age (for men from 72 to 76 years, and for women from 73 to 80), and the proportion of deaths which, on simple assumptions, might be regarded as "senescent" (i.e. of those who attain a predetermined lifespan) has in- creased (for both sexes from about 40 per cent to about 70 per cent). Previous analysis of the so-called "curve of deaths" is reviewed and some suggestions are made for future analysis. REFERENCES Barnett, H. a. R. (1955). J. Inst. Actu., 81, 105. Barnett, H. a. R. (1958). J. Inst. Actu., 84, 212. Beard, R. E. (1950). Proc. Centen. Assembl. Inst. Actu., 2, 12. Clarke, R. D. (1950). Proc. Centen. Assembl. Inst. Actu., 2, 12. Actuarial Aspects of Human Lifespans 15 GoMPERTZ, B. (1825). Phil. Trans., 115, 513. Makeham, W. M. (1867). J. Inst. Actu., 13, 325. Ogborn, M. E. (1953). J. Inst. Actu., 79, 170. Pearson, K. Karl Pearson's Early Statistical Papers. Cambridge University Press (1948). Perks, W. (1932). J. Inst. Actu., 63, 12. Phillips, E. W. (1935). J. Inst. Actu., 66, 17. Phillips, E. W. (1954). J. Inst. Actu., 80, 289. Rich, C. O. (1940). J. Inst. Actu., 70, 314. Thiele, T. N. (1871). J. Inst. Actu., 16, 313. DISCUSSION Danielli: What is the basic mechanism underlying this shift of the death curve to the right — that is to say, why was there an improve- ment in mortality? Benjamin: I cannot hazard a guess. All I could deal with was the observed deaths, and it is a fact that the peak has shifted to the right. This might be because, as Phillips has suggested, successive generations are tending to attain a sort of ideal curve of deaths — that is, a very sharp peak, even further over to the right than has yet been observed. Or it may simply be that more people are in fact surviving to a constant natural lifespan, which means that the curve, instead of being pulled to the left by what Clarke called anticipated deaths, is allowed to grow up more on the right-hand side. Rockstein: Does the initial portion of your curve correspond closely with that for data in other countries ? There seems to be a rather high mortality rate for males in Britain during the early years. Benjamin: I have not yet made much comparison with the life- tables of other countries. I should not have thought it was unusual for the Western countries, at least for Western Europe. Rotblat: The peak for the anticipated deaths seems to become sharper in the 1950-52 curve. Why should this happen ? I would have expected this peak to become flatter and spread over the whole span of life, rather than sharper. If you used a skewed distribution rather than a symmetrical one perhaps you would not get this sharp peak. Benjamin: I am very grateful that you have made that point because this sharpening of the peak in the anticipated deaths is in fact phoney; it is simply an accident of the simplified type of analysis and has no meaning so far as I can see. One can do as Clarke did: make an arbitrary assumption about the proportion of deaths which are senescent and so avoid this assumption of a sym- metrical distribution. But my difficulty is that I do not know where 16 Discussion to draw this line. At the end of this meeting I may know a Uttle bit more about where it should go. Jalavisto: One should perhaps think about three curves — one independent of age, running horizontally, a second one beginning at middle age and rising with advancing age as a result of inadequate living habits, and a third one representing the random distribution of species-specific lifespan. Have you studied the difference between the mortality curves for males and females ? In your symmetrical component curve I would expect the base to be narrower for the females. In the female and male mortality rates the Gompertz rule holds fairly well for males but in the females the assumption of a random distribution around the age of 75 years would fit the facts better. Benjamin: The width of the symmetrical distribution for women is only slightly narrower than that for men. You would like to take a slice off the bottom right across ? Jalavisto : Yes, but furthermore there would be one group which is quite clearly a result of externally induced pathological changes, i.e. through accumulation during decades of carcinogenic substances, of cholesterol deposits following high fat diet, slowly developing deficiency diseases, etc. Of course, they tend to increase the mortality with advancing age because they are just a function of the chrono- logical age. These three groups, as far as I can see, can result in any form of mortality curve according to their mutual relationships. But in any case the end-point as part of a normal distribution curve comes out very clearly, I think. Berg: There is no difference in the nature of the diseases that cause death at age 50 and those that cause death at age 80. The diseases of senescence also occur at age 50. In animals, as well as in man, the so-called diseases of senescence occur in early life. Benjamin: That is precisely what I want to learn. Maynard Smith: The anticipatory deaths you mentioned were presumably deaths due to causes which would kill a person at any age, and not merely if they were old. What worries me is that, whatever those causes are, it is assumed in your analysis that no- body dies from them after the ages of about 60 to 70. If there are causes which will kill people at any age they presumably will kill old people. If you allowed for the fact that old people are dying from accidental causes, as well as young people, there might be no in- crease in the modal age at death in the later life-tables. In other words if you continued the anticipatory death curve throughout the whole period of life, the peak might stay at about 70 to 72 instead of shifting to the right. Discussion 17 Benjamin: Two different kinds of error may be introduced by this over-simplified analysis. I am more worried about my first assump- tion that no one under the age of 55 can die of senescence. I am not so much worried about this second assumption that no one over the age of 76 can die otherwise than by senescence, because if an old person dies of an accident, it is very difficult to know whether they would have had the accident if they had not been old. In other words, although a deaJth appears as accidental in the statistics it may still be a death of senile origin. Danielli: To what extent can you correct your curves for acci- dents ? There must be some proportion of accidents occurring to which the person who was killed makes no contribution; for example there are people who get killed in railway accidents, and as passengers in cars. Then there are other accidents, such as those to car drivers and motor cyclists, to which the individual concerned does make a contribution. These two groups may show some variation with age so that the non- contributory accident, so to speak, would be more independent of age. Benjamin: You could calculate the deaths which are due to accidents to which the individual may have made no contribution so far as these are shown by the certified causes of death. You would of course get rid of some part of that peak of "anticipated" deaths. Comfort: In the curve for deaths of pedestrians in road accidents by age the mode is a very close fit with the curve of general deaths if allowances are made for the different risks to infants. Pedestrian deaths are an excellent measure of general vigour — the power to see a vehicle coming, jump fast enough to avoid it, and recover if it hits you (see Comfort, A. (1957). Ciba Found. Coll. Ageing, 3, 7. London: Churchill). Benjamin: Isn't it true to say that pedestrians do make a contri- bution to the accident ? Comfort: They make a large contribution and that is the point. An inestimable amount of this pedestrian mortality is of an age- distributed, or age-conditioned kind. The same point arises over my horses, when I have to decide which I am going to call natural deaths and which I am going to call non-natural deaths. It is a point where the scoring convention becomes very difficult to determine. Jalavisto: Death from appendicitis might almost be described as an accident. Mortality in appendicitis was formerly a nearly horizontal line, especially in males, and it did not rise much with age. Later on when conditions improved, it can be seen that it is especially young people who escape death from appendicitis. The result is that the curve rises and begins to resemble the form of the 18 Discussion usual curves of mortality seen in nearly any disease. Obviously therefore, even in these age-independent causes of death, the aged do not profit from improved conditions. Sacher: Your analysis of the curve of deaths into three components. Dr. Benjamin, implies the hypothesis that the population is divided into three mutually exclusive sub-groups, each of which is subject to just one of the three modes of death. An alternative hypothesis is that each individual in the population is subject to all three risks. On this basis one should consider, for each individual at each age, the joint probability of dying of these diseases. The basic actuarial func- tion for the discussion of mortality on this probability model is there- fore the rate of mortality function rather than the curve of deaths, for in this model it is the contingent probabilities, the mortality rates, that combine additively. The multiple risk model seems to me to conform better to our intuitive judgments about the nature of the mortality process. The same data that you analysed in terms of the rate of the curve of deaths can be analysed in terms of the rate of mortality function. The conclusion reached is that one component, identifiable with mortality from infectious disease, has decreased markedly, whereas the component identifiable with mortality from degenerative disease (your senescent mortality) has changed very little. The changes in mortality over a period of a century were assignable primarily to the progressive change in numerical value of two parameters (Sacher, unpublished). Gerking: What did you really mean by saying theoretical life- tables are never actually reproduced. Dr. Benjamin ? Benjamin: A life-table is made up of a large number of genera- tions, and people who are dying at the older ages have been born a long while earlier, so that it is only a model. You could construct a generation life-table in which you had only the mortality appro- priate to people born at one particular time, but of course it would take a long while to accumulate this data because you would need to follow a generation right through. So the life-table, which is based on a short period of observation, is actually never reproduced because the environmental conditions are changing all the time. The actuary in normal practice merely uses the life-table as a model to indicate the variation of mortality from age to age, to obtain a basis for his calculations ; he makes no claim that it will be reproduced in the future. I agree with Sacher that you would have to use rates if you were applying probabilistic theory because that is more fundamental, but it did seem to me to be easier to look at this from the point of view of lifespans by stretching out the curve of deaths. If you did Discussion 19 what I suggested that you should do here and took a look at the application of probability theory to people who, as it were, made the change from organization to disorganization, when degeneration begins, you would have to revert to operation on the force of mortality. Danielli: It has been alleged that people exposed to increased radiation have a decreased expectation of life. What would happen to the right-hand peak in your death curves with this particular group ? Benjamin: I don't think there is sufficient evidence yet. Rotblat: The general population has had such a small increase in radiation compared with natural background radiation that one would not expect to see any effect on the death curves. On the other hand we should consider various theories of ageing, for example the recent theory of Szilard, who suggests that we all start with a certain load of faults, or mutants as he calls them, and that these determine the lifespan of a population. If we accept the fact that radiations cause mutations which may influence the lifespan, then one might have expected that in the course of time there would be an increase of these faults within us, because we accumulate the radiation from generation to generation. I would expect, therefore, that the whole curve would gradually shift to the left rather than to the right. Prof. Griineberg, is there an equilibrimn which we may expect to reach and which would take care of this ? How would such an equilibrium be affected by the fact that nowadays people who formerly would have died younger, now live longer because of con- ditions in a welfare state ? Griineberg: I think the effects of radiation fall into two categories. We may expect the ultimate appearance of more or less deleterious mutations in the homozygous condition ; however, as close inbreeding is restricted in man, the appearance of recessive mutations in homo- zygous condition will take a long time. On the other hand, right from the beginning, we may experience the effect of these same mutations in heterozygous condition. It is now becoming increas- ingly clear that at least some of these mutations in heterozygous condition can be advantageous even if they are disadvantageous in homozygous condition. One might thus expect an advantageous result in the early stages of radiation when we are mainly concerned with heterozygotes ; this would ultimately be counteracted by dis- advantageous effects when the mutations become homozygous. It is between these opposite effects that equilibrium is likely to estab- lish itself in the end. 20 Discussion Rotblat: What are the advantageous effects which one may expect from radiation in heterozygous conditions ? Griineberg: It appears that these are improvements of general viability. In Drosophila the chronic irradiation experiments of Bruce Wallace (1957. Proc. not. Acad. Sci. (Wash.), 43, 404) indicate that the exposed population on the whole increased in viability rather than the reverse. The effects on general viability are ap- parently due to hybrid vigour, due to the presence of freshly arisen mutations in heterozygous condition. PARENTAL AGE EFFECTS ON MAN Eeva Jalavisto Institute of Physiology, Helsinki The life expectancy of offspring in relation to parental age has received very little attention. This is easily understood because of the difficulties in the collection of material suitable for such a study. In official vital statistics the dates of birth of the parents are not recorded, and the only feasible method is to compile data from family records. Holmes and Wilson (1925) and Holmes (1928) have collected material from European royal lineage, but although it extends over about eight or nine centuries the data are nevertheless rather scanty. Holmes and Wilson showed that the life expectancy at 20 years is not dependent upon maternal age if families of the same size only are compared. However, on the whole the older brothers tend to live longer than their younger brothers. The same relationship was shown by Beeton and Pearson (1901). Ansell (1874), Yerushalmy (1938, 1939), Gardiner and Yerushalmy (1939), Burns (1942), Tabah and Sutter (1948) and Hoogendoorn (1953), amongst many others, clearly demonstrated that the age of the mother influences the rate of stillbirths and early postnatal mortahty. With increasing age of the mother the death rate of children under one year of age increases steadily even in children of the same birth order (Yerushalmy, 1938, 1945), but the question of the postnatal mortality is controversial, the opposite relationship also being found (Heady et a/., 1955; Knox and Mackintosh, 1958). For references MacMahon and Gordon (1953) may be consulted. Material and Methods The material to be presented was collected from Swedish and Finnish published family records (Jalavisto, 1950, 1951). 21 22 Eeva Jalavisto It comprises 17,986 cases in which the years of birth and death of the children as well as those of their parents were known. The families were therefore completed when recorded. All cases of violent death were excluded. The material goes back to the 16th century, but most of it is from the 18th and the first half of the 19th century. Most cases come from 35 30 - 25 Xr^ Ov 0f / \ s -20 0) / / 1 \ \ \ o / \ \ « 15 . (/ \ \ a. \ \ 10 . •Born before 1830 O .. after 1830 \ 5 — 1 1 i_ %1U 240 25-29 30-34 35-39 Maternal age, years Fig. 1. Percentage of eases in various maternal age groups of children born before and after 1830, respectively. Finnish middle-class and noble families, the rest from Swedish middle-class and noble families. The social level is thus rather high. The life expectancy is calculated as mean age at death minus the years lived. The time of observation must be at least 100 years if correct values for the entire expectation of life are to be recorded. However, this necessarily implies inconsistency in the material, caused by changing environmental conditions during such a Parental Age Effects on Man 23 long time period. In this study, therefore, no attention has been paid to observance of the 100-year span. The expecta- tion of Hfe was calculated for different maternal ages at the birth of the children, notwithstanding the excess of short- lived among those born in the latter half of the 19th century. It was assumed that this would give erroneous absolute values for the total expectation of life but that it would not affect the parental age effects. In order to determine whether 50 40 30- 920 UJ t 10 Total material 50 40 30 o II ■X LU T Born_ before 1830 Total material cT 20h j^All parities 2 Second and subsequent children S2A 25-29 30-36 35-39 S40 ^24 25-29 30-34 35-39 ^40 Maternal age, years Fig. 2. Mean length of life {Ex=o) in various maternal age groups of boys and girls born before 1830 (12,786) and in the whole series (17,980). this assumption was justified the relative number of the mothers in various maternal age groups, in those born before and after 1830, respectively, was calculated. The result is shown in Fig. 1. As may be seen, the distribution is not uniform: among tnose born after 1830 the number of young mothers is smaller than in those born before 1830. The deficit is evenly compensated for by the other maternal age groups. This means that there is a bias in the expectation of life in favour of the progeny of the youngest mothers. Fig. 2 records the expectation of life calculated for those born before 1830 and for the whole material. It seems evident that most 24 Eeva Jalavisto of the decrease in mean length of Hfe is due to inconsistency in the distribution of young and old mothers. Whether other factors were operative, possibly in the reverse direction, during the earlier centuries is difficult to assess. It seems essential, therefore, to avoid calculation of the entire expecta- tion of life and to search for some other method of study. Maternal age: cTlI-^" < 25 years Expectation of life as function of age (years) 10 15 20 ■►AGEjx) 30 ^0 50 60 Fig. 3. Expectation of life (Ex) as function of age in the male progeny of young (<25 years) and old (>40) mothers; first- born excluded. However in one respect the expectation of life calculated for each age and for different maternal age groups is instructive. Fig. 3 shows the expectation of life as a function of age for the progeny of young and old mothers (first-born excluded). It may be noted that from 6 years of age onwards the difference in expectation of life between the progeny of young and old mothers begins to diminish and between the ages of approxi- mately 15 to 40 years the difference is constant. This can be interpreted as evidence that at least in the middle-aged the Parental Age Effects on Man 25 maternal age does not affect the mortality. Therefore, it should be possible to use a group between the ages of 15 and 40 years as a basis of reference when the maternal age effects have to be elucidated. The frequency of births in different age groups cannot be determined directly because of the fictitious nature of the population to be studied. Since only cases of death are recorded, variations in death rate affect the age structure of such a population. Furthermore, since the material comprises data from individuals whose birth may be several centuries apart it would not be possible to use data given by vital statistics. The distribution of births varies in different historical periods and for Finland the births grouped according to the age of the mother are given only from 1871 onwards. The present author therefore thought that since the mortality in the age group 20-29 years is probably not affected by maternal age the percentage distribution of cases of this age would reflect the frequency of births in the dif- ferent maternal age groups. If similar percentage distribution curves are constructed for ages at death below 6 years of age a possible parental age effect would manifest itself in an excess of cases in one of the maternal age groups, and this is what is actually seen. Fig. 4 gives the distribution of cases in various maternal age groups. As can be seen, the curves representing deaths during the first and the second years of life have a dif- ferent course from all the other groups studied. The maternal age group of 25-29 years seems to be particularly favourable for the infants whereas the maternal ages over 35 years seem to have an excess mortality of infants less than 2 years of age. If the distribution of cases is studied by the x^ method it shows that the distribution of deaths of children under 2 years, from 2 to 19, and between 20 and 29 years, of these maternal age groups, differs from a random distribution quite significantly: x^ =25-7 (degrees of freedom = 6) P< 0-001. If, however, deaths at 2-4 years are examined it may be noted that the distribution of deaths in this age resembles more closely the distribution of deaths in the 20-29 year group than 26 Eeva Jalavisto that in the group under 2 years old. Furthermore, if a group is formed containing the cases of 2-19 years it differs from the 20-29 years of age group only slightly, the probability for random occurrence being 0-05 > P > 0-02. Since the age group 20-29 is small, the age group 2-29 years may be used as the basis of reference. 35- 30 25 20 C 15 « a 10 X Dead during first year of life • ■• •• second O ., „ 3rd & 4th D •• between ages 2-29 years ^ 20-29 '• i2U 25-29 25-29 30-34 235 30-34 235 i2L Maternal age, years Fig. 4. Distribution in various maternal age groups of deaths of children under 30 years of age. Total series, 5,590 cases ; 2,735 were born before 1800. It may be asked whether the difference found is a real one, or could be due to differences in recording, i.e. due to the fact that in the latter part of the 19th century the number of recorded infant deaths is relatively greater than in earlier years. In order to exclude this fallacy, the same comparison has been made with cases born before the year 1800. The result is exactly the same except that the number of cases in the 20-29-year group is rather small and gives a somewhat abnormal distribution of deaths. The 2-29-year group is Parental Age Effects on Man 27 therefore to be preferred as an indication for the distribution of births in this population. Since on an average only 29 years separate the deaths occurring during early infancy from this "standard distribution" it is not conceivable that differences in external conditions could affect the result. The conclusion would therefore seem to be that a maternal age of over 35 years increases the mortality of the progeny during the first two years of life. The optimurn age with the least deaths is that of 25-29 years of age. Beyond early infancy, namely between the ages of 2 and 4 years, the effect of advanced maternal age is already quite small or non-existent, and it cannot be demonstrated in older groups, either because there is no influence or because of methodological difficulties arising out of incompatibility of life-expectancy data col- lected from an over-long (at least 100 years) time period. No mention has so far been made of the possible role of paternal age effects. Since maternal and paternal ages tend to be correlated the question is rather intricate. If the material is divided into paternal age groups and a dependence of life expectancy or infant mortality on paternal age is recorded, this may simply reflect the effect of maternal age. If on the other hand no difference is noted, then the paternal effect possibly acts in the opposite sense to the maternal age, high paternal age being favourable for the progeny. Curiously enough, in the total material, in spite of the bias shown to arise from the different distribution of young and old mothers in the material collected before and after 1830, no paternal age effect upon the total expectation of life could be demon- strated. The most probable explanation would be that maternal and paternal ages may after all not be strictly correlated. The rather large mean difference found between the age of spouses (7-10 years) obviously allows fairly wide discordant variations in the material considered here. The same objections can, however, be made against calculation of the total expectation of life in the paternal age series as in the maternal series. Therefore the same procedure for elimination 28 Eeva Jalavisto of external bias arising from the 100-year time span necessary for studies of total life expectancy has also to be applied in the paternal age series. The percentage distribution of deaths at various ages (0, 1, 2-4, 20-29 and 2-29 years) in different age groups of fathers is calculated. The paternal age groups are formed so that the limits are five years higher than in the maternal series, and the 30r 25 20 ^ 15 10 Total 30 c25 X Dead during first year of life • ■■ second O •■ •• 3rd & 4th D •• between ages 2-29 years ^ 20-29 •• <30 30-34 35-39 UO-UU 245 <30 30-34 35-39 40-U 245 Paternal age, years Fig. 5. Distribution in various paternal age groups of deaths of children under 30 years of age. Total series, 5,590 cases; 2,735 were born before 1800. last group extends ten years beyond the highest maternal age group. The result is seen in Fig. 5. It may be seen at a glance that differences in paternal age do not affect the distribution of cases at any age : there is hardly any excess of infant deaths with fathers of advanced ages. The difference between the maternal and paternal age distribution of infant deaths strengthens the impression that the maternal age effects are not artifacts, and that this method of study is practicable. Parental Age Effects on Man 29 Discussion One-third of the total material is made up of deaths under the age of 30 years (5,590 cases). These are, however, divided into several parental age groups and groups with different ages at death. The cases in a group are thereby reduced to quite a small number, always less than a thousand and some- times as low as 200. A further reduction, e.g. resulting from grouping according to birth order, would not give relevant information because of the smallness of the material. Since parity and maternal age tend to correlate it is not possible to study the effects of birth order on infant mortality in this series. However, the main result of this investigation suggests that the maternal age effect is restricted to the first two years of life. It is therefore not necessary to use genealogical material for the elucidation of questions on the effects of parity, etc., since recent statistics are available and more appropriate for this purpose. The difficulties when dealing with family histories are many, and lie mostly in their heterogeneity. The only uniform feature is the rather high social rank of the families. Family records of labourers' or peasants' families, for example, are rarely available. The advantage of the uniformity is, however, lost by the necessarily long time period, usually covering several centuries, needed to compile a population sample of a sufficient size. During such a long period of time famines, epidemics, general hygiene and the standard of living have changed and influenced mortality rates and causes of death. The results concerning maternal age effects are rather con- troversial. Whereas there is agreement concerning perinatal mortality, which increases with age of the mother, infant mortality is found sometimes to increase and sometimes to decrease with maternal age. When family histories have been used for investigation of effects on longevity a slight de- leterious effect of advanced maternal age has usually been found. Gibson and McKeown (1950) have pointed out that 30 Eeva Jalavisto favourable economic circumstances in the population studied tend to eliminate the maternal age effects, which may explain the discrepancy between earlier and more recent data. In spite of the high social level of the family histories on which this study is based they represent a population with a very high mortality. It is quite obvious that if a maternal age effect exists, it is never very pronounced. Therefore it is easily understood that low mortality in favourable conditions does not allow such small effects to manifest themselves. Consequently recent population statistics in western "Welfare States" can no longer be used for detection of maternal age effects. However, there still exist countries with sufficiently low standards of living for such studies. It would be interest- ing to compare records from these countries with data from countries with a high standard of living. Summary Parental age effects were studied in material comprising 17,986 cases collected from Scandinavian noble and middle- class family histories. It was concluded that total expectation of life is not suited for elucidation of such effects, because of the necessarily extended period of observation with resulting inconsistency of the data. Since, however, such calculation made it probable that advanced maternal age had no effect on deaths after the age of 15 years, the group of dead between the ages of 20 and 29 could be used as indicative of the frequency of births in different parental ages. When the dis- tribution of dead during the first and second years was com- pared to that in the 20-29 year group, a slight excess of infant deaths during the first and second years of life was noted in the highest maternal age group ( ^ 35 years). In the maternal age group 25-29 years the relative number of infant deaths was remarkably low, this maternal age being the most favourable for the survival of the offspring. A similar com- parison in the paternal series did not show any significant paternal age effect on the survival of the progeny. Parental Age Effects on Man 31 REFERENCES Ansell, S. (1874.). Quoted by MacMahon and Gordon (1953). Beeton, M., and Pearson, K. (1901). Biometrika, 1, 50. Burns, C. M. (1942). Quoted by MacMahon and Gordon (1953). Gardiner, E. M., and Yerushalmy, J. (1939). Amer. J. Hyg., 30, 11. Gibson, J. R., and McK:eown, T. (1950). Brit. J. soc. Med., 4, 221. Heady, J. A., Daly, C, and Morris, J. N. (1955). Lancet, 1, 395. Heady, J. A., Stevens, C. F., Daly, C, and Morris, J. N. (1955). Lancet, 1, 499. Holmes, S. J. (1928). Univ. Calif. Publ. ZooL, 31, 359. Holmes, S. J., and Wilson, I. (1925). J. Hered., 16, 47. HooGENDOORN, D. (1953). Ned. T. Geneesk., 97, 2130. Jalavisto, E. (1950). Rev. med. Liege, 5, 719. Jalavisto, E. (1951). Ann. Med. intern. Fenn., 40, 263. Knox, G., and Mackintosh, J. (1958). Brit. J.prev. soc. Med., 12, 131. MacMahon, B., and Gordon, J. E. (1953). Amer. J. med. Sci., 226, 326. Tabah, L., and Sutter, J. (1948). Population (Paris), 3, 63. Yerushalmy, J. (1938). Amer. J. Hyg., 28, 244. Yerushalmy, J. (1939). Hum. Biol, 11, 342. Yerushalmy, J. (1945). Hum. Biol., 17, 65. DISCUSSION S acker: Aren't many of the early deaths at present due to con- genital defects and could that be strongly related to parity ? Jalavisto: In young mothers the mortality rate increases according to their parity, but in older mothers that does not matter so much. Sacher: So you do not think that parity is a major factor ? Jalavisto: It might be. Rockstein : There are two distinct ways in which parental age may affect the offspring. There are the direct effects upon the foetus, which might possibly be environmental, and there are those upon the overall longevity of the successful offspring, which might be hereditary. In the housefly, for example, the effect of the advanced age of the parent upon the effective potential of the total population is deleterious, primarily through fewer eggs, of lesser viability, being produced by older parents. On the other hand, where we do have surviving offspring, the longevity of the male housefly is extended considerably (by about 20 per cent). We may likewise have tw^o distinct effects in human populations : where an offspring of an older parent survives, i.e. where an old mother successfully produces young, the young will be longer-lived. The mother in this way may be selecting (or producing a selective effect upon the 32 Discussion population) to the extent that she is herself a long-lived individual. That is, she may be still fertile at the age of 40 or 45 and producing a large family. These offspring will then be long-lived by virtue of this and the net effect will be not to have an adverse effect upon the successful offspring. In other cases, offspring from older parents will be stillborn or malformed at birth. However, Sonneborn has analysed the statistics for a large population of offspring from the New York City vital statistics records. He found that there was a stronger correlation, in a positive sense, between 'paternal age and the number of stillbirths, than with maternal age. This is usually ignored in the statistics, because normally the age of the mother is known rather than the age of the father. Berg: Was the normal lifespan about 45 years in 1800 ? Jalavisto: Yes. Berg: That would be before the onset of diseases of senescence. Causes of death were chiefly physical or infectious in nature. Danielli: There are three distinct factors which might affect the mortality of the children. First, there may be changes in the Mendel- ian characteristics transmitted through the chromosome material. If it were true that there were no parental effects, one would think that no type of deterioration of the chromosome genes was contri- buting to your results. Prof. Jalavisto. Second, there is the possi- bility that you may get cytoplasmic inheritance effects coming in. Then, of course, the actual environment to which the embryo is exposed is also changing as the age of the mother changes, but not as the age of the father changes. So the mother has two chances (or three, if cytoplasmic inheritance effects are included) of altering the expectation of life of the child, whereas the father has only one chance. Maynard Smith: Even in organisms without placentae, e.g. Drosophila, there is plenty of evidence that the age of the female laying an egg will influence the probability that this egg will hatch. In human data one might suppose that the age of the mother, through the cytoplasm of the egg she produces, influences the rate of still- births. There is no difficulty in supposing that either egg cytoplasm or uterine environment might influence the survival during the first two years of life, but they are less likely to influence long-term survival. Danielli: To what extent are more children surviving in recent years from older mothers than was the case 100 years ago ? If more are indeed surviving, the shift to the right of the final curve in Dr. Benjamin's death curves might be due to such an effect. Comfort : There are several social factors which affect the parental Discussion 33 age at first birth, and there must be considerable differences in its distribution now, compared with populations where parenthood was not so optional. Children born to very young mothers now are often also socially underprivileged or illegitimate. But in spite of this, they have a statistically better performance than first children born to predominantly prosperous mothers between 35 and 40 (Baird, D., Hytten, F. E., and Thomson, A. M. (1958). J. Ohstet. Gynaec. Brit. Ejnp., 65, 865). There is also the point in regard to parental age effects that when there are very large disparities in age between husband and wife, the proportion of cases where the husband is not the father increases very considerably. Maynard Smith: There is another point on this genetic question of whether, if more old mothers are having children which survive, this could have a genetic consequence on the expected longevity of a population. It would be very dangerous to assume without evidence that there will be a positive correlation between the longevity of parents and offspring. Beeton and Pearson (1901) found a correla- tion, but it was very small compared with, for example, the cor- relations for stature or for other metrical characters. There are genetic reasons why one might expect, for a character closely associated with fitness, to get small parent-offspring correlations. Certainly Dr. Comfort and I in independent work have found low parent-offspring correlations for longevity in Drosophila. Yet Prof. Rockstein has implied that he gets quite a considerable correlation in the housefly, for the male offspring only. We require better evidence in the human population as to whether there is or is not a high correlation between the longevity of parents and their children. Danielli: We do not really know whether there are significantly more children of older parents surviving, because once there is a probability through advancement in social techniques that a woman of potentially childbearing capacity will survive longer, then her opportunities for not having children, so to speak, also increase and this effect may be working in the opposite direction. So unless we have actual evidence on this point, we have not even got the raw material to find out whether there is any genetic effect at all. Muhlbock: In our inbred mice some of these factors are not apparent. We can control the genetic constitution equally well in all the animals. We observed first that there are more deaths in utero with a higher age and that the litter size is also much smaller. This is not just the effect of the number of eggs, but it is also a uterus effect. Maybe the hormonal stimulation in the endometrium of the uterus is not good enough in old age. Then we tried to find out whether the lifespan of these inbred mice is influenced by the age of AGEING — V — 2 34 Discussion the mother. We did that with the C57 black strain. We considered the Hfespan only of female animals living longer than one year, i.e. half the usual lifetime of a mouse. The fertility age of this strain is up to twelve months. We divided the offspring according to the age of the mother, first three months, then four to six months, and then six to twelve months. There was no difference in the lifespan of these offspring. The environmental factors were the same and there is no difference in genetic make-up. Berg: We have been able to extend the age of fertility considerably in the rat, by regulating the food intake. For example, 700-day-old female rats fed ad libitum have a fertility rate of 12 per cent. On a controlled food intake fertility is increased to 67 per cent. Jalavisto : One feature of these family records is that the families are very big. There are great age differences between the members of one family, although the environment is fairly consistent. Hinton: Your main maternal effect may be restricted to the first two years of the life of the offspring partly because the older mothers have more experience. For example, we have children 17 years apart, and we took care of our later children much better than we did of the first ones. Comfort: Even so I think there is a higher early death rate among first babies born to prosperous mothers over 40 than among those of underprivileged mothers of 18. I doubt if this is entirely a matter of maternal experience. Jalavisto: I have divided the material into first and second children and there is no difference. The second and subsequent children show the same effect as the first children. Danielli: It may just be that the mothers do not bother so much with the later children. Benjamin: The present discussion does seem to tie in with what Dr. Berg said earlier about deaths of children being due to acute infections or other endogenous diseases. This factor of the experience of the mother helping her to deal with infections would seem to be important. You would find great difficulty in making such researches in this country now, because there seems to be a strong tendency for women to have their children very young and very early in married life. Jalavisto: In Finland we have on record what is probably the highest maternal age — a woman nearly 60 years old ! STUDIES ON THE LONGEVITY AND MORTALITY OF ENGLISH THOROUGHBRED HORSES* A. Comfort Department of Zoology, University College, London The General Stud Book records the year of foaling, and in many cases the year of death or disposal, of the thoroughbred racehorses foaled in Britain since the end of the 18th century. This record has been compiled with careful attention to identity, and is greatly superior in quality and quantity to the other non-human mammalian vital statistics so far examined. The possibility of using it as a source of biological data has been recognized before (e.g. Vitt, 1949) but never fully ex- ploited, chiefly, no doubt, because for most purposes each life history must be individually extracted, and statistical treatment of a large sample is therefore very slow work. Since actuarial figures for large mammals are scarce and theoretically important, the Stud Book has been examined to see how far it could be expected to yield useful material for comparative age studies, especially in relation to parental age effects and the inheritance of longevity. Study of parental age effects on lifespan in man is complicated by the high cor- relation between ages of spouses (Sonneborn, 1957); there is no such correlation between ages of sire and dam in horse- breeding, and both mares and stallions commonly remain at stud to advanced ages. One special object of the study was to examine Vitt's (1949) claim that the Stud Book records indicate a large parental age effect on the vigour and longevity * The work described in this paper was carried out during the tenure of a Nuffield Research Fellowship in Gerontology. Part of it received a Ciba Foundation Ageing Award in 1958. 85 36 A. Comfort of thoroughbreds. Another was to see whether the records could be used to determine the extent of the parent-offspring and sib-sib correlations for lifespans. The following account combines and summarizes the results already described elsewhere (Comfort, 1958a, b, 1959a). Materials and Method The Stud Book consists to date of 34 volumes, published at four-yearly intervals since 1808. It is essentially a nominal list of brood mares, giving details of the serving, foaling and ownership of each since the last entry. Arabian mares are listed separately. From this record can be obtained (1) the year of birth of every thoroughbred foal under the name of its dam, (2) the life history of every filly which returns to stud as a brood mare, from her first covering by a thoroughbred stallion until death or disposal from the stud. The life histories were extracted by following each in- dividual animal by name from its first appearance as a brood mare until its last ; in all, including those required to establish parental longevity, and additional lives scored in the course of coat-colour studies, about 10,000 histories were extracted in this way. Initial samples taken were ( 1 ) all the thoroughbred fillies foaled in Britain, excluding Ireland, in the years 1875-80 (Sample A) and 1860-64 (Sample B) which subsequently re- enter the record as brood mares, (2) all the Arabian mares foaled in 1880 and the 35 subsequent years — a compact group of manageable size, where over half the fillies returned to stud. The original six annual cohorts of thoroughbreds (Sample A) were chosen so that their survival period avoided the World Wars. After this sample had been analysed. Sample B was taken to obtain more data upon the relationship of longevity to parental age. The thoroughbred cohorts of 1900 and 1901 were scored later, to see whether secular changes had occurred in the course of the record. Longevity of English Thoroughbred Horses 37 Each life history normally ends in an entry that the animal was shot, died, was exported, or was disposed of (put out of stud, sold, given away as barren). Some end in unexplained disappearance from the record or in "no further return". Each unaccounted absentee was sought by name in each volume to the end of a 32-year period from its year of foaling, and for two volumes thereafter to allow for corrections. The forms used in the terminal entries are "died", "shot", "sold", "sent abroad", and qualifications of these, e.g. "broke leg and shot", "died after foaling" (the last of these implies only that the mare had foaled, not that foaling was the cause of death). The entries therefore discriminate between losses to the record, deaths due to accident, animals destroyed, and deaths presumed to be due to natural causes ; but of animals entered as "died" only those dying at or after foaling, and of animals shot only those which met with an accident, are usually distinguished. Deaths not attributed to accident or shooting account for 1,009 out of the total 2,742 lives in the cohort samples. Only these, which include deaths at or after foaling, were scored as "natural" (d^) in computing the age- dependent mortality, all other losses, and all animals dying at imprecise ages, being scored as lost to the record (a^). Where the last surviving individual is lost to the record, it is assumed to have died in that year. Life-tables from 4 years of age were then prepared by cal- culating the yearly mortality under this convention, with the usual correction for losses. Details of the calculation have been given elsewhere (Comfort, 1958a). Standard errors of the expectation of life were obtained by Irwin's (1949) approxima- tion. Results and Discussion General form of the survival curve Survival curves drawn for the 2,742 thoroughbred mares in samples A and B, and the tabulated figures from which they 38 A. Comfort are derived, are given in Fig. 1 and Table la and b. The curves for the samples coincide closely; they are of typical Gompertzian form, with a high early survival and a steadily increasing force of mortality. The modal age of adult death Table la Life-table for 1,492 thoroughbred mares, foaled 1875-80 Year interval dx «x qx Lx Corrected deaths (10,000) ex 4- 4 11 0-0027 10000 27 17 043 5- 7 49 0048 9973 48 16 04 6- 11 48 0079 9925 78 15 05 7- 16 56 0120 9847 118 14 06 8- 20 69 0159 9729 155 13 08 9- 15 48 0127 9574 122 12 12 10- 18 50 0162 9452 153 11 17 11- 17 45 0162 9299 151 10 23 12- 19 42 0194 9148 178 9 31 13- 22 35 0237 8970 213 8 40 14- 23 43 0265 8757 233 7 51 15- 36 40 0448 8524 382 6 65 16- 33 37 0452 8142 369 5 82 17- 37 34 0560 7773 436 5 02 18- 41 41 0699 7337 513 4 •26 19- 32 49 0639 6824 436 3 •56 20- 31 52 0742 6388 474 2 •90 21- 40 40 1173 5914 694 2 •28 22- 61 38 2328 5220 1216 1 •72 23- 34 33 2054 4004 822 1 •26 24- 24 25 2342 3182 744 0 •90 25- 12 24 2222 2438 542 0 •62 26- 10 11 4082 1896 774 — 27- 2 2 2500 1122 280 — 28- — 1 00 0842 0 — 29- 1 1 2857 0824 241 — 30- — 2 (10) 0601 601 — 31- — — 00 — ■ — n = 566 926 10,000 Ve = 0-046 (Je = 0-214 Standard deviation = 5-5 yrs. Longevity of English Thoroughbred Horses 39 is 22 years. The proportions are very like those of prosperous human survival curves; the time equivalence, obtained by fitting the median and second quartile to the U.S. Census curve for white males, 1941, beginning at the age of 10, is Table lb Life-table for 1,250 thoroughbred mares, foaled 1860-64 Year interval Corrected dx Ox Ix Lx deaths (10,000) ex 4r- 1 00 1-0000 17-314 5- 3 15 0024 10000 24 16 31 6- 4 30 0033 9976 33 15 32 7- 11 55 0094 9943 93 14 23 8- 16 48 0145 9850 143 13- 33 9- 14 80 0136 9707 122 12 35 10- 16 59 0170 9585 163 11 41 11- 18 45 0206 9422 194 10 44 12- 8 50 0099 9228 91 9 50 13- 23 37 0303 9137 277 8 57 14- 25 41 0359 8860 318 7 69 15- 23 25 0360 8542 308 6 82 16- 29 30 0493 8234 406 5 98 17- 25 31 0473 7828 370 5 •17 18- 33 41 0706 7458 527 4 41 19- 32 36 0808 6931 560 3 69 20- 30 33 0910 6271 580 3 •03 21- 33 38 1250 5791 724 2 •42 22- 35 31 1718 5067 902 1 •87 23- 19 22 1407 4165 586 1 41 24r- 22 24 2366 3579 847 1 03 25- 15 18 3000 2732 820 0 •71 26- 6 9 2791 1912 534 — 27- 1 5 1176 1378 162 — 28- 1 1 2222 1216 270 — 29- 1 1 4000 0946 378 — 30- — 1 10 — 568 — n — 443 807 10,000 Ve = 0 044 Ge = 0-214 Median = 22-07 yrs. 40 A. Comfort roughly X 3-2. The last part of the curve calculated from q^ values at ages of 25 and over is, of course, largely arbitrary, and in the late intervals losses, many of them from age- dependent causes, equal or exceed deaths. The survival curves and expectations of the separate cohorts are roughly similar, but there is substantial scatter (Table II), lOOi- SURVIVAL CURVES OF THOROUGHBRED MARES -GENERAL STUD-BOOK AGE IN YEARS Fig. 1. Survival curves of mares foaled in 1860-64 (- -), in 1875-80 ( ), of brood mares by three selected long-lived stallions (Hampton, Galopin and Hermit) ( ), and of grey mares ( ). the last three years of the 1875-80 sample having a shorter expectation of life than any of the others (0-05 < P < 0-02 for the largest difference). The proportion of scoreable lives to total fillies foaled is rather higher in sample B (33-0 against 26-8 per cent) but there is no immediately obvious reason for the differences in cohort performance. There may be bias in sampling, since the two lowest-scoring cohorts of sample A contain an unusually low proportion of late entrants to the record. Mares are normally entered when first covered, but animals among the unaccounted majority which have not Longevity of English Thoroughbred Horses 41 Table II Longevity of thoroughbred mares by years of birth Total Number yeafs scored fillies ( Years) Median 1860 242 726 33-33 17-44 0-21 22-70 1861 255 763 33-42 17-28 0-18 21-86 1862 244 757 32-23 17-33 0-35 21-82 1863 241 783 30-78 16-58 0-21 21-62 1864 268 756 35-45 17-39 0-17 22-03 1860-64 1250 3785 33-03 17-31 0-044 22-07 1875 285 910 31-32 17-26 0-34 22-43 1876 245 863 28-39 17-64 0-21 22-43 1877 233 915 25-46 17-44 0-21 22-49 1878 248 969 25-59 17-49 0-30 22-15 1879 261 952 27-42 16-13 0-16 20-81 1880 221 954 23 17 16-01 0-18 21-81 1875-80 1492 5565 26-82 17-04 0-046 22-18 1900 1910 501 650 1674 1566 33-41 41-50 15-20 1511 <0-l <0-l 19-55 20-65 Arabians 1880-1915 183 314 58-26 18-81 0-43 23-66 Thoroughbred mares — longevity and survival by coat colour Group ^Z = 4 V, Median Bays (1875-79) Blacks (1854-1900) Chestnuts (1875-79) Greys (1845-1920) 568 16-68 0 08 22-03 358 16-53 0-15 21-95 262 17-23 0-20 22-50 200 15-57 0-28 20-43 42 A. Comfort been regularly at stud may enter the record at any age if they produce a thoroughbred foal; some of these have probably been missed. But on the basis of mortality rates calculated for all the cohorts at 4 years of age, omitting subsequent entrants, it appears that this bias is only enough to account for a small part of the difference observed. The 1900 and 1910 cohorts gave much lower figures for mean and median ex- pectation of life than either of the earlier samples. This apparent fall in performance might be influenced (1) by the 1914-18 war: the increased losses are concentrated early in the 1910 table and about 10 years later in the 1900 table, (2) by a change in breeding policy — mares over 23 are rare in recent volumes, being now apparently put out of stud at earlier ages. A secular trend in lifespan cannot be assumed without scoring further cohorts, but there is some prima facie evidence of it here. The mean expectation of life of Arabians was significantly higher than that for any cohort of thoroughbreds {e^^^ = 18-81 ±0-66 years) and the rate of decline slower. These lives are spread too thinly over too long a period for any secular trend to be made out. Maximum age records The highest ages in the series were reached by Arabians, three mares reaching 31 years, and one dying in its 33rd year (born 1911, died 1943; last covered, but barren, 1942); these ages may only indicate more conscientious returns for Arabians past breeding age, compared with thoroughbreds. The two oldest thoroughbred mares in the sample were alive at 30 years. The Stud Book has not been searched in detail for higher records — the oldest mare so far encountered (Blue Bell, by Heron out of Jessie) was foaled in 1851 and died in 1885 at the age of 34. Pocahontas (Stockwell's, Rataplan's and King Tom's dam) was foaled in 1837, died in 1870, and bore her last foal in 1862. The frequency with which such ages are Longevity of English Thoroughbred Horses 43 actually reached by thoroughbred mares is largely determined by human intervention, since many which disappeared from the record at ages of 25 or over were probably capable of living longer — some no doubt did so, dying unrecorded. The stallion Matchem (1749-1781) reached a reputed age of 33; in the obituary lists of the Stud Book one other stallion reached 32, and four reached 31. These ages agree with maximum authenticated records in other breeds (Hokkaido ponies 32 -f, Matsumoto, 1935; Hafling mares, over 32, Schotterer, 1939; Lipitsa horses (J31, 533, Kadic, 1949). Claims of higher ages have been reviewed elsewhere (Comfort, 1956). Many of these refer to ponies, and none is supported by Stud Book records. Thirty- eight years is recorded in a captive zebra (Weber, 1942). Effects of parental age on the longevity of progeny Vitt (1949) has claimed that the longevity and racing per- formance of thoroughbred horses are substantially influenced by the age of both dam and sire, and that impairment of vigour by the use of old breeding stock is cumulative. He found that in a sample of 100 mares from the early years of the General Stud Book, the progeny of dams twelve years old or less developed more slowly, judged by the age at first foaling, and lived longer {6^=4^ =19-5 years) than the progeny of dams aged 13 or more {e^^4^ = 16-4 years). Absolute figures and standard errors are not given, and it is not clear whether the estimates are corrected for losses or based on the distribution of recorded deaths alone. Vitt also compared the fertility and racing form of foals by old and young stallions, and concluded that there was an equally marked paternal age effect, the optimal performance being reached by the foals of stallions 8-16 years old out of mares 6-13 years old. To test this the lives in sample A were distributed (a) by age of dam at foaling, (b) by age of sire at covering, one year earlier, (c) by age of dam at foaling and sire at covering, where 44 A. Comfort these fell in the same grouping interval. Of 1,492 lives, 1,342 were scored and grouped by age of dam, 1,355 by age of sire, and 719 by both, the missing lives among these being scored for one parent only — chiefly the progeny of imported horses, of stallions whose dates of birth were not easily ascertainable from the record, or of mares covered by more than one stallion in the season. The distribution of parental ages is shown in Fig. 2, and the results of the calculation in Table III. DISTRIBUTION OF PARENTAL AGES THOROUGHBRED MARES 1875-1880 DAM • SIRE O AGE:YRS 2 Fig. 2. Mares foaled in 1875-80: distribution of ages of sires at covering (O) and dams at foaling (O) (Comfort, 1958a). Reproduced by courtesy of the Editor, Journal of Gerontology. There was no significant difference in expectation of life between foals of mares under and over 13 years of age ( ^ 12, e^^^ = 16-89; ^ 13, e^^^ = 16-86 years). With further sub- division the progeny of the oldest mares had the shortest lifespans, but the largest difference was less than twice its standard error. Still smaller differences were obtained for the same lives grouped by paternal age alone. Of the 719 lives grouped by age of both parents, those whose dam and sire Longevity of English Thoroughbred Horses 45 Table III Longevity of thoroughbred mares by parental ages (sire at covering, dam at foaling) n €z = i ( Years) Ve Median 1492 17 04 0-046 22-18 1250 17-31 0-044 22 07 Whole sample Foaled 1875-80 (A) „ 1860-64 (B) Mares foaled 1875-80 {Sample A) and 1860-64 {Sample B) Dam (Sample A) : under 8 yrs. 8-12 13-16 ^ 17 Sire (Sample A) : under 8 yrs. 8-12 13-16 17-19 ^ 20 Dam and sire : ^ 12* Sample: A B A + B ^ 13* A B < 9 ^ 16 A + B A B A B Progeny of Hermit, Galopin and Hampton All mares : Got in or after sire's 20th year : Got in or after sire's 16th year, dam ^ 16 at foaling : All mares {A, B and selected sires) by parents $5 16 yrs 154 297 16-90 0-12 21-07 537 16-66 0 09 21-06 303 17-29 0-17 22-21 248 16-15 0-23 22-83 250 16-73 0-16 22-10 537 1711 0-10 22-35 352 17-26 0-20 21-83 151 16-87 0-28 22-08 65 16-87 0-73 21-03 449 17-39 0-22 22-28 531 17-07 0-11 21-90 980 17-33 0-07 22-10 270 16-45 0-23 21-61 150 17-24 0-28 22-64 420 17-03 0-16 22-05 128 17-91 0-22 22-28 220 16-51 0-27 21-50 70 15-71 0-69 20-41 44 17-26 1-10 23-17 ampton 412 16-60 015 21-73 124 16-65 0-33 21-83 41 16-29 0-89 2119 16-45 * Include extreme groups ( <9, > 16). 0-38 2112 46 A. Comfort were under 13 years old lived slightly longer (17-39 ±0-36) than those whose parents were over 13 (16-45 ± 0*48; t ^ 1-4, 0-2 > P > 0-1), and the difference was greater in the extreme segments of these groups (dam and sire ^ 9, 17 • 91 ± 0-47; ^ 16, 15-71 iO-83; t ^ 2-3, 002>P > 0-01). This difference is much smaller than that described by Vitt from maternal age alone, and is of the order of the difference between cohorts. In view of this result, the five additional cohorts (sample B) were extracted and scored for parental age, with the results shown in Table III. The differences found in the 1875-80 sample were not repeated here. The longest-lived group were the progeny of parents of 16 years and over, but the standard error was very large (17-26 i 1-90); the 220 animals which were the progeny of two young parents had numerically the shortest lifespans (16-51 i 0 • 52) ; none of the differences was significant, and all were in the reverse direction to those in the 1875-80 sample. The mean expectations of life were also calculated for the foals sired early and late in life by three selected stallions: Hermit, by Newminster (1864-1890); Galopin, by Vedette (1872-1899); and Hampton, by Lord Clifden (1872-1897), for comparison with Vitt's analysis of the progeny of Swynford. These three stallions produced in their lifetime 141, 119 and 159 fillies which returned to stud. Forty-six, 13 and six of these came from the cohorts already scored, the remaining 347 being new lives. The combined curve of survival for all Hermit, Galopin and Hampton mares coincided closely with that for the original six cohorts ; their mean expectation of life at 4 years was 16-50 ±0-39 years. The 121 mares got during or after their sire's 20th year had a slightly, but not a signi- ficantly, higher expectation than the global mean (16-65 db 0-57) (Table III; Fig. 3). Only 41 mares were got by the three selected stallions in or after their 16th year upon dams 16 years old or more; these had a mean expectation of life of 16-29 ±0-94 years. By combining these mares with the Longevity of English Thoroughbred Horses 47 progeny in samples A and B of parents 16 years old or over, 154 lives were obtained, with e^^^ = 16-45 ±0-62 years, which is less than any of the three global means, but not sig- nificantly so. These results, taken as a whole, seem to afford no good evidence of any consistent effect of parental age on the longevity of mares. lOO 9 PROGENY OF STALLIONS HERMIT. GALORN & HAMPTON ALL (412) 0 0 0 SIRE$20 YRS (121) Fig. 3. Survival curves of lifetime brood mare progeny of Hampton, Galopin and Hermit (O O) and of mares got in or after their sire's 20th year (Comfort, 1958a). Reproduced by courtesy of the Editor, Journal of Gerontology. Correlation betw^een lifespans of parents and offspring Since age of death depends in part upon heritable factors there should be a measurable difference in longevity between the foals of long-lived and short-lived parents, though Beeton and Pearson's (1901) results in man suggest that it would not be large. All the mares in sample A were scored for the longev- ity of their dam, and as many as possible for the longevity of their sire; the date of death could be ascertained only for stallions appearing in the obituary lists, or rather less than half 48 A. Comfort the sires contributing to the sample. Life-tables were made (a) for all the mares in sample A whose dams were known to have reached the age of 25, or died before the age of 14, (b) for mares in sample A whose sires reached 25 or died before 15, (c) for the female progeny of those mares in sample A which reached 25 or died before 14, (d) for 113 mares in the sample whose dam and sire both reached 23 years. The grouping limits in all Table IV Longevity of thoroughbred mares by longevity of parents ( Years) n ex=t Ve Median Mares foaled in 1875-80 Sire reached 25 yrs.* 132 17-54 0-50 22-63 Sire died ^ 14 113 16-27 0-60 22-53 Dam reached 25 yrs. 238 16-96 0-16 22-25 Dam died ^ 13 53 16-35 0-59 21-18 Dam and sire reached 23 yrs. 113 18-07 0-34 23-74 Progeny of mares foaled in 1875-80 Dam reached 25 yrs. 168 16-33 0-31 22-54 Dam died ^ 13 58 16-67 1-25 20-02 Progeny of Hermit, Galopin and Hampton Dam reached 24 yrs. 100 17-11 0-82 21-52 Dam died ^ 18 yrs. 58 15-39 0-83 21-51 • Stallions whose date of death appears in the obituary lists ; mares by Hermit, Galopin and Hampton are excluded from this figure, but included in the figure for dam and sire > 23 years. these cases were fixed to secure enough lives for the calcula- tion; the relative contribution of short-lived mares to the sample was so small that it was not possible to prepare a table for the survival of their fillies by short-lived stallions. The mares by the three long-lived stallions, Hermit, Galopin and Hampton were also grouped by longevity of dam. The calculated means (Table IV) show differences of less than twice the standard error in favour of all the groups with one long-lived parent, except the progeny of long-lived mares in the 1875-80 sample; the 113 mares with two long-lived parents had a mean expectation of life at 4 years of 18-07 i Longevity of English Thoroughbred Horses 49 0 • 58 years, which is significantly more than the global mean, or the mean for any other group. The true difference is more- over likely to be minimized, since nearly half the dams and more than half the sires contributing to the global total died at unkno^vTi ages or from accidental causes, and these losses must include some potentially or actually long-lived pairs. The difference in lifespan of fillies by the three selected stallions out of long- and short-lived dams was about 1 • 3 times its standard error, but the short-lived mothers contri- buted only 53 lives, even when the grouping limit was raised to 18 years, and the comparison means little. In view of Vitt's opinions, a table of the early progeny of long-lived parents was also made, taking all the available mares from all the scored samples w^hose dam and sire were aged 15 or less at the time of foaling or conception, but lived eventually to an age of 23 or more, thus avoiding competition between any age effect and inheritance of longevity. The performance of these mares was in fact numerically but not significantly poorer than that of the groups scored without regard to parental age (e^^^ = 17-09 ±0-89 years). Goat colour Five cohorts (1875 — 79) were scored for coat colour — of 1,271 mares composing them, 588 were bays, 181 browns, 262 chestnuts, 27 blacks, 5 greys, 3 roans and 2 grey-roans, the balance being of doubtful or unstated colour. In order to compare the less common coat colours, further records of greys and blacks were collected from other volumes of the Stud Book. Brown mares were excluded because of the heterogeneity of coat colours included under this description; so were all individuals of doubtful colour, e.g. " black or grey", at first registration. Of the colours examined (Table II), only greys appear to diverge significantly from the means calculated for all mares (P < 0-01) (Fig. 1). There was no significant difference 50 A. Comfort between the longevity of greys in the early and late years of the sample. Large factors of selection may well operate — at many periods grey horses appear to have been selectively exported and they show a high proportion of losses to the record. Most of the apparent reduction in their expectation of life is due to early deaths, the expectations at 10 and 15 years being 10-96 and 3-85 years in greys, as against 11-17 and 2 • 93 years in the whole of sample A. Causes of death are not given, and there is consequently no information about the incidence of melanomas, to which grey horses are sometimes subject (McFadyean, 1933). Longevity of stallions The Stud Book does not contain records of stallions com- parable to those of mares. Three different estimates of thoroughbred stallion longevity have been obtained, with the help of other records — two are based (unlike the mare studies) on cross-sectional samples, and the third is a longitudinal study of the earliest age group for which a list of names could be had. All three leave a good deal to be desired, but they give some provisional indications of the rate at which the expectation of life declines with age in entire males. Cross-sectional samples were taken of (a) all the animals listed in volumes 1 and 2 (1910 and 1913) and (b) all the animals listed in volume 5 (1921) of the Register of Thorough- bred Stallions (excluding the appendix). Thirty-three animals from sample (a) were still alive in 1921 and figure in both samples. The cohort sample was obtained by taking, from the lists of sires of brood mares in volumes 17 and 18 of the General Stud Book, all the stalhons (180) foaled in 1880-84 inclusive. Life-tables were prepared by calculating age speci- fic death rates from death and disposal records in the obituary and export lists of the Register, the General Stud Book, and the lists of premium stallions of the National Hunters and Light Horse Society. Longevity of English Thoroughbred Horses 51 Table V gives the full life-table for the two cross-sectional samples (correcting an error in the L^ column of the 1921 sample as originally published — Comfort, 1959a); and Table Table V Abbreviated life-tables for thoroughbred stallions Listed 1910-1913 Listed 1921 Age n Qx Lx ex oe n qx Lx ex oe 4 9 0 0 1 0000 17-71 ±0-84 2 00 1 0000 19 - 502 ±0-75 5 38 0526 1 0000 12 0 0 1 0000 6 73 0 0 9474 30 0 0 1 0000 7 126-5 0079 9474 38-5 0260 1 0000 8 162 0 0 9399 56 0357 9740 9 195-5 0102 9399 63 0 0 9392 10 220 0227 9303 12-87±0-58 81-5 0 0 9392 15 -6 ±0-53 11 233-5 0385 9092 93 0 0 9392 12 240 0208 8742 111 0 0180 9392 13 244-5 0123 8560 112 0268 9223 14 237-5 0337 8455 110-5 0090 8976 15 217-5 0276 8170 9-31 ±0-59 115-5 0087 8895 10 -13 ±0-50 16 201 0199 7945 110-5 0 0 8817 17 190-5 0210 7787 104 0288 8817 18 176-5 0453 7623 97 0619 8563 19 159-5 0376 7278 89 0562 8033 20 146-5 0956 7004 5 -41 ±0-64 74-5 0402 7582 6 -28 ±0-47 21 122-5 1061 6334 58-5 1026 7277 22 100 1100 5662 44 0682 6530 23 71-5 1399 5039 35-5 0282 6085 24 57 1404 4334 28-5 1404 5913 25 43 2093 3726 3-01±l-ll 18-5 2162 5083 3 -05 ±0-52 26 27 2222 2946 10 1000 3984 27 16-5 2424 2291 8 2500 3586 28 11 3636 1736 6 3333 2690 29 4 2500 1105 3 0 0 1793 30 2 0 0 0829 2 1 0 1793 31 1 1 0 0415 0 0 VI gives the mean further expectation of life at 4, 10, 15, 20 and 25 years for all three samples of stallions, compared with the two main samples of mares. There is no significant dif- ference at any age between the expectation of mares and of the 52 A. Comfort stallions in the cohort sample; the two cross-sections differ significantly from one another, from the mares, and from the cohort. Table VI Expectation of further life (years) for thoroughbred MARES AND STALLIONS AT DIFFERENT AGES. (MeANS AND STANDARD ERRORS.) Stallions Mares Age Listed 1910 and 1913 Listed 1921 Foaled 1880-84 Foaled Foaled 1860-64 1875-80 4 10 15 20 25 17-71 (0-84) 12-87 (0-58) 9-31 (0-59) 5-41 (0-64) 3-01 (1-11) 19-50 (0-75) 15-60 (0-53) 10-13 (0-50) 6-28 (0-47) 3-05 (0-52) 17-29 (0-72) 11-95 (0-73) 8-08 (0-69) 5 01 (0-67) 2-83 (0-43) 17-04 (0-21) 17-31 (0-21) 11-82 (0-21) 11-90 (0-20) 7-80(0-20) 7-98(0-20) 4-54(0-22) 4-76(0-21) 2-54(0-45) 2-60(0-35) We can take our choice among these findings. The cohort sample is closest in date and method of treatment to the samples of mares, but it is small, and depends on only 38 conventionally "natural" deaths, none of them under 8 years of age, while of the two cross-sections, that for 1910-13 is probably the better, on grounds of size and absence of intervals in the middle of the table where q^ = 0. Over most of the lifespan the plot of log qjt for both samples of mares is a presentable straight line with a doubling time of 3|-4 years (see Sacher, this colloquium, p. 115). The present data are too poor for inference about its shape in stallions. Most of the apparent gain in male survival occurs over the years when mares may die of causes connected with foaling. The only valid conclusion from the figures is that contrary to the impression given by the uncorrected table (Comfort, 19596), stallions are not shorter-lived than mares under these conditions of performance. Longevity of English Thoroughbred Horses 53 Conclusions The main use of the study has been in providing a survival curve for a large mammal, sufficiently detailed to be used in criticizing hypotheses about lifespan determinants, for which maximum age-records are unsuitable. More such curves are badly needed. By the criteria needed in dealing with a matter as theoretically important as the supposed paternal age effect the study was too small and the chance of systematic biases, especially in losses, too large for convincing subdivision by parental ages, though the data were better than those on which such effects have sometimes been claimed. The literature of mammalian parental effects on longevity is con- tradictory, and has been reviewed elsewhere (Miner, 1954; Comfort, 1956). Yerushalmy (1939) found an increased still- birth rate in babies with very old or very young fathers, but the difficulty arising from correlation between ages of spouses (Sonneborn, 1957) affects the value of such data. Our study was confined to parental age effects on longevity; paternal age has been held to affect other characters of stock perform- ance, particularly by Russian breeders (e.g. Zamyatin et at., 1946; Isupov, 1949; Ponomareva and Spitskaya, 1953; Pospelov, 1952; Eidrigevits and Polyakov, 1953; Barton, 1951 ; Frankland, 1955). Man and the horse, since they continue to breed into old age, are clearly the mammals in which such an effect on longevity is most likely to be demonstrable; our figures do not bear out the suggestion that old stallions have short-lived offspring, but it might still be desirable to examine the stillbirth rate in the mares which they covered. This, unfortunately, could not be done from stud-book records. The small but positive correlation of filial with parental longevity is in accord with Beeton and Pearson's (1901) work, and with Haldane's (1949) interpretation of it — if heterozygos- ity is an important correlate of vigour, the sib-sib correlation of lifespan should be larger than that between generations. Perhaps the most striking feature of the study is the smallness 54 A. Comfort of the contribution, under these breeding conditions, made by short-hved parents to the next generation. There is, effect- ively, intensive spontaneous selection for longevity, which is not a deliberately-sought character in racehorses. Murie's work on sheep (1944) suggests that for some large mammals the wild adult survival curve is not very different from that in domestication. If this were true of wild horse populations, longevity of the dam, and probably, in the presence of com- petition between males, longevity of the stallion, would have a large selective advantage; the upper limit of the lifespan would have in this case to be fixed by something other than decline of selection pressure. REFERENCES Barton, R. A. (1951). Nature (Lond.), 168, 37. Beeton, M., and Pearson, K. (1901). Biometrika, 1, 50. Comfort, A. (1956). The Biology of Senescence. London: Routledge. Comfort, A. (1958a). J. GeronL, 13, 342. Comfort, A. (19585). Nature (Lond.), 182, 1531. Comfort, A. (1959a). J. Geront., 14, 9. Comfort, A. (19596). IV Int. Congr. Geront., 1, 133. EiDRiGEViTS, E. v., and Polyakov, E. V. (1953). J. gen. Biol., Moscow, 14, 435. Frankland, H. M. T. (1955). J. agric. Sci., 46, 180. Haldane, J. B. S. (1949). Ann. hum. Genet., 14, 288. Irwin, J. O. (1949). J. Hyg. {Lond.), 47, 188. Isupov, A. P. (1949). Horsehreeding, Mosk., No. 4, 29. KADid, M. (1949). Jugosl. vet. Glasn., 2, 602. McFadyean, J. (1933). J. comp. Path., 46, 186. Matsumoto, K. (1935). Z. Gestutk. Pferdez., 30, 127, 165. Miner, R. W. (ed.). (1954). Ann. N.Y. Acad. Sci., 57, 451. MuRiE, A. (1944). The Wolves of Mt. McKinley. Washington: U.S. Dept. Int. Nat. Parks Service. Ponomareva, L. I., and Spitskaya, T. D. (1953). Horsehreeding, Mosk., No. 23, 28. PosPELOV, S. P. (1952). J. gen. Biol, Moscow, 13, 445. ScHOTTERER, A. (1939). Dtsch. landw. Tierz., 43, 228. Sonneborn, T. M. (1957). Proc. Ageing Conference, Gatlinburg. Washington, D.C. : A.I.B.S., in press. ViTT, V. O. (1949). J. gen. Biol., Moscow, 10, 161. Weber, R. (1942). Zool. Gart. (Dusseldorf), 14, 208. Yerushalmy, J. (1939). Hum. Biol., 11, 342. Zamyatin, N., Stolbova, A., Chugaeva, M., and Kuznetsova, G. (1946). Trud. Novosibirsk selskh. Inst., 7, 107. Discussion 55 DISCUSSION Rockstein: Perhaps the small but significant difference obtained for longer-lived animals from the older parents might have resulted because you started with a fairly long-lived strain, and you were dealing with inbred animals. Comfort: This may be so; it has been said of at least one other strain of horses that the longevity depends on the proportion of English thoroughbred blood they have. Rockstein: You said that among thoroughbreds there is a tendency to breed for longevity as well as for racing ability. Comfort: This is not deliberate. It happens because short-lived animals contribute surprisingly little progeny to the total. Mares are not usually covered until they are taken out of training. As you know, horses in flat racing very rarely run over the age of four, whereas horses which are raced under National Hunt rules may go on being ridden up to quite high ages. Most of these animals here, if they had been raced, would have been raced before they were used as brood mares. The successful ones would then have been valued as brood mares; likewise the stallions — as soon as a stallion has made its name as a promising racehorse its value goes up enormously and it will be used to sire just as many foals as can be got out of it during the rest of its life. Maynard Smith: I should like to make a few comments about the genetics of longevity. The consideration that was at the back of my mind in suggesting earlier that there might not be very much cor- relation between parents and offspring was as follows : if a character has been influenced by natural selection for a long time — if there has been natural selection tending to move it in one direction — then most of the genetic variability that is left will not be additive in the genetic sense, and will not give a positive correlation between father and child or mother and child, though it will, of course, do so between brother and sister. Resemblances of the kind that Prof. Jalavisto showed are very similar to the ones which I shall show you later on in Drosophila; that is, there are resemblances between parents and ofl'spring of the same sex, but not of different sexes. This pattern is what one would call sex-limited, and would be expected if the causes of death were to some extent different in the two sexes. On the other hand, if you get a resemblance between mother and son, and father and daughter, as you might in some cases, then this is what would be expected in sex-linked inheritance. 56 Discussion Comfort: I do not think one can get much along those Hnes out of my data. The differences for the two sexes scored separately was very small. The subjects were all fillies, but you have to score both parents to get a significant difference. Hartwig: Have you also studied the influence of parental age on the fertility of the offspring ? Comfort: No. I could probably do so now, by going through the data again. But it would mean following each life from start to finjsh and counting the number of foals. In many instances there is some doubt whether the animal missed or whether it miscarried. The Stud Book usually distinguishes cases where the foal was born dead, or where there was a miscarriage, from those where the mare failed to conceive ; but one would have to be sure of differentiating between unsuccessful pregnancies, and pregnancies which did not take place at all. The other trouble is that as the animals are not crossed twice with the same stallion in succession, one would have to allow for the fertility of the stallion, which varies a great deal. In these thoroughbreds there is a surprisingly high rate of infertility. Kershaw: You showed a death curve [not printed] which starts with a slow rise on the left. The figures which we had on industrial horses, that is draught horses and police horses, show that while the general survival curve is the same, the death curve may have its slow fall on the right (Chalmers, T. A., Kershaw, W. E., and King, J. O. L. (1956). Nature {Lond.),178y4<8). I assume that the arbitrary end-point in the racehorses is in part economic; in those figures of ours for the draught horses and the police horses it was certainly economic. This suggests that the curve of death is arbitrarily determined by the index that one uses. We had assumed that the working life had some relation to natural longevity. It does seem now, in animals for which one can get the same data for different indices, that the different indices may produce curves which are made up differently. Comfort: What was the maximum age for the police horses ? Kershaw: Fifteen to twenty years. Comfort: These thoroughbred mares live a good deal longer. My figures refer to breeding practice during the last century. The maxi- mum age is lower today, particularly in mares from commercial studs, which are put out at 23-24 years, but rather less so in stallions. The number of stallions which "fall dead" and the number of them recorded as having died of old age is large, even now. They are kept, if they were famous animals, very much as pets of the establishment. One of the common, and I suppose the most enviable, terminal entries, is "fell dead after serving a mare". LIFESPAN OF CATTLE AND HORSES UNDER VARIOUS CLIMATIC CONDITIONS AND THE REASONS FOR PREMATURE CULLING W. Hartwig Institutfiir Tierzucht, Martin-Luther-Universitdtf Halle In cattle breeding performance can be separated into two major components : (1) the special productivity (of milk, meat, wool), (2) the general performance. General performance means: fecundity, longevity, power of resistance to disease (particularly to those hereditary diseases caused by failure to adapt to environmental changes), food utilization and so on. Certain factors of the general performance, including longevity, also come under our definition of constitution. Besides the increase in special performance, the improvement of the general performance of agricultural domestic animals is a primary object of cattle breeding; in particular increased lifespan must be considered. In the determination of the real duration of life of large agricultural domestic animals there are, however, considerable difficulties. First of all domestic animals are kept for their economic productivity; their lifespan is therefore affected by economic considerations and is more or less variable. Factors such as decreased performance, hard milking, price relations between milk and meat, shortage of space, period of feeding, bad fodder in some years, technical developments, etc., may lead to the sale for slaughter of completely healthy animals that could have lived longer. That is why, in considering the average age of living animals, we have to deal not with a real biological parameter, as in human vital statistics, but with arbitrarily biased values. Other conditions are to be found, 57 58 W. Hart WIG however, where exterior circumstances are not controlled directly by man (for instance, diseases, epidemics, inferior constitution and its consequences, all influence the lifespan of animals). Here we get approximate real values of the life- span, man and animals in their struggle with nature being exposed to these influences everywhere. The major task in cattle breeding and veterinary science is to analyse the struggle against the various factors which shorten the age of productive animals and contribute to their premature culling. In the following discussion of the age of cattle and horses first of all the average age of living animals and then the recorded average age of herd-book animals of various species will be given. Finally the most important reasons influencing the premature culling of animals will be considered. The lifespan of cattle has been examined in detail in recent years, but there are still only a few results regarding horses. This fact is easily comprehensible because cattle breeding is of great economic importance in many countries, while the importance of the horse has diminished considerably as a result of increased mechanization. Milch cows The research listed in Table I covers about 25 years, from 1932 to 1957. It will be noted that cattle of highland breeds show a longer lifespan, 1-1 • 7 years on an average, than low- land cattle. It is not possible to explain clearly how far this difference is based on breed-conditioned, genetic factors and how far it may be attributed to climatic circumstances, quality of soil or economic reasons. The average lifespan of the different breeds is between 4-7 and 9-35 years and the general average age is 7-1 years. Three years may be sub- tracted for breeding and the productive period therefore amounts to 1-7 to 6-3 years. It is a fact proved by many authors that the productivity of the animals rises according to the number of lactations after the fifth to the seventh calf; Lifespan of Cattle and Horses 59 that is to say the highest productivity is reached between 7 and 9 years, depending on the breed and the individual animal, and then it decreases again gradually. This optimal age of productivity is not attained by the average animal. Table I Average age of living cows of various breeds Breed Lowland cattle Scottish cows English breeds American breeds and Iowa Black-pied Schleswig-Holstein Black-pied, East Prussia Black-pied and red-pied (Rhineland) Black-pied (Rhineland) Various breeds (North-western Germany) Unicoloured red cattle Shorthorn Anglia cattle Red-pied lowland cattle (Minsterland) Black-pied coloured (Saxony-Anhalt) cattle Highland cattle Swiss brown cattle Spotted cattle (Simmental) Spotted cattle (Bavaria) Brown cattle (Wurtemberg) Number Average of cows age Author examined (y^s.) Wright (1933) — Smith, Buchanan and — Robinson (1932) Cannon and Hansen — (1940) Ripke (1938) 4,000 Bauer (1940) — Mannes (1947) 4,000 6-4 (milk- 5-5- fattened) 6 0 4-7 Schieren (1948) Winnigstedt (1949) Winnigstedt (1949) Winnigstedt (1949) Ziegenhagen (1951) Rottgermann (1953) 1910-1950 controlled cows Hartmann (1953) Engeler (1947) Stockklausner (1937) Hagel (1939) Piel and Rumbaur (1948) Spotted cattle (Upper Bavaria) Martin (1950) Spotted cattle (Northern Wurzel (1952) Baden) 4,000 1,455 4,361 30,000 3,000 2,150 166 4 2 81 81 ■4^71 7-2 6-6 6-7 •6-70 7-4 8-5 6-5-7-6 90 9-35 9-23 711 It is easy to see that the average productivity of many cows would be considerably higher if they lived to a greater age. This alarming circumstance has been pointed out repeatedly in recent years. We cannot, however, fully agree with the 60 W. Hartwig opinion sometimes expressed, that the age of the animals would decrease as a consequence of higher productivity. Mannes (1947) and Dietrich (1956) succeeded in proving, in the course of their researches on red-pied coloured cows in the Rhineland and on black-pied cows in Saxony, that long-lived cows produce more milk, even during their earlier years, than those having a short period of produce, and that highly pro- ductive cows live to the greatest age. From the results shown below it can be seen that there has been a general rise in the average lifespan of herd-book animals, even though the pro- ductivity has also risen considerably. Table II Average age of black-pied cattle of the Middle-Weser association of cattle breeders (from bottcher, 1952) No. of Average lifespan Date cows iv^s.) 1930 420 5-5 1935 752 6 0 1940 1,094 5-9 1945 1,618 6-5 1949 1,722 6-9 The rise in the average age in 20 years was 1 • 4 years. Ziegen- hagen undertook the same examination of the Anglia breed (Table III). Table III Average age of Anglia cattle (from Ziegenhagen, 1951) No. of Average lifespan Date cows Yrs. Mos. A: Herd-book cows 1924 9,851 5 10 1929 9,458 6 10 1932 7,797 7 2 1939 8,492 7 10 1947 9,831 7 B : Herd-book and not herd-book 11 1930 15,830 6 1 1939 14,991 6 1 1947 29,350 6 7 Average age Yrs. Mos. 5 6 7 8 11 3 Lifespan of Cattle and Horses 61 The rise in average age of herd-book cows of this breed there- fore amounts to 2 years, 1 month, and cattle not registered in herd-books showed the same tendency. Finally, let us mention the values for grey-brown highland cattle (Table IV). Table IV Average age of grey-brown highland cattle (herd-book) Date 1900 1925 1949 Like the lowland cattle they show an increase in age, amount- ing in this case to 1 year, 5 months. This generally observed tendency to an increase in age is due to improved feeding, to improved methods of keeping and breeding and to the struggle against epidemics and disease. Besides the age of living cattle, the age of cattle at death is of great interest, for this gives a clear idea of the average longest lifespan of the animals. In this case it does not matter w^hether the animals were killed because of insufficient pro- ductivity or whether they died as a result of epidemics, disease or accident. Bottcher (1952) gives the figures shown in Table V. Table V Average age at death of black-pied cattle of the Middle- Weser breed (Bottcher, 1952) No. of Age Date cows {yrs.) 1930 559 6-4 1935 1,482 7-2 1940 2,205 72 1945 3,171 7-9 1949 2,665 8-2 Table VI gives results found by Konig (1951) for another breed. 62 W. Hartwig In both these studies on the age of cows at death we find the same trend as in the ages of living animals already quoted. Although there is a rise in age at death, it still remains true that the majority are culled before they reach the years of greatest productivity. Table VI Average age at death of grey-brown highland cattle (herd-book) (Konig, 1951) Date Yrs. Mos. 1901 6 9 1925 8 4 1949 9 4 Rottgermann (1953) has carried out research on the age structure of red-pied cows in Westphalia and the Rhineland. From 1944 to 1951 an average of 36 per cent of cows were aged up to 5 years, 51 per cent were aged from 5 to 10 years, and only about 13 per cent were older than 10 years. Thus 87 per cent of cattle died before attaining their tenth year. Freudenberg and Francke (1956) found that in black-pied cattle of the Central-German arid region the highest number of deaths occurred between the fourth and the eighth years. At the age of 3, 14-7 per cent of the cattle died; at the age of 5, 21-3 per cent; at 6, 22-6 per cent; at 7, 13-2 per cent; and at the age of 8, 9-8 per cent. Thus 81 per cent of all the cows died in these five years, and the average age at death was 6 • 43 years. This ratio of losses is high, considering that these figures are for herd-book cattle which are valuable for breeding and might be expected to be kept alive longer. The figures for cattle not registered in herd-books reveal still worse results. In this connexion we need to ask what are the causes that have a decisive influence on the age of the animals and lead to their premature culling or death. Freudenberg and Francke (1956) performed special researches along this line on 469 cows from 12 big farms of the arid Central-German district. They found the following reasons for suppression. First, 54*8 Lifespan of Cattle and Horses 63 per cent of the animals were sterile ; the authors pointed out that this is due to genital tuberculosis as well as to Bang's disease, which causes most of the sterility. Secondly, 12-79 per cent of them had tuberculosis (udder tuberculosis included). Thirdly, inflammation of the udder accounted for 7-89 per cent. The remaining 115 animals died from various other diseases, such as cancer of the lungs and cardiac weakness. Dinkhauser's (1940) investigations in Lower Saxony showed that 23 per cent of all cattle had to be culled because of steril- ity. From research in the Central-German dry region Marlow (1951) succeeded in showing that softening of the bone was the most important cause of death. In his researches on cattle from smaller farms in Baden, Gerner (1952) obtained the following results: 1. Sterility 32 • 0 per cent 2. Tuberculosis 14-2 „ „ 3. Swallowing foreign bodies 13-4 „ „ 4. Other diseases 130 „ „ 5. Age 21-0 „ „ 6. Insufficient productivity 6-1 „ „ Other diseases here means those that are specially connected with the sexual organs, such as dropping of the womb and emergency slaughtering at calving. Piel and Rumbaur (1948) studied the causes of death in first-class brown cattle in Wurtemberg. They obtained the following results : 1. Sterility 24-3 per cent 2. Tuberculosis 4-6 59 ?> 3. Brucellosis 1-7 5> 5> 4. Age 23-4 >> 5J 5. Slaughtering without indication of cause 26-0 ?J J5 6. Garget 11-7 >J J> 7. Foot-and-mouth disease 2-9 J5 >5 8. Emergency slaughtering 5-4 JJ JJ 64 W. Hartwig From research into the family history tuberculosis was found in 41 families, six of which had two or more cases. The authors therefore suppose that there is a hereditary inclination to tuberculosis, as the same phenomenon is to be found in human medicine. Martin (1950) performed research on this line on 2,160 cows in Baden. Piel (1951) examined the reasons for culling 2,507 brown Wurtemberg cattle from 1939 to 1944, while Ziegen- hagen (1951) examined the causes of death of 5,244 Anglia cows. These authors found that the causes were more or less the same as described above. All these statements clearly demonstrate that sterility, tuberculosis, garget, brucellosis and softening of the bones are the most important diseases and deficiencies that lead to culling of breeding animals and shorten their lifespans and utilization as brood cattle. Using black-pied cows of various ages from the Central- German dry region, Spohde (1948) undertook research on death from the three most important diseases (sterility, tuberculosis and garget). Of all the cows culled because of sterility, 78 per cent were aged from 4 to 8 years. Sixty-eight per cent of tuberculous animals were culled between the ages of 5 and 7 years; 78 per cent of the animals infected with garget were also culled at these ages. These three principal diseases cause the relatively premature suppression of animals in the second to fifth years of life, that is to say at an age when their maximum capacity has not yet been reached. The different measures that might be taken in order to raise the average age of the animals include breeding, by scrupulous selection, for fecundity, longevity, and intensified resistance to diseases, with simultaneous stress on optimal productivity, as well as improvements in breeding, keeping and feeding. Bulls In his studies on the ages of bulls in Brunswick slaughter- houses Hogreve (1955) found the following average ages: bulls from the South Hanover-Brunswick region, 3 years, 5 months ; Lifespan of Cattle and Horses 65 from Luneberg, 3 years, 4 months; from East Friesland, 3 years, 4 months. He concludes from this low age of death that the premature culling of bulls is due more to private and economic reasons than to the proper physiological limit of productivity of the individual animals. Hartwig (1959, un- published) examined the average age at death of 2,000 herd- book bulls from Saxony-Anhalt. An average age of 4 years, 9 months was noted; 83-5 per cent of the animals had been killed because they were no longer used for covering. The reasons for this were principally economic, as Hogreve had also found. Schroder (1958) found a higher average age of 5 years, 4 months among suppressed bulls kept for fertilization. In insemination stations we find that the bulls are kept by means of better utilization of breed and better conditions so that the animals in general grow older. Schroder found 19 bulls which were older than 10 years. The most important causes of death of the bulls were : culled for genetic reasons 60 bulls = 20-27 per cent diseases of the genital system 59 ,, =19*9 ,, ,, tuberculosis 58 ,, =19-5 ,, ,, diseases of the legs 33 ,, =11*1 ,, ,, Five per cent of the bulls were slaughtered after swallowing foreign bodies and 2 per cent because of viciousness. Here again economic aspects play the leading part, so that these factors make the determination of a real average age impos- sible. Cows and bulls can both reach 18 or 20 years of age, though these figures are exceptional. Horses It is very difficult to make corresponding observations on the regional breeding of horses, therefore these studies were carried out on stallions that were kept as long as they con- tinued to breed. For that reason a better estimate of their real age at death can be made. Research along this line has been undertaken on stallions in the district stud of Kreuz by AGEING— V — 3 66 W. Hartwig Wussow and Hartwig (1956-57). All of the 635 stallions were examined carefully and classified according to breed in order to find the average age of each breed. The following values were noted: Average age {yrs.) 20 East Prussian stallions 15-3 24 Hanoverian 13-6 52 Oldenburg 12-9 The average lifespan of warm-blooded* stallions was 13-8 years. Average age {yrs.) 150 original Belgian and Dutch stallions 13-2 53 stallions of the Rhineland 12*6 42 imported English cold-blooded stallions 10 • 5 284 Belgian stallions, born in Saxony 10-4 The average lifespan of cold-blooded* stallions was 11-5 years. Table VII gives a summary of the ages at death of different breeds. This table shows that well-bred or thoroughbred warm- blooded horses have the longest lifespans. They also have the highest percentage of stallions reaching an age of more than 15 or 20 years, whereas by the age of 15, 81 per cent of the cold-blooded horses have died and only 0 • 7 per cent reach an age over 20 years. Oldenburg stallions, being heavy but warm- blooded, show a lifespan that is between the well-bred warm- blooded and the cold-blooded horses. From the above it seems that thoroughbred and well-bred warm-blooded horses have a better constitution and live to a greater age than do cold- blooded ones. The constitution of the animals seems less good the less thoroughbred blood they have. This tendency can also be found in examining the lifespan of mares. Flade (1958), for instance, while doing research on 64 Arab thorough- bred mares born between 1921 and 1945 in Poland, found an * "Warm-blooded" horses are thoroughbreds; "cold-blooded" horses are heavy ones. Lifespan of Cattle and Horses 67 Table VII Ages of stallions at death 53 thoroughbred and well-bred warm-blooded horses : Age group No. of stallions Percentage 3-5 5 9-5 3-10 22 40-6 3-15 31 58-5 3-20 42 79 0 10 stallions attained an age over 20 years = 19-0 52 heavy warm-blooded horses (Oldenburg and Friesian) : Age group No. of stallions Percentage 3-5 9 160 3-10 21 40-3 3-15 84 63-3 3-20 50 96 0 2 staUions reached an age of more than 20 years = 4*0 534 cold-blooded horses : Age group No. o' stallions Percentage 3-5 63 11-8 3-10 229 42-9 3-15 434 81-0 3-20 530 99-3 4 stallions reached an age of more than 20 years = 0-7 average age of 16-7 years. Konopinski and Detkens (quoted by Flade) found an average age of 11 • 5 years for 598 half-bred Poznan mares in Poland. The most important causes of the premature culling of stallions are, according to Wussow and Hartwig (1956-57), the diseases and constitutional faults shown in the table on p. 68, which were observed in 390 stallions whose causes of death were evident. In summary we can state the following points concerning the average lifespan of cattle and horses : 1. Research on living cows of 19 different breeds revealed an average lifespan of 7-1 years (4 -7-9 -35 years). This average is undesirably low, because the majority of the animals are culled before attaining the years of highest pro- ductivity. 68 W. Hartwig 2. Research on various breeds shows a small rise (1-5 to 2 years) in the average lifespan both of living cows and of cows which died during the past two or three decades. We cannot agree with the often expressed opinion that increased pro- ductivity will cause the premature death of the animals. Stallions Percentage 1. Colic 135 34-0 2. Dermatitis verrucosa (malanders) 74 19-5 3. Disease of the heart, lungs and chest 58 15-1 4. Bad covering 52 13-5 5. Weakness of the legs 30 8-0 6. Sepsis 12 31 7. Viciousness 11 2-8 8. Sleepy staggers (Borna) 1 9 2-3 9. Cancer Total 8 390 1-8 100-0 3. The lifespan of cows is considerably influenced by economic factors, which frequently lead to the premature killing of healthy animals. 4. The most important causes of death that have a great influence on the lifespan of cows are sterility, tuberculosis and garget (mastitis). 5. The average lifespan of bulls is still lower. Economic factors act even more intensively here, so that the animals only attain an average lifespan of between 3 years, 5 months and 5 years, 4 months. 6. The average lifespan of Arab mares in Poland was stated to be 16*7 years and that of Poznan half-bred mares 11-5 years. 7. Warm-blooded stallions at the Kreuz stud attained an age of 13-5 years and cold-blooded stallions of Belgian origin 11*5 years. Lifespan of Cattle and Horses 69 8. The lifespan of horses seems to decrease the less they are related to English and Arab thoi'oughbreds. 9. The most important causes of loss of the stallions were: colic, dermatitis verrucosa (malanders), infirmity of the heart, chest and lungs, and bad covering. REFERENCES Bauer, K. (1940). Z. Tierz. ZuchtBiol, 47, 261. BoTTCHER, T. (1952). Volkerode, Sonderdruck, 18. Cannon, C, and Hansen, E. (1940). Zuchtungskimde, 15, 193. Dietrich, H. (1956). Dissertation, Halle. DiNKHAUSER, F. (1940). Dtscli. landw. Tierz., 44, 153. Engeler, W. (1947). Das schweizerische Braunvieh. Flade, J. E. (1958). Arch. Tierz., 1, 354. Freudenberg, F., and Franke, G. (1956). Dtsch. Landw., Berl., 7, 301. Gerner, K. (1952). Tierzuchter, 4, 589. Hagel, L. (1939). Dissertation, Munich. Hartmann, W. (1953). Dissertation, Halle. Hartwig, W. (1959). Unpublished results. HoGREVE, F. (1955). Tierzuchter, 7, 60. Konig, K. (1951). AUgau. Bauernblatt. KoNOPiNSKi and Detkens (quoted by Flade, 1958). Mannes, a. (1947). Dissertation, Hannover. Marlow, H. (1951). Dissertation, Halle. Martin, W. (1950). Ziichtungskunde, 22, 97. PiEL, H. (1951). Ziichtungskunde, 23, 79. PiEL, H., and Rumbaur. (1948). Dissertation, Hohnheim. RiPKE, G. (1938). Dissertation, Munich. RoTTGERMANN, W. (1953). Z. Ticrz. ZiichtBioL, 62, 1. ScHiEREN, J. (1948). Dtsch. tierdrztl. Wschr., 55, 49. Schroder, G. (1958). Thesis, Halle. Smith, Buchanan, A. D., and Robinson, O. J. (1932). Z. Tierz. ZuchtBiol, 25, 307. Spohde, H. (1948). Thesis, Halle. Stockklausner, F. (1937). Dtsch. landw. Tierz., 41, 570. Winnigstedt, R. (1949). Ziichtungskunde, 20, 193. Wright, N. C. (1933). Ziichtungskunde, 8, 424. WuRZEL, W. (1952). Zuchtungskunde, 23, 239. Wussow, W., and Hartwig, (1956/57). Wiss. Z. Martin-Luther -Univ., 6, 13. Ziegenhagen, G. (1951). Z. Tierz. ZuchtBiol, 59, 331. 70 Discussion DISCUSSION Benjamin: The improvement in average lifespan in the more recent groups you mentioned was very much smaller than in the cattle in earlier years. Is there some sort of resistance to increasing the lifespan or is there some other factor involved ? Hartwig: The overall tendency was towards an increased lifespan, and the smaller increment there is not significant. Bourliere : Is there any correlation between the average lifespan of the different breeds of cows and their weight ? Hartwig: The lighter cattle live longer than the heavy ones. Danielli: Have you any data for highland cattle living in the lowlands, and vice versa ? Hartwig: There is the example that when cows were exported to Africa the highland cows adapted themselves better to those condi- tions than the lowland ones. Kershaw: When were these cattle exported ? Hartwig: At the end of the 1920's. Kershaw: In that case there is a complicating factor, because I have seen their progeny in the Cameroons and they are remarkably resistant to sleeping sickness. Danielli: To what extent is this increase in average age due to better veterinary services? Hartwig: That is too complex to answer and I am unable to decide the cause of it. Danielli: Unless one has some idea of the extent to which some specific disorders have been suppressed by veterinary work, it is very difficult to evaluate the data at all. Hartwig: The veterinarians claim that they helped towards this increase, but on the other hand the breeders say it is due to their work. Comfort: Has it been possible, in your data on horses, to compute lifespans or life-tables in the same sort of way as I have done, on the basis of age-specific natural death rates allowing for the animals lost or culled ? Hartwig: Yes, that has been done. Comfort: The agricultural lifespans are a very different matter indeed from those of horses, which are kept to advanced ages. I am sure these cows would live a great deal longer if they were not culled. Hartwig: I agree with you. Comfort: It would be very useful if we could get a good life-table for one of the heavy breeds for comparison with thoroughbreds. Discussion 71 Hartwig: The great difficulty is that if these animals are not kept for breeding purposes they are mostly sold and records are very difficult to obtain. MiXhlhock: Cancer research workers are very interested in keeping cattle to the end of their lives, because no one has ever seen a mam- mary tumour in cattle. The question is are they naturally resistant, or are they slaughtered before they reach the age at which tumours would develop ? A considerable amount of money is now being collected just to keep cattle to the end of their natural lives to see what happens. There is another cancer in cattle which is very peculiar, and that is the so-called "cancer eye", or cancer of the conjunctiva. It is a very interesting condition, hardly ever seen in Europe, but more often found in countries with plenty of sunshine. It is thought to be a virus infection, and studies are being carried out in Texas where a great number of cattle are kept just for this purpose. Gerontologists could therefore find out from this material what the normal lifespan is. Wolstenholme : Could one obtain any figures from the cattle in India, which are allowed to live their normal lives to the full ? ComfoH: There is an institute in India which studies these cattle, but their ages are not known, and unless one has reliable stud-book records and the actual date of the calving it is no good. If I was un- justifiably sceptical about centenarians who have birth certificates, I would be ten times more sceptical about a sacred cow which has not. Wolstenholme: Have any sacred cows been seen to have mammary tumours ? MiXhlhock : They are in such a poor condition that it is no wonder that they do not get any. ONSET OF DISEASE AND THE LONGEVITY OF RAT AND MAN Henry S. Simms, Benjamin N. Berg, and Dean F. Davies Department of Pathology, College of Physicians and Surgeons, Columbia University, New York It is self-evident that the longevity of any species, such as the rat or man, depends upon the age range at which the major diseases of that species result in death. This, in turn, depends upon three factors : First, the ages at which early lesions of the major diseases are most likely to appear in individuals of the species. (This we call the "probable age of onset.") Second, the time required for lesions to develop from the early stages to the severe lesions that cause death. Third, the effect of advancing age on the onset of lesions in that species. The combined action of these three factors results in the mortality curves with which we are familiar (Fig. 1). One hundred and thirty-four years ago Gompertz (1825) published his law relating mortality with age. More than one hundred years later this law was rediscovered by one of the present authors — and this started him on a career in gerontol- ogy. The law may be expressed in the following form : Log Pm - Log Pmo = ^W where P^ is the probability of death (mortality rate) at age t, P^Q is the (extrapolated) probability of death at the age of birth, and k^ is a constant having a positive value. Fig. 1 shows a plot of the logarithm of mortality rate against age. This is seen to be approximately a straight line through- out adult life. The equation of this line is the one just given. 72 Onset of Disease and Longevity of Rat and Man 73 For nearly 15 years we have been studying the onset of lesions in rats in relation to age and longevity. For the purpose of these studies we established a special rat colony with conditions that were unusually favourable and uniform. The temperature was kept constant (76° f). The humidity Avas kept at about 60 per cent. The lighting was indirect and .o^s^ V Fig. 1. Logarithm of human mortahty, plotted against age (From Simms, 1946). Figs. 1-5 reproduced by courtesy of the Editor, Journal of Gerontology. was the same in all cages, with 12 hours of light and 12 hours darkness each day throughout the year. The diet was uniform and the quarters were quiet and clean. Data on growth and disease in rats kept under these conditions have been reported by Berg and Harmison (1957). Finally, we were able to reduce respiratory infection (and other infections) to such a low level that they were practically non-existent. Under these conditions it was possible to study 74 H. S. SiMMS, B. N. Berg, and D. F. Davies the onset of lesions (of a non-infectious nature) in relation to age and in relation to mortality and longevity. Fig. 2 shows the percentage of male rats that showed detect- able lesions of five selected diseases (Simms and Berg, 1957). Each curve is an S -shaped curve approaching a maximum level of incidence. Two of these curves approach 100 per cent, whereas the other three approach lower levels of incidence. It will be noted that the curve for chronic nephrosis and glomerulonephritis and also the curve for myocardial degener- ation are spread over a wide age range. This means that onset 100 PERCENT 80 (,0 """■^ ADENOMA OF - 40 20 PITUITARY ^::^:^^ 0 100 200 AGE IN DAYS 300 400 500 600 700 B05 900 iooo" 1100 Fig. 2. Percentage of rats having detectable lesions of five major diseases, plotted against age at autopsy (From Simms and Berg, 1957). of these lesions was observed in some rats at an early age but in other rats it was not seen until they were much older. On the other hand, onset of muscular degeneration occurred in all rats within a narrow age range (lasting only about 500 days). In the majority of the rats the onset of detectable lesions of this disease occurred between 700 and 900 days of age. Hence our rat colony was quite homogeneous in regard to this disease — but much less homogeneous in regard to other diseases. Fig. 3 shows the slope of these curves at various ages. Each curve in this chart represents the age distribution of onset of a disease. For example, muscular degeneration (Berg, 1956) had Onset of Disease and Longevity of Rat and Man 75 50 PERCENT i 1 PER 100 DAY INTERVAL 40 -100/il 1 1 1 1 ^1 1 ImuscI Ioegen\ ! 30 — 20 CHR NEPH. /jPER! \ 10 j^W^ )WA \ ^ 0 1 r<^ "^^^l^S "t ^N" 1 AGE 100 200 IN DAYS 300 400 500 600 700 800 900 1000 1100 Fig. 3. Age distribution of onset of rat lesions. These curves show the probabihty of onset at various ages (among the total number living at each age) (From Simms and Berg, 1957). its greatest probability of onset at 750 days of age, although a few rats acquired lesions at 500 days — and some not until 1,000 days. Similarly, the peak for chronic nephrosis and glomerulonephritis was 580 days and the peak for periarteritis was 710 days. 0 300 200 AGE IN DAYS 900 1000 Fig. 4. Probability of onset of new lesions among the surviving rats having no lesions (From Simms and Berg, 1957). 76 H. S. SiMMS, B. N. Berg, and D. F. Davies The reason why the curves fall off after reaching a maximum is simply that there are fewer remaining animals without lesions of a given disease and which consequently can acquire lesions for the first time. There is no decrease in the tendency to form lesions among those individuals that have survived to advanced age without them. This is shown in Fig. 4, where it is seen that the tendency to form lesions increases progressively with age. — . AGE IN DAYS 0- 100 200 — 1 r- ^iL5 3di ^>- ^20. Fig. 5. Logarithm of the probabihty of onset of new lesions (among surviving rats having no lesions). Also a curve for rat mortality (From Simms and Berg, 1957). That this increase is a logarithmic function of age is shown in Fig. 5, where the logarithm of the probability of onset (among rats not having lesions) is plotted against age. These curves approximate to straight lines. Note that they parallel the bottom curve for mortality of rats. We may now compare rat and man. Fig. 6 contains two charts each having data from both rat and man. At the left end of the top chart are curves for occurrence of lesions in rats. These are identical with the curves in Fig. 2, except that here the time scale is very much condensed, making the curves much steeper. Similarly, at the left end of the bottom chart are curves for the distribution of onset of lesions in Onset of Disease and Longevity of Rat and Man 77 rats. These are identical with those in Fig. 3, except that the time scale is here very much condensed. At the right-hand end of these charts are data on man — plotted on the same time scale as the rat data. These data were obtained by one of us (Dr. Davies) in 1945-46 while working in this department. Examination was made of over 500 100 80 o 40 o " 20 0 RAT MAN ,\hr. neph. •^myo. degen. ^^periarteritis PERCENT OF INDIVIDUALS HAVING LESIONS 10 20 30 80 90 AGE IN YEARS RAT At r 100 DAYS MAN At =10 YEARS 40 - 20 ^MUSC. DEGEN. XHR. NEPH. PERIARTERITIS rMYO. DEGEN. DISTRIBUTION OF AGE OF ONSET HYPER. PR05T 0 10 AGE 20 30 YEARS COR. THROM. 90 Fig. 6. Data of rat and man plotted on the same age scale. Top Chart : Percentage of individuals having detectable lesions of diseases of their species. Bottom chart : Age distribution of onset of lesions (probabiUty of onset). autopsy records of the Presbyterian Hospital. The data were tabulated and some of the findings on males are reported here. We recognize that such data are open to the criticism that the autopsies were performed by numerous pathologists and that the cases came from a very heterogeneous population. How- ever, these objections apply equally to all the age groups reported. Unfortunately, the number of cases in the upper 78 H. S. SiMMS, B. N. Berg, and D. F. Davies age groups was small (56 males in the 70-79 group and 12 males in the 80-89 group). It will be seen from Fig. 6 that when the data of rat and man are plotted on the same time scale, the curves of the two species are identical in form, except that those of the rat occur much sooner, and are compressed into a much shorter time range, than those of man. This applies not only to one or two diseases of each species, but also to other diseases not reported here. It should be pointed out that the rat data were obtained from a fairly homogeneous colony of animals raised under uniform conditions. On the other hand, the human data were obtained from a heterogeneous collection of individuals who had lived under differing conditions. Had the humans been as homogeneous as the rats, their data on these two charts would be characterized by much steeper curves within a much narrower time range — perhaps approximating to the curves of the rats in shape, but not in their location on these charts. Hence, we may conclude that the factors which determine the longevity of these two species (rat and man) appear to be identical, except that in one species the lesions of the major diseases have their onset much sooner and over a much shorter range of time, than do those of the other species, when compared on the same time scale. That it is possible to alter the age of onset of lesions has been shown by dietary experiments in which rats on a re- stricted food intake were found to have a considerable delay in the onset of their major diseases, as compared with rats receiving as much food as they wanted to eat. This suggests that there may be other methods for modifying lifespan which may result in altering the age of onset of lesions of major diseases. Summary Observations on the accumulation of lesions in rats in relation to age have shown that the lifespan of these animals Onset of Disease and Longevity of Rat and Man 79 is directly related to the age of onset of lesions of the major diseases of this species. The age of onset of these lesions, when plotted against age, gives smooth curves which are character- istic of the diseases in question. Data on human lesions, found from autopsy data at dif- ferent ages, give curves similar to those of rats, except that the onset of the human diseases does not occur until many years later than is the case with the rats. This indicates that there is a mechanism for the deferment of onset of lesions, which accounts for the difference in lifespan between the species. REFERENCES Berg, B. N. (1956). J. Geront, 11, 134. Berg, B. N., and Harmison, C. R. (1957). J. Geront.y 12, 370. GoMPERTZ, E. (1825). Phil. Trans., 115, 513. SIMMS, H. S. (1946). J. Geront, 1, 13. SIMMS, H. S., and Berg, B. N. (1957). J. Geront.y 12, 244. DISCUSSION Gerking: The curve showing the age distribution of onset of dis- ease in your rats was a normal, or nearly normal distribution. The right-hand portion of the curve of deaths that Dr. Benjamin showed us, indicating senescence in the human, was also the normal distri- bution. These both occurred late in life and they may actually reflect the same phenomenon. Berg: That is true. We find the same type of distribution curve for the onset of lesions as for mortality — in both rat and man. This is because the period for the full development of the disease, from the time of the early lesion to the late lesion, does not change with age. Gerking: The correspondence between the shape of these ciu'ves, over a wide range of ages, is very remarkable. GriXneherg: I have some confirmatory evidence on what you said about the relation between the age of onset of the disease and the environment. In mice of the A strain, in animals above a certain age, nearly all individuals suffer from the deposition of a substance called amyloid in the kidneys and various other organs. A group of in- vestigators in the National Institutes of Health, Bethesda, has suc- ceeded in influencing this condition by a change of diet. If these animals are fed on a protein-poor diet the onset of the condition is delayed and the total incidence is greatly reduced. Still more in 80 Discussion keeping with your results is the fact that if, in addition to the reduction in proteins, the total intake of food is reduced, amyloidosis in this strain can be made to disappear almost completely. In the A strain of mice there are three different entities which tend to kill these animals: (1) high incidence of mammary tumours (but only in breeding females), (2) high incidence of lung tumours, and (3) the deposition of amyloid. So far as I know the appearance of the other two conditions cannot be easily retarded. To what extent were your rats inbred in these experiments, Dr. Berg ? Were they genetically homogeneous or were you dealing with a mixed colony ? Berg: They were not homogeneous but were closely related. We have two lines of rats that have been randomly inbred for 65 generations. These derive from a small group of Sprague-Dawley rats acquired in 1945. Obviously, if the rats were more homogeneous the curves would be steeper than those we have shown. Qruneherg: That being so, additional information might be ob- tained by using highly inbred strains of mice, which have very dif- ferent lifespans. Some strains die early because they have an early onset of manunary tumours, such as the C3H strain; and the I strain has a short lifespan because it develops lesions in the stomach. Other strains differ in lifespan for reasons which are not yet fully understood. In addition one can use homogeneous but not inbred material (first generation hybrids) and segregating material (Fg generations) and so on. It should also be possible to subdivide the causes of mortality further by systematic changes in the environ- ment, particularly change in diet, as you discussed, but also by dif- ferences in crowding, temperature, etc. If this were done it would probably turn out that there is no unique life expectation of the species ; presumably each genotype and each type of environmental situation has its own expectation of life, and in man the survival curve is presumably a superimposition of a whole family of curves. Muhlbock: Amyloidosis is one of the major diseases in mice. In different inbred strains there are differences in incidence of amyloido- sis. In some strains there is a very high incidence at 18 months of age, whereas in other strains it is nil, or very low. Therefore one should investigate a number of different strains before generalizing and saying that for that species this is the age of onset of the disease. Berg: You are quite right. Our results apply only to our particular Sprague-Dawley strain, maintained under our conditions. Sacher: Amyloidosis has been under investigation in our laboratory by Dr. S. Lesher (1957. J. nat. Cancer Inst, 19, 1119). He finds a high incidence in the A strain and in F^ hybrids with an A-strain Discussion 81 parent. Amyloidosis in mice is a disease with maximum incidence in middle age. Incidence falls to zero in later life. Thus it resembles some diseases in man, particularly some liver diseases. If a single X-ray dose is given to the young animal, this whole sequence is moved to the left on the time axis so that a given incidence of amyloidosis is seen at an earlier age, and about the same total incidence is seen. Perks: I was also struck by the symmetry of the curves of distri- bution of onset of the different lesions in your rats, Dr. Berg. We should bear in mind that when you combine symmetrical curves of this kind and get a curve of onset of lesions of all kinds, you do not necessarily reproduce symmetry. Further, as the mean delay between onset and death is significantly different for the different lesions, the final death curve could well be highly skew as compared with the component symmetrical curves of onset. The other point I was particularly interested in was the delay in onset for females. My mind always goes to the question of the useful- ness of statistics, and in the life assurance world we are seriously concerned with the differential mortality between males and females, particularly in view of the growing volume of pension business. In this country there is a difference of about five years in effective age between males and females at the older ages. In some of the Scandi- navian countries there is a considerably narrower difference and the mortahty for males is much lower than in this country. We actuaries do not really know why male and female mortalities differ, nor do we know why the Scandinavian male mortality is much more favourable than in this country, although some of us have thought in terms of environmental factors. It does seem to me that there may be some clue in the figures given here today. If, in fact, there is delay in the onset of some of these degenerative diseases in females as compared with males, maybe the answer is essentially constitutional. Jalavisto: In Finland the excess of male over female mortality is about the highest in the world, especially in the 50-year-old group. This seems to be because coronary death is so common in Finland. I have the impression from Finnish data that at that age or in that sex which has a greater disposition for a certain disease, it is very difficult to lower the mortality in that disease by improving external conditions. For example, in cholecystic diseases in which females have a higher mortality than males, the reduction which has taken place in recent years is greater in the male — although in most diseases the reduction is much greater in females. Perks: When I mentioned the Scandinavian countries I was not thinking so much of Finland as of the Netherlands, Denmark, 82 Discussion Norway and Sweden — countries which have an exceptionally low mortality amongst the males. While your rats were in captivity and being observed, Dr. Berg, were they allowed their normal reproductive functions? Berg: No. The experimental animals were kept separate from the breeders and w^re never mated. Verzdr: The rat colony kept in our laboratory has a 50 per cent survival rate at 23-5 months, even under very varying conditions. (This is, of course, not counting early deaths, since each mother is allowed to feed only four or five young ones.) Spontaneous death appears more frequently after the tenth month of age. The ages of 0 2 4 6 8 lb 12 14 16 18 20 22 24 26 28 30 32 34 36 Age in months (876 cf) Fig. 1 (Verzar). death of 1,602 rats (876 ^ and 726 ?) are shown in Figs. 1 and 2, which also show the number of tumours present in these animals. Tumours first appeared in 8-month-old animals, in both males and females. The number is much higher in females than in males but the occurrence is fairly constant during life. It decreases only between the 29th and 37th months, because so few animals survive. This means that the relative quantity of tumours in all deaths is smallest when the death rate is highest, i.e. between the 18th and 24th months. The main cause of death was lung disease (bronchiectasis), in which the lung becomes infected and has large amounts of white pus. This is found in 24 per cent of females and 35 • 5 per cent of males. There are more tumours in females than in males (21 per cent against 11 per cent) and this difference thus counterbalances the greater Discussion 83 number of deaths due to lung disease in the male. Unexplained deaths were due to vertigo as signs of destruction of one labyrinth were found (9-5 per cent in females, 8-5 per cent in males). Other causes of death which were diagnosed account for about 16 per cent. We have no obvious cases of nephrosis leading to death, but all rats over the age of 10 months have proteinuria and this becomes very high in some very old animals. Very old animals, of about 36 months, generally show no pathological findings which could explain their death. The possibility exists, therefore, that natural death may 20 n 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Age in months (7269) Fig. 2 (Verzar). occur without obvious illness being suffered. (About 15 per cent of our rats were not autopsied.) The survival curves show that after the 10th month "ageing" occurs at the same rate as resistance to external damage (infections) diminishes. This leads to maximal mortality at about 23-5 months, which is also the time of survival of 50 per cent of all individuals. Another point is that creatinuria in our rats begins at 500 days, at exactly the same time as the muscular dystrophy in Dr. Berg's rats. Creatinuria is also present with old age in man. Rockstein : How much, if at all, did you extend the life of those rats by restricted diets. Dr. Berg ? Berg: Lifespan studies have not been completed. At 800 days the survival rate of ad libitum-ied males was 48 per cent as compared 84 Discussion with 87 per cent for restricted rats. Incomplete data indicate that the onset of lesions was delayed nine to twelve months. Rockstein: What was the maximum weight attained ? Berg: The body weight of restricted rats was 25 per cent lower than maximum weight of ad libitum-fed animals. Rockstein: Is the protein content restricted to the same extent as the caloric value of this diet ? Berg: The protein content of the diet was 20 per cent and was the same for the restricted diet. BourUere: Have you measured the basal metabolic rate in both restricted and unrestricted animals of the same age ? Berg: No, I have not. Comfort: In view of the interesting similarities between the life- span curves in man and in rats, it pays to remember, when consider- ing dietary restrictions, that there are differences in growth patterns between them. Your rats show virtually determinate growth, but under some conditions the rat grows in weight and in bone length for most of its life. This is a very different situation from that in man. Berg: After 170 days, skeletal growth practically ceases in the rat though some of the epiphyses remain open for the entire lifespan. There is no evidence of osteogenesis in the cartilage plate of the tibial epiphysis of old rats. Increments in body weight of ageing rats are due largely to fat accumulation associated with prolonged inactivity. Comfort : But my point was that in man you could not, I imagine, restrict growth. The effect of underfeeding on the growth pattern and on the appearance of sexual maturity in man may be different from the effects you can produce in rats. McCay kept his rats infantile for over 1,000 days; I doubt whether a comparable effect could be produced in man. Berg: I think that if we had a comparable inbred strain of glutton- ous men, and could perform a similar experiment, we might obtain results corresponding to those in the rat. Tanner: I am accustomed to dealing with growth data rather than with data dealing with the other end of the lifespan. But methodo- logically there are very great similarities. We must consider, for example, the implications of the use of chronological age in all these data. One possible interpretation of your data could be as follows. We think of children or animals growing in the same way as we think of trains moving along a railway line. You can think of the various diseases as trapdoors on the railway line. You can either consider that those trapdoors have been moved nearer the start so that the train gets to them earlier, or you can consider that the Discussion 85 train has been slowed down and therefore does not reach the trap- doors, which have not altered their positions, until later. Your graph of the numbers dying, such as these that died before 800 days, somewhat supports the latter interpretation. You have got straight lines against age, but the slope of the line for the restricted rats is less than it is for the ad libitum-ied rats. This is the equivalent of the velocity of the train — to follow my analogy — being less. If you altered the time scale you could superimpose those two lines. This is a possible explanation, though not necessarily a correct one. The specific question following that is how much was puberty, the opening of the vagina, actually delayed in your rats ? Berg: There is a delay of about three to four weeks. Tanner: That is relatively small. We do not really know how an animal measures the passage of time — except that it is not by a calendar ! We see this particularly in children, and we have various measures of what we call developmental age : the stage of ossification of the bones of the wrist, the number of teeth erupted, the menarche (the first menstrual period in girls) and so on. These stages are similar to the developmental horizons of the embryologists. It seems to me that in gerontology we are all badly in need of some equivalent of this. Berg: We all recognize that various functions (such as sexual development, bone growth, and onset of lesions) take place more rapidly in lower animals than in man. This difference between species in rate of growth, development and ageing is a problem in itself. What we have attempted to do in our paper is to show the similarity between the species in that they follow the same laws in regard to onset of lesions and death. Tanner: The growth curves of man and rodents (and your rodents, of course, are living in dreadful conditions) are really considerably different. But the growth curves of primates as a whole are very similar. This points out the tremendous importance of doing this sort of work with the cheapest monkeys you can get hold of. Berg: The rats in our colony really live under very fine conditions. Except for being in captivity, they live under better conditions than most humans. However, the cost of these conditions is very high. A similar colony of monkeys would cost millions of dollars. Comfort: Monkeys survive extremely poorly in captivity. Another trouble is the time factor. A baboon can live for over thirty years (Duetz, G. H. (1938). Lab. Rep. zool. Soc., Philadelphia, 66, 31). Verzdr: You have just underlined one of the main points of gerontological research. Dr. Tanner — that we need tests of biological age. We speak far too little about tests, and all our work should 86 Discussion depend on them. You can all judge the age of a man, but how do you do it ? One form of tests in rats is adaptation methods : with ageing the power of adaptation, such as to cold, or lack of oxygen, decreases. You can also measure the age of the collagen in the rat's tail tendon (see Verzar, F. (1957). Ciba Found. Coll. Ageing, 3, 60. London: Churchill). But then it turns out that everything ages differently, and rats age differently in their brains than in their tendons. We irradiated rats with 700 r. and they died quickly, but their collagen had not aged. Holt: I was going to raise the same point, because my experience is also in the comparison of growth curves which are completed at relatively different rates in different species. The dispersions of the age of incidence curves which you showed. Dr. Berg, in the compari- son between man and rat, seemed proportional to their means. You thought that the curve for man had a higher dispersion because you were dealing in that case with a heterogeneous group ; my interpreta- tion was that both distributions were equally dispersed, because I mentally converted them to equivalent relative time scales. Rotblat: Have you ever drawn graphs on which you plot age not in years but in the fraction of the span of life, so that you can compare the spans of life directly ? Otherwise how do you know that the onset of disease is the same in the rat as in man ? Berg: Prof. Simms and I have discussed this extensively. He feels that such a plot would be meaningless. The lifespan is determined by the age of onset of lesions. Hence, to use lifespan as a standard for comparing age of onset would have no significance. It would be like comparing the speed of two racehorses — not in terms of minutes per mile — but in terms of minutes per mile multiplied by miles per minute. Rotblat: From the change in the slope of the Gompertz curve which you showed us I would expect a large extension of the time scale. Tanner: This works out as the equivalent in the human of around 17 to 18 years. In other words if you multiply the scale 30 : 1, which is roughly right, the curves you showed for the rat would be almost superimposable on those for man. Rotblat: This is what I wanted to know: whether they are really the same if they are superimposed. Maynard Smith: To me the most surprising thing you told us, Dr. Berg, was that the ages of onset for a whole variety of at first sight causally unrelated lesions were all shifted in the same direction by the same environmental treatment, i.e. restriction of the diet. I do not believe that one can tell very much about the causes of ageing in any Discussion 87 organism by just looking at curves. But if you experimentally interfere with the conditions and find that those curves move, then you do know something, and that seems to me very exciting. People working on mice have mentioned cases where specific diseases have been shifted to a younger or older age by specific environmental causes. What they have not discussed is whether either dietary restrictions or irradiation, or any other environmental treatment, have a common effect on a number of apparently unrelated diseases. If you delay one disease in the mouse by restricting diet, do you expect to delay the others or not ? This is of enormous theoretical importance and it may one day be of great practical importance. Rothlat: It has been dealt with to a certain extent by Curtis, who has tried six different environmental effects (2nd International Con- ference on the Peaceful Uses of Atomic Energy, September 19.58). Sacher: Curtis reported only on the after-expectations, and not on the kinds of pathology present. Maynard Smith: I want to know whether the ages of the onset of tumours, of kidney diseases, etc., are shifted in the same direction by the same environmental causes. Berg: Yes. The delay in onset of all lesions including tumours produced by dietary restriction points to a single factor that controls the time of onset of disease. Before the discovery of the tubercle bacillus the various forms of tuberculosis involving different organs were considered to be different diseases. With the discovery of the tubercle bacillus these conditions were found to be various consequences of a single cause. Although this analogy is not exact it is possible that a single mechan- ism may be involved in the onset of many widely different diseases. Wigglesworth : An even better analogy is that of malaria. If you reduce malaria in a region the mortality from many other diseases is reduced. Have you had the opportunity yet to switch over the diet at some stage of life in your experiments ? In other words, is it indulgence in youth or indulgence in age which is significant in these effects ? Berg: We are planning such experiments. Comfort: I can think of two factors which could produce exactly such a non-specific effect on many diseases. One is the so-called stress response. I do not know whether you measured the adrenal weight in these creatures. The other is immunological ; I am thinking of auto-immunization processes taking place in the body, and depending on the escape of cell antigens with the passage of time. Either of those could quite readily produce marked changes in many apparently unrelated diseases. 11 88 Discussion Berg: We do have adrenal weights but we have not studied them in relation to time of onset or incidence of lesions. Comfort: Do these rats appear to produce more cortisone than the ordinary animal of that size ? Berg: I do not look upon these restricted animals as being under stress. Comfort: Even if you spend your life in a deck chair, dietary retardation is still a physiological stress. These animals have less to eat than they would normally have — although wild rats do not get all they want to eat. Berg: Within certain limits a state of hunger in the restricted animal appears to be nearer normal for the rat than the satiety of the ad libitum-fed rat. Comfort : It is still possibly a stress. The domestic rat is the result of selection for equanimity and low adrenal weight. Wild-caught rats are quite unmanageable. We have got laboratory animals which, whether we like it or not, have been adapted by covert selection to living under conditions of captivity. Berg: We have in progress stress experiments based on variations in light, noise, and other unfavourable conditions. The adrenal weights of these rats will be compared with the adrenals of rats under standard conditions. Tanner: Maynard Smith asked about the possible common mechanism whereby the times of incidence of these various diseases could all be brought forward together. I think that is rather a dif- ferent situation from the amyloid disease one. As Comfort said, this somewhat nebulous concept of stress does provide a basis. I have recently been to the Mental Health Research Fund conference on "Stress in relation to mental health and disorder" at Oxford (1959. BlackweU's Scientific Publications, in press). Prof. Hans Selye was there and talked about stress as almost equivalent to ageing. The two concepts were being pushed very close together. Selye dis- cussed some very interesting data on the effect of myocardial degeneration of various balances of deoxycorticosterone-type hormones and cortisol-type hormones. He evidently regarded the ordinary circulation of the blood as constituting a stress, while we would regard it as something which perhaps produces ageing. It is particularly in such endocrinological regulations that the general mechanism Maynard Smith is querying probably lies. Sacher: Ionizing radiations, which are normally deleterious and shorten life, can in some circumstances increase life expectation, although without increasing the maximum lifespan. When this occurs in mice and rats, it is observed that the infectious diseases Discussion 89 that are enzootic in the particular populations (such as pneumonia in the rat, various kinds of enteric infections in the mouse) have a lower incidence in the irradiated populations than in the controls. What one sees then is a much more nearly rectangular life-table (less mortality in early and middle life) with no actual increase in the maximum span. L. D. Carlson, W. J. Scheyer and B. H. Jackson (1957. Radiat. Res, 7, 190) at the University of Washington, Seattle, found evidence of this sort, as did E. Lorenz and co-workers (1954. In Biological Effects of External X and Gamma Radiation, ed. Zirkle, R. E, p. 24. New York: McGraw-Hill). They irradiated rats and mice, respectively, with small daily doses of gamma rays. We have obtained similar results at Argonne Laboratory (unpublished). LIFESPANS OF MAMMALIAN AND BIRD POPULATIONS IN NATURE F. BOURLIERE Centre de Gerontologie Claude Bernard and Ldboratoire de Physiologie, Faculte de Medecine, Paris The various marking and banding techniques devised by mammalogists and ornithologists during the last 25 years have provided us with a good many data on the maximum lifespan of numerous species of mammals and birds belonging to a large number of families of these two classes of verte- brates. There are still too many gaps, especially for some groups peculiar to certain geographical areas (such as tropical species in general, small Australian marsupials), or for families with specialized ecology (as cetaceans among mammals and humming-birds or swifts among birds). Nevertheless we already have a preliminary idea of the potential lifespan of most families living in temperate countries. When we turn from individuals to populations, the situa- tion is far less satisfactory. Very few species of mammals and birds can indeed be aged accurately or have been marked in sufficient numbers and followed long enough to provide us with data which can be used to construct adequate life-tables of natural populations. There are none the less a few figures available and the purpose of this review is to bring together this scattered information, bearing in mind that most of the data at hand have been gathered haphazardly during eco- logical studies made for other purposes and that none of them are quite satisfactory from our present point of view. The more accurate observations we have for mammals con- cern some of the larger ungulates which are of interest in game management in Western Europe and North America. All 90 Mammals and Birds : Lifespans of Wild Populations 91 these ungulates breed once a year, have a herbivorous diet and a long maximum lifespan. For such species sources of eco- logical data for the construction of life-tables are of three kinds: (1) knowledge of age at death for an adequate and reasonably random sample of the population; (2) knowledge 1000 500 100' MALES OALL SHEEP Fig. 1. Survival curves for males of five ungulate populations. of the fate of individuals of a single cohort, at frequent intervals; and (3) knowledge of the age structure among the living. As Deevey (1947) pointed out, the first and third types of information can be used only if one is prepared to assume that the population is stable in time. The survival curves of males and females of four species (roe deer, red deer, black-tailed deer and dall sheep), based on 92 F. BOURLIERE the ly. columns in their Hfe-tables, are shown in Figs. 1 and 2. Survival curves of both sexes taken together are given in Fig. 3 for barren-ground caribou. Original figures on which these calculations were based have been published by Evans (1891) for Cervus elaphus of the island of Jura off Scotland, by FEMALES black-tailed oeer (chaparral) OALL SMEEP I Fig 2, Survival curves for females of five un- gulate populations. Murie (1944) for Ovis dalli of Mount McKinley National Park in Alaska, by Andersen (1953) for Danish Capreolus capreolus of the Game Research farm of Kalo and by Taber and Das- mann (1957) for Odocoileus hemionus of California. Data for Canadian Rangifer arcticus are from Banfield (1955). All these populations, except that of dall sheep, were hunted, either by sportsmen or natives. Mammals and Birds : Lifespans of Wild Populations 93 When we compare these survival curves, we can quite clearly distinguish some interesting similarities and differences. (1) A very steep initial slope, indicating a very high mortal- ity during the first year of hfe, is found in both sexes of all the species. The actual mortality rate for calves is probably even greater, since their skulls ar^ more easily overlooked and more Fig. 3. Survival curve graduated on a logarithmic scale, for a series of 292 barren-ground caribou. quickly destroyed than those of the adults. This age class is therefore quite probably under-represented in most samples. (2) During the second year of life there is a small loss in both sexes of all species, with but two exceptions, that of the male roe deer where there is emigration and that of the male black- tailed deer of the Californian chaparral where some yearlings are killed. (3) During the third year of life there is a heavy loss among male and female roe deer and male black-tailed deer from both 94 F. BOURLIERE range types. The roe deer loss is due to emigration (Andersen, 1953) and that of the black-tailed deer to hunting (Taber and Dasmann, 1957). Among the male dall sheep and the male red deer there is little loss during the third year. (4) From the fourth year onward, to old age, the hunted populations (roe deer, black-tailed deer and red deer) show fairly steep losses in both sexes. The rate of loss tends to lessen in full adulthood in the male black-tailed deer, pre- sumably because learning and behaviour make these individ- uals less vulnerable to hunting (Taber and Dasmann, 1957). In the red deer, on the contrary, the rate of loss becomes heavier in full adulthood because of the selection of prime stags by sportsmen. (5) The dall sheep, which is not hunted, shows very little loss from adulthood to 9 years old. If it were not for hunting, the other ungulate populations would probably exhibit survival curves rather more similar to those of Ovis dalli. As Taber and Dasmann (1957) pointed out, it is nevertheless doubtful that they could ever attain as high a survival as long as their ranges were fully stocked and starvation was a com- mon cause of death. In that connexion, it is interesting to note that the survival curve of the barren -ground caribou, which is hunted mainly by natives, is closer to that of the dall sheep than to that of hunted deer. (6) In old age there tends to be in most cases a steepening of the survival curve ; this accelerated loss may be due directly or indirectly to senescence. In dall sheep we know, for instance, that both the very young and very old animals were preferably killed by wolves. Heavily hunted species do not display such a pattern because the high kill permits few individuals to grow old. On the whole, it seems that most of the differences between these various populations of wild ungulates are not inherent in the species, but rather imposed by environmental condi- tions. If a roe deer population is fenced, emigration is pre- vented but winter mortality due to starvation becomes high. Mammals and Birds : Lifespans of Wild Populations 95 In such a case, it would be expected that the population dynamics would be different from those found by Andersen (1953) in Kalo; the older animals would be competing with younger, physiologically more efficient animals. Similarly, if a dall sheep herd were not culled by predators, the population would presumably be limited by food supplies, and the mortality among prime adults would increase accordingly. Taber and Dasmann are thus certainly right in pointing out the danger of considering the population dynamics of a given animal under given circumstances as typical of that species in general. All the species of long-lived and slow-breeding ungulates considered above, belonging either to hunted or unhunted populations, show in most cases age-specific mortality rates. The situation seems very different in small mammals, which are both short-lived and fast-breeding. All the species investi- gated so far appear to have age-constant mortality rates (after very early life). In his study of the survival of wild brown rats on a Mary- land farm, Davis (1948) shows, for instance, that no more than about 5 per cent of the rats live for a year. In the tropical environment of Malayan jungles, the situation looks very much the same for the 12 species or subspecies of Murids studied by Harrison (1956). Table I indicates the estimate of mean survival rates per month for marked animals, together with the mean and maximum length of life (in months) and the age at which only 5 per cent of the population can be expected to survive (effective maximum, 95 per cent). The white-footed mice (Peromyscus leucopus) of the George Reserve in south-eastern ^lichigan have likewise a very low survival rate (Snyder, 1956). With the high rate of mortality of the winter 1950-1951, only one mouse in a thousand could be expected to reach an age of 93 weeks; with the lower rate of the previous winter, 34 mice could be expected to reach 93 weeks, and at least one would probably reach 197 weeks. The mean length of life from birth for such individuals would 96 F. BOURLIERE be 17-4 and 31-5 weeks respectively. Such figures contrast sharply with a potential longevity of six to eight years, recorded in captivity. The same situation occurs in the Tulare kangaroo rat. Fitch (1948) reported that 35-2 per cent of the Dipodomys heermani taken during a four-year live-trapping programme Table I Summary of survival figures for 12 species and SUBSPECIES OF RATS IN MaLAYA (AFTER HARRISON, 1956) Species Survival rate per month Length of life in months Mean Maximum recorded Effective maximum 95% Chiropodomys gliroides Rattus rattus diardii $ R. rattus diardii ^ R. rattus jalorensis, sheltered R. rattus jalorensis, scrub R. rattus argentiventer R. rattus jarak R. exulans R. miilleri R. bowersi R. whiteheadi R. rajah R. sabanus R. canus 0-88 0-75 0-72 0-88 0-76 0-85 0-90 0-73 0-84 0-85 0-75 0-86 0-78 0-82 7-8 3-5 30 7-8 3-6 6-2 90 3-2 5-6 6-2 3-5 6-5 41 50 26 14 4 9 10 + 9 + 10 + 23 10-5 9 22 11 18 28 10 18 18 10 20 12 15 had records that extended over not more than one month, 34 • 2 per cent had records of one to six months on the study area, and only 4-7 per cent had records extending for more than a year. The longest record was 33 months for an in- dividual marked as a partly-grown juvenile. Other instances of short expectation of life at birth in wild rodents are reported by Blair (1953) and Bourhere (1954). Larger rodents have a shghtly longer mean longevity in natural conditions. Kalabouchov (1933) found that 31-5 per Mammals and Birds : Lifespans of Wild Populations 97 cent of a population of little souslik {Citellus pygmaeus) reached the age of one year in the Caucasus and Fitch (1947) reported 9 to 38 per cent of cottontail rabbits reaching one year in central California. Shrews appear even more short-lived than rodents. In England, Crowcroft (1956) estimated that the common shrew {Sorex araneus) lives for only 18 months at the longest and that most individuals die before reaching one year of age. Bats, on the contrary, seem to have a higher survival rate than other mammals of similar size. The expectation of life in the colony of Myotis mystacinus ringed by Sluiter, Van Heerdt and Bezem (1956) in Holland, was 4*4 years, with a maximum lifespan exceeding 20 years! Such longevity is quite unexpected in such small mammals and probably has some- thing to do with the very peculiar metabolic pattern of these animals. To sum up the available data on mammals, it seems certain that, in natural conditions, small and fast-breeding species, with a low ratio of mean to potential duration of life, show very high and age-constant mortality rates. The only exception, that of bats, is probably due to the rather peculiar physiology of these animals. On the other hand, large and slow-breeding species, with a high ratio of mean to potential duration of life, tend to have age-specific mortality rates; this pattern is nevertheless strongly influenced by ecological conditions. In birds, we find the same difference as in mammals between small and fast-breeding species on the one hand and relatively large and slow-breeding ones on the other. Most of the available data concerning this group have already been sum- marized by Lack (1954), Farner (1955) and Hickey (1955) and need not be mentioned again here. We will therefore limit ourselves to a few examples. The best life-table we have for any bird of relatively large size is that of the common terns {Sterna hirundo) studied by AGEING — ^V — 4 98 F. BOURLIERE 2^9S. , ■ I I I . , , , I I I I » » < 1 T T I I f Years of age Fig. 4. Survival curve of a cohort of common terns banded as chicks in 1934 in Cape Cod colonies (After Austin and Austin, 1956). Austin and Austin (1956). Adults (6,965) banded as chicks in the Cape Cod colonies were subsequently trapped in the same places by these ornithologists. To overcome the usual bias caused by band loss in long-lived sea birds, the Austins took the precaution of adding new bands to every bird wearing a band it had carried eight or more years. Their results are therefore more reliable than those of other observers. Fig. 4 Mammals and Birds : Lifespans of Wild Populations 99 shows the survival curve of a cohort of these terns which were banded as chicks in 1934. In this sample, the first year mortality was over 94 per cent and the mean annual mortality rate from the fourth to the 18th year averaged 26 per cent. Beyond the 18th year the curve continues downwards, showing a steady increase in a mortality rate that had remained a straight line during the previous 14 years. The continuous rebanding of all the older birds handled during this study reduces the possibility that this sudden increase in the death rate can be explained by band loss ; the Austins therefore consider that it might be due to senility. The composite life-table based on all the returns of common terns banded as chicks and trapped in nests by these observers, 1940 through 1955, shows the same increase in the death rate after the 18th year. This tendency towards an increased mortality rate in old long-lived birds may be counteracted by a progressive im- provement in the survival rate as the birds get older, at least in the species which are heavily hunted by man. Inexperienced immature or young adults appear to be shot much more frequently than older ones. Such an improvement in the survival rate of older individuals is quite apparent in the survival curve of Scandinavian buzzards (Buteo huteo)^ com- mon herons (Ardea cinerea) and tawny owls {Strix aluco) drawn by Olsson (1958) and shown in Fig. 5. In small and short-lived passerines, the situation is very similar to that of the small rodents and shrews, and the population turnover is very rapid. The mortality rate is always very high, especially at the nestling stage and in the first four months of life. Summers-Smith (1959) has found, for instance, a mortality rate of about 87 per cent for the juveniles (one to four months old) and about 40 per cent for the adults in the urban populations of house sparrows {Passer domesticus) he has studied. We have found even higher figures in a tropical population of the red-billed fire finch {Lagonosticta senegala) now under study in the lower Senegal valley. Kluijver (1951) 100 F. BOURLIERE has found an average annual adult mortality of 49 per cent for great tits (Parus major) in Holland. The lowest adult mortality rate for small birds is found among swifts, where it averages only 18 to 20 per cent per 1000 500- 100 - 5 10 AGE IN YEARS Fig. 5. The number {Ix) of surviving buzzards, herons and tawny owls at beginning of age group {x). Abscissa : Age in years. Ordinate : Numbers of surviving individuals, on a logarithmic scale. year. It should be remembered that swifts, like bats, have a very poor temperature regulation. The lowest figures in the whole class of birds are those of two sub-antarctic birds. In the yellow-eyed penguin (Mega- dyptes antipodes) population studied by Richdale (1957), the adult mortality rate was only 12-9 per cent per year; in the Mammals and Birds : Lifespans of Wild Populations 101 royal albatross {Diomedea epomophora) this figure even goes down to 3 per cent. A summary of investigations on mortality rates in non-passerine birds is given in Table II (based on Farner, 1955 and completed after Bendell, 1955; Hickey, Table II Summary of ustvestigations ox mortality rates IX xox-passerixe birds Approximate range of Approximate range of Order juvenile mortality adult annual mortality rates rates /o o/ /O Sphenisciformes 13-30 Procellariiforines (Diomedeidae) about 3 Pelecaniformes ( Phalacrocoracidae ) 35-80 12-30 Ciconiiformes (Ardeidae) about 60 about 30 Anseriformes (Anatidae) 50-85 17-65 Falconiformes (Accipitridae) about 60 about 30 Galliformes 20-50 50-83 Charadriiformes (Charadrii) 15-50 (Lari) 40-60 18-30 Columbiformes (Columbidae) about 80 55-58 Strigiformes (Tytonidae) 50-79 28-57 (Strigidae) about 50 about 30 Apodiformes (Apodidae) about 30 18-20 1955; Lack, 1956; Richdale, 1957; Boyd, 1957, 1959; Summers- Smith, 1959; and Southern, 1959). The adult mortality rates given there apply after the first 0-5 to 1-5 years, according to the species or groups concerned. The juvenile mortality rates are calculated for a year beginning with the fledgling leaving the colony, or sometime later during the summer of hatching. 102 F. BOURLIERE We therefore find very much the same dichotomy in birds as in mammals, between small and fast-breeding species on the one hand and large and slow-reproducing ones on the other. The highest ratio of mean to potential lifespan is indeed found in a bird, the royal albatross, which reproduces only every second year and may reach an age of at least 25 years in the wild. Reproductive and mortality rates are thus closely adjusted. Both vary mainly with ecological conditions and an increase in mortality rate in older individuals, which may be due to the onset of old age, is apparent only in large, slow-reproducing and long-lived birds and mammals. REFERENCES Andersen, J. (1953). Dan. Rev. Game Biol., 2, 127. Austin, O. L., and Austin, O. L., Jr. (1956). Bird Banding, 27, 55. Banfield, a. W. F. (1955). Canad. J. Zool., 33, 143. Bendell, J. F. (1955). Canad. J. Zool., 33, 195. Blair, W. F. (1953). Advanc. Genet., 5, 1. BouRLiERE, F. (1954). The Natural History of Mammals. New York: Knopf. Boyd, H. (1957). Bird Study, 4, 80. Boyd, H. (1959). Ibis, 101, in press. Crowcroft, p. (1956). Proc. zool. Soc. Lond., 127, 285. Davis, D. E. (1948). Ecology, 29, 437. Deevey, E. S. (1947). Quart. Rev. Biol, 22, 283. Evans, H. (1891). Some Account of Jura Deer. Derby, privately printed. Earner, D. S. (1955). In Recent Studies in Avian Biology, p. 397. Urbana : University of Illinois Press. Fitch, H. S. (1947). Calif. Fish Game, 33, 159. Fitch, H. S. (1948). J. Mammal, 29, 5. Harrison, J. L. (1956). Bull. Raffles Mus., 27, 5. Hickey, J. J. (1955). In Recent Studies in Avian Biology, p. 326. Urbana: University of Illinois Press. Kalabouchov, N. I. (1933). Rec. Trav. Sci. Univ. Moscou, 1, 29. Kluijver, H. N. (1951). Ardea, 39, 1. Lack, D. (1954). The Natural Regulation of Animal Numbers. Oxford: Clarendon Press. Lack, D, (1956). Swifts in a Tower. London: Methuen. MuRiE, A. (1944). The Wolves of Mount McKinley. Washington: U.S. Dept. Int. Nat. Parks Service. Mammals and Birds: Lifespans of Wild Populations 103 Olsson, V. (1958). Acta Vertebratica, 1, 86. RiCHDALE, L. E. (1957). A Population Study of Penguins. Oxford: Clarendon Press. Sluiter, J. W., Van Heerdt, P. F., and Bezem, J. J. (1956). Arch. neerl. ZooL, 12, 63. Snyder, D. P. (1956). Misc. PubL, Mus. ZooL Univ. Mich., 95, 1. Southern, H. N. (1959). Ibis, 101, in press. Summers-Smith, D. (1959). Ibis, 101, in press. Taber, R. D., and Dasmann, R. F. (1957). Ecology, 38, 233. DISCUSSION Rotblat : Is anything known about the lifespans of the same species in captivity ? Bourliere: Not for the swift, because it is at present impossible to keep them caged. What we have are good figures on the maximum lifespan of some individuals. We also know that in both large and small mammals and birds the maximum lifespan in captivity is always far greater than in the wild. Nevertheless, in wild popula- tions, at least in those species for which we have data, a very small percentage of very old individuals is found ; but in order to find these very scarce old animals, you need to study a very large population for a very long time. Rotblat: I understand that bats kept in zoos live much longer than the 20 years which you mentioned for the albatrosses. Bourliere: Twenty-five years is the longest recorded lifespan for the royal albatross in the wild in New Zealand, but as far as I know, nobody has ever kept an albatross in captivity for more than a few years because it is very difficult to feed them. Comfort: A chaffinch has been kept for 29 years (Moltoni, E. (1947). Riv. ital. Orn., 17, 139), and even an inbred budgerigar is reported to have reached nearly 20. Scheidegger: One swift in a big colony near Basle lived for 18 years, but the rest had an average age of about 5 to 6 years. Danielli: Your remarks about birds which survive longest being large does not fit in with data for the swift, as you pointed out. Prof. Bourliere. Do you attribute the advantage the swift seems to possess to the fact that it more or less hibernates ? Bourliere: That explanation was advanced by Farner (1955) and it may be true because we have the very same phenomenon in mammals. We may compare rodents and bats of similar size and weight. Small mice, for instance, never live in captivity or in the wild for more than four years, whereas bats of the same weight will live for 20 years ; so there is certainly some correlation between a long 104 Discussion maximum lifespan and the ability to lower the body temperature and the basal metabolic rate for more or less prolonged periods. Danielli: Have experiments been done in which groups of animals which normally hibernate have been prevented from doing so ? Bourliere: At the present time I know of no good observations * which have been made on mammals. The difficulty is not only to house a large number of animals during 30 or 40 years, but also to secure an investigator who could study such a problem for three decades. Danielli: Even if one did, of course, you would still be up against some difficulty in interpreting the facts, because it seems to me that the advantage which is gained may either be that the "biological time-scale" is altered by hibernating, or alternatively the hibernat- ing animal may be protected from all accidents and so forth. One does not know which of these two alternatives is involved. Comfort: Many small birds presumably die in winter. If a species hibernates it has not got to search for food, and so it is not so liable to die from lack of it. Bourliere: In swifts there is no true hibernation during winter, because they migrate to tropical Africa at that time of the year, but their temperature control is nevertheless very different from that of passerines. The studies on swifts which have been made in Oxford and Switzerland by D. Lack and E. Weitnauer have shown (see Lack, 1956) that during bad weather, especially in early spring when the parent swifts are unable to obtain enough flying insects to feed their young, the young then undergo a kind of pseudo-hibernation or torpid stage instead of dying as other species do. J. Huxley, C. S. Webb and A. T. Best (1939. Nature (Lond.), 143, 683) have des- cribed the same feature in adult humming-birds at night, and they are also long-lived animals. One humming-bird lived in captivity in the Copenhagen Zoo for more than eight years, which is quite a record for a bird so difficult to keep in captivity. Danielli: Humming-birds might be the right material, as they have this diurnal "hibernation". Maynard Smith: In the terns there was a very low mortality in the second and third years of life. Is this associated with the fact that this species does not breed until its fourth year ? Bourliere: Yes, common terns do not return to breed in quantity until their fourth summer. Sacher: Have the life-tables of animal populations in wild-life preserves, such as the European and American bison, been studied ? Bourliere : As far as I know, no such studies have ever been made on bison in America or in Europe. The first good study on the Discussion 105 behaviour and ecology of the American bison was pubhshed a year ago (McHugh, T. (1958). Zoologica, N.Y., 43, 1), and we are still waiting for a similar work on the European form. The so-called wild European bison are in fact so domesticated that they are not a good example to study. If you want to have samples large enough to be studied, you need to choose rather common species which can live in national parks or some place where human distui'bance is very rare. That is why most of the available information has been drawn from the field of game management or rodent control. Chitty: Did you say there was no age-specific mortality rate for small rodents in the wild ? Bourliere: I do not know of any study showing such an age- specific mortality rate in wild rodents. Chitty: I do not really see how this kind of information could be obtained very easily for wild populations. Such evidence as I have published (1952. Phil. Trans. B., 236, 505) shows that there is a higher mortality rate with increasing age but of course it is exceed- ingly difficult to separate the environmental components from it. There is an increase in mortality rate as the winter goes along, but it is not known whether that is because of changing ecological condi- tions, or because of an increase in average age. Kershaw: This may be compared with observations on insects in the wild. There seems to be evidence now that the survival of insects with a rapid population turnover is modified by predators and natural hazards, whereas those with a long and slow population turnover maintain their own intrinsic survival. In mosquitoes it seems likely that the intrinsic survival, having a Gompertz function with a sloping straight line, is altered completely by field conditions, and has a flat Gompertz function. For the last ten years we have been studying the life-cycle of one of the West African flies, Chrysops, which turns over once a year. In the laboratory that fly has a normal rectangular survival, both in those bred from the pupae and in wild-caught flies. We have been following through natural populations of flies coming in to bite man, which of course is a selective, but functionally selective, population. We have found that throughout the year the population is made up of separated succes- sive cohorts, each behaving with its own particular rectangular survival, so that in this particular fly the intrinsic survival is the natural one. I think one has to go back to mosquitoes now that one has biological markers for them, based on parity and so forth, to see whether what is true of Chrysops is also true of mosquitoes. This, of course, is of importance in producing mathematical models. ARTERIOSCLEROSIS IN BIRDS S. SCHEIDEGGER Institute of Pathology, University of Basle Arteriosclerosis ranks first among diseases affecting the arteries, and it is of much more importance in human beings than in animals. All experiments designed to produce such a form of degeneration in the vessels of animals are doubtful, since a severe form of this disease can only be found in man. In animals we never find such severe forms as regards the spreading of the disease or transformation of the tissue, and it_ can never be described as a disease producing clinical symp- toms. In other words, whereas arteriosclerosis is one of the principal diseases in man and a frequent cause of death, in animals it is only of secondary significance, showing only slight malformations, possibly with transformation of the^ vessel wall. The term arteriosclerosis, which etymologically means a hardening of the vessels, in fact stands for a combina- tion of different lesions. At the onset of the disease, we only find some inclusions in the wall. In the final stage there is accumulation of degenerations, malformations, and trans- formations ; all layers of the vessel walls are now affected and we can see calcifications, ossifications, and often an occlusion of the lumen. In other cases the same calcification and ossi- fication process can produce an enlargement of the vessel. At the beginning of the illness we can see a series of various re- actions of unknown and doubtful origin. Some fox^ms of arteriosclerosis constitute pure inflammations, whereas some can be classified as pure degenerations. In some cases this disease of the vessels is restricted to a single organ or to a specific system, i.e. it may be found only in the vessels of the heart, or the vessels of the brain, or perhaps the finer vessels 106 Arteriosclerosis in Birds 107 of the extremities. It is a known fact that the vessel can be affected in various ways. The disease can be a diffuse one, and then the whole vessel is in an arteriosclerotic state. Alternatively, the severe degeneration of the vessel — often the arteria coronaria — is restricted to small defined areas. With this form, we speak of a so-called arteriosclerosis in plaques. Furthermore, we know — and this is another important factor — that the organ has an influence on the development and type of arteriosclerosis. In the brain, for example, ^ve never find the same forms of vessel degenerations as in the kidneys or in the heart. In the brain vessels, especially in the finer vessels of the brain substance, we never find atherosclerosis with deposits of fat and cholesterol esters, etc. In this organ we have a more diffuse hyaline transformation, in others fine fibrillar changes, a so-called fibrillar vessel degeneration, similar to that which can be seen in the neuroglial tissue (of the neurofibrillar changes and in the senile plaques of the Alzheimer disease). This form, which Scholz (1938; Scholz and Nieto, 1938) was the first to describe, is a typical mal- formation of the brain vessel. The final stage is always the same : the occlusion of the vessel. This occlusion is the only pathological symptom which can be found. The pathological process and the way of evolution can, however, vary. In most observations of arteriosclerosis it is not possible to give a key to its origin and development. We know that arteriosclerosis of the vessels of the heart, which develops only in restricted areas, is often the result of an inflammation whereas the diffuse forms mostly result from a primary pure degeneration. Experimental studies of this disease are not possible. It is possible that some types of arteriosclerosis are the cause of either an acute or a chronic arteritis. The acute arteritis can in turn be the result of an infection with bacteria or with toxins. In the place of the inflammation of the vessel wall, parietal thrombi develop. The bacterial infections may disappear. After the inflammation a degenerative form of arteriosclerosis may occur. An injured vessel wall, or 108 S. SCHEIDEGGER infiltration by parasites, can often be the causative factor in chronic arteritis. Domestic animals, especially cattle, dogs and pigs, show a slight thickening of the arterial walls through proliferations of the intimal tissue. Scherer (1944) describes a case of a 20-year-old chimpanzee with sclerosis of the main vessel of the brain without degeneration of the tissue. In four Macacus rhesus monkeys the same author could observe fine degenerations of the cortex of the brain with multiple small areas showing a softening of the tissue. In the cerebellum the granular layer was in a state of degeneration with multiple areas in which the cells had disappeared. In birds more severe forms of such degenerations can be seen. Two different phenomena may be the cause : some species of birds reach an advanced age, and some are (and this is another important fact) purely carnivorous. The distribution and the anatomy of arteriosclerosis in birds present some difficulties. In most cases the aorta is the seat of the malformation. The aorta valves are free of lesions. In the intima layers we often find hyperplasia with an increasing of fibres. The media shows a muscular granularity in this part of the vessel wall. Micro- scopical examination of such vessels shows a media with ruptured muscle fibres and and a split or broken elastica. The intima is often covered with a fibrocellular exudate. One of the best and most complete reports on the problem of arteriosclerosis in birds has been published by Fox (1923), who gives a survey of the different types of diseases affecting the animal in captivity and in the wild state. To illustrate the problem the author has collected some thousands of post- mortem examinations in the Washington Zoo. Psittaci have a high percentage of arterial disease. Some findings are of interest. Often the central vessels are not the principal seat of an atheromatosis as in other classes, and the lesion shows reactions with tissue proliferation. Accipitres have the great- est percentage of any order. The arterial lesions are frequently accompanied by renal, myocardial, and valvular disease. Degenerations are equally severe in the media and intima. Arteriosclerosis in Birds 109 Calcifications are not seldom found in this group of birds. In parrots the arterial damage is caused in the arteria carotis or in the small wing arteries, but the most common seat of the process is the lower thoracic region. Galli often demonstrate a vascular disease combined with myocardial disease. Cormo- rants, pelicans and gannets often have arteriosclerotic mal- formations and intimal proliferations. Ducks and geese present a considerable number of cases which demonstrate arteriosclerotic degenerations accompanied by cardiac and general pathology. The character of the lesions is similar to that found in the Accipitres. Arteriosclerosis is common to many zoological orders. Fox gives a good general review of this problem. The order of percentage incidence is Accipitres 6-6; Anseres 3-4; Psittaci 1-8; GalK 1-6; Passeres 0-22. Carnivorous birds have^the highest incidence of chronic^ arterial disease. Next in order are the~ungulates, the anserine birds and the carnivorous mammals. Fox remarks on the interesting fact that the orders with great activity, such as the primates and the passeres, are at the end of the list. This is perhaps due to the fact that their food consists chiefly of carbohydrates. Arteriosclerosis in mammals and birds in captivity is often accompanied by nephritis, chronic infectious disease or chronic enteritis. In some cases of this disease, aneurysm in the vessels can occur. This develops above the valves and arises from a degenerative arteritis. The present author has examined cases of birds with slight, medium, and severe arteriosclerosis. These observations are all autopsies in the Basle Zoo. The animals were vultures, storks, cranes, flamingoes and geese. Milder forms of this disease are quite common. In these cases the aorta showed a thickening of the intima and plaques could be seen, especially where the main vessels branch off. In these parts we found small deposits of fat or a pure fibrosis without cholesterol deposits or atheromatous ulcerations. (See Figs. 1-10.) 110 S. SCHEIDEGGER In all the cases observed the elastic membranes were split off or interrupted. In the medium forms of atherosclerosis yellow areas could be seen in the vessel wall and with increas- ing age the disease became more severe. Wolkoff (1925) described a typical arteriosclerosis in a 40- year-old parrot. Observations of arteriosclerosis in a 85- and a 42-year-old parrot with severe atheromatosis are reported by Nieberle (1931), Nieberle and Cohrs (1931), by Pallaske (1930), and by Beneke (1931). In his thesis Fahr (1935) dis- cussed arteriosclerosis in chickens. The comprehensive article written by Krause (1939) about the pathology of animals states that the beginning of the aorta is not the place of the primary degeneration. He found that in most cases the principal seat of this disease was the part of the aorta between the kidneys. He reported that in his observa- tions atheromatous ulcerations could never be seen. The intima was always intact, hardened and thickened by a fibrosis. Inflammatory processes are not of importance. In his opinion lipoidosis and sclerosis are two diff'erent processes. He found the primary infiltration of fat in the aorta between the kidneys, never in the aorta ascendens or in the arcus aortae. Such forms of degeneration could be observed in 5 to 6-year-old poultry. A calcification could very seldom be found. He suggests that only in some cases with avitaminosis or hypervitaminosis can a primary necrosis produce small calcareous deposits. Fahr (1935) suggests that no connexion exists between the degree of the infiltration of fat and age. Sclerosis is, however, a process that can only be found in advanced age. He was never able to see a coronary sclerosis. In these studies of arteriosclerosis in birds only one animal developed an arterio- sclerotic disease of sufficient clinical significance to cause severe functional disturbances. A vulture nearly 60 years old developed an arteriosclerotic disease with such clinical symptoms as can be found in man. The cause of death was a diffuse vascular sclerosis. All the different stages could be lit. ' '<^'"'^ Fig. 1. Diffuse pure mucoid degeneration with a fine lipoidosis in the arteria carotis {Sarcorhamphus gryphus L.). K|?jaCv«% « . - -* w. i-*^ • ^-^' ■!^^ Fig. 2. Diffuse thickening of tlie aorta by infiltration of mucoid and fat substances with degeneration of the elastic membranes {Balaeniceps rex Gould). facing page 110 Fig. 3. Arteriosclerosis of the basilar artery. Thickening of the intima with degeneration of the media {Sarcorhamplms gryphus L.). Fig. 4. Coronary vessels with a fine thickening of the intima and deposits of cholesterol esters and fat (Sarcorhamphus gryphus L.), ^i*w«;if^"''^^i;^pft^BB^ • ~ ^ #■ T.ir-^'*'' Fig. 5. Diffuse infiltration of mucoid substances in all the dif- ferent layers of the vessel wall, with degeneration of the media (Phoenicopterus). Fig. 6. xVorta of the red-breasted goose with degeneration of all the different layers of the vessel wall and a splitting off of the elastic membranes in the media {Branta ruficollis Pall.). Fig. 7. Thickening of the media and intima with a fine calci- fication and multiple small deposits of cholesterol esters {Balaeniceps rex Gould). X 0^. tf Fig. 8. Severe thickening of the arterial wall by large deposits of fat and cholesterol esters {Balaeniceps rex Gould). Fig. 9. Vascular hyalinosis with perivascular softening {Sarcorhamphus gryphus L.). Fig. 10. Sections throuoh the brain of a vulture showing post- malacic cyst secondary to cerebral arteriosclerosis, with enlarge- ment of the lateral ventricle {Sarcorhampfms gryphus L.) Arteriosclerosis in Birds 111 seen in this case: proliferations of the intimal cells, with con- centric lamellations which completely filled the lumen, hyaline transformation, swelling and necrosis of the media, com- pression of the lumen, medial calcification and ossification. Atheromatous plaques were found in the main vessels. In the observation under review the concentric lamellations and the necrosis of the media produced a compression of the lumen of the vessel which was accompanied by a softening of the brain with subsequent formation of cysts. The coronary vessels showed a high degree of arteriosclerosis. The lumen was often nearly obstructed by intima proliferations. Several infarctions of the myocardium could be seen. The present author's suggestion is that only in captivity can such severe forms of arteriosclerosis occur. In nature the animal concerned would have died earlier of hunger since the higher functions of life would have been handicapped. The atherosclerosis induced by several conditions in animals may be reversed by removing those conditions but there is no evidence that the same possibility exists as far as the disease in human beings is concerned. Atheromatosis in the human adult must be considered as irreversible. Atherosclerosis is induced in rabbits and chickens by adding cholesterol in large quan- tities to the food. But animals species differ greatly. The concentration of cholesterol in the blood is a very important factor in atherosclerosis both in the human being and in animal species. Atherosclerosis can be seen in any artery but the major concern is the coronary artery. It is common in young persons, and quite common m persons over 30, but this type of degeneration cannot be seen, or extremely rarely, in animals. Arteriosclerosis is of much more importance in human beings than in animals. This fact is important. Unfortunately there is no possibility of studying this disease which so frequently affects the human being, in animal experiments. 112 S. SCHEIDEGGER REFERENCES Beneke, R. (1931). Beitr. path. Anat., 87, 285. Fahr, H. O. (1935). Dissertation: Giessen. Fox, H. (1923). Disease in Captive Wild Mammals and Birds. Phil- adelphia: Lippincott. Krause, C. (1939). Ergebn. allg. Path. path. Anat., 34, 226, 367. NiEBERLE, K. (1931). Verh. dtsch. path. Ges., 26, 239. NiEBERLE, K., and Cohrs, P. (1931). Lehrbuch der spez. path. Anatomie der Haustiere. Jena: Fischer. Pallaske, G. (1930). Frankfurt. Z. Path., 40, 64. Scherer, H. J. (1944). Vergleichende Pathologic des Nervensystem der Saugetiere. Liepzig : Georg Thicme. ScHOLZ, W. (1938). Z. ges. Neurol. Psychiat., 162, 694. ScHOLZ, W., and Nieto, D. (1938). Z. ges. Neurol. Psychiat., 162, 675. WoLKOFF, K. (1925). Virchows Arch. path. Anat., 256, 751. DISCUSSION Verzdr: Was there any difference in the diet of the birds which had arteriosclerosis and those which had not ? Diet is important in view of the present theories that arteriosclerosis is connected with fatty acids. Scheidegger: All these birds are carnivores, and it seems to me that this is one of the most important factors in arteriosclerotic disease. The birds all came from the Basle Zoo, and this form of arterio- sclerotic disease is seen especially in vultures, flamingoes, cranes and so on. I have never seen severe forms of arteriosclerosis in other birds than the kinds mentioned here. BourUere: Can it really be said that arteriosclerosis in birds is always related to a carnivorous diet ? European flamingoes are "mud-eaters", feeding mainly on small crustaceans and inverte- brates, and these kinds of food should contain a lot of unsaturated fatty acids. Geese also feed as much on plants as on invertebrates. In some zoos flamingoes may receive a different diet than they do in the wild. Scheidegger: In Basle Zoo the flamingoes ate crustaceans and so on, and they also received small pieces of meat in this food. I do not know about geese. Hinton: I know there are a lot of differences between arthropods and the rest of the animal kingdom, but small crustaceans are bits of meat, aren't they ? Rockstein: Arteriosclerosis has been observed in the common fowl, of course, and it is primarily a herbivore, so I do not think one can generalize so easily on this point. Discussion 113 Lindop : There are two studies going on in this country on athero- sclerosis in poultry (Saxl, H., in press; Hall, D. A., in press). One is done by the Ovaltine workers who are keeping two groups of animals on identical diets, but one group has limited exercise. These groups are being followed for the development of atherosclerosis. The other study, at Leeds, is more on the cytological and the biochemical side. There they have been able to produce atherosclerosis in poultry which have limited exercise, and they have also been able to reverse atherosclerosis in poultry which have been allowed to exercise after they had been inhibited. Atherosclerosis developing in zoo animals might therefore be caused by the comparatively limited exercise they get. Scheidegger : The problem really concerns the age of these birds. This old vulture lived in the zoo for about 5Q years. It came as a young bird, at the age of 3 or 4 years, and the food was always the same. In its last year the bird had severe arteriosclerosis. The ages of the other birds are not known, and I cannot tell you what the birds had to eat in captivity. Jalavisto : Orma has made observations on cholesterol-fed poultry and it seems that even a little exercise may be enough to retard development of atherosclerosis. He put the food on one side of a fence, and the water on the other side. The poultry drink after they have eaten and therefore have to go over the fence and back again all the time. That could be very easily arranged in a zoo, and it would be one means of checking easily whether it is exercise or something else which is operative in this case. Nigrelli: Arteriosclerosis and atherosclerosis can be induced experimentally by dietary methods in poultry, as Dr. Lindop said, and I think there is some relationship with the pantothenic acid in the diet. Comfort: The blood pressure of these birds in relation to their con- figuration may have some effect. It is just conceivable that birds of the flamingo type, which have both a long neck and a considerable change of posture between their head-erect and head-down positions, have larger fluctuations in blood pressure than other animals. It is a point which needs investigation. The other point is that I am quite sure all zoo birds and animals are better fed, or more consistently fed, than they are in the wild. I rather suspect that, like Dr. Berg's rats, these birds would be less likely to get atherosclerosis if they were only fed intermittently. I suspect that in the wild the vulture does not feed every day, or even every week, unless he is lucky. Bourliere: I understand that a study is already under way in 114 S. SCHEIDEGGER South Africa concerning the blood pressure of the giraffe. It was found to be extremely high, but as far as I know no increase in arteriosclerosis was observed in that species, at least in the wild. Comfort: There is the slight drawback that the giraffe is such an extremely timid animal that even in a zoo it is a nightmare to file its hooves, or anything like that — you have to give it an anaesthetic. It must be rather a difficult animal on which to estimate the blood pressure, unless you could telemeter it. Rockstein: The birds which show this atherosclerotic condition seem to be among the longest-lived birds. For example I have here records of a common crane which lived to more than 42 years, a flamingo to more than 22, a stork to 30, and a vulture to close to 60 years of age. The question is whether atherosclerosis is indeed a factor in longevity as such ? Berg: Prof. Scheidegger's presentation underlines again the necessity of knowing causes of death in lifespan studies. Important, too, is the question of whether degenerative diseases are inevitable with ageing. Kershaw: The Royal Air Force has been doing routine post- mortems on its fittest people — those who fly jet planes. Most of these are aged around 19, 20, and 21. In the coronary arteries of these men a high and astonishing degree of coronary atheroma was found. Whether that means that most of us have vessels which have atheroma in them which is of no significance, or whether that particular selected group is more likely to have atheroma, is not clear. Danielli: It should be perfectly easy, shouldn't it, to get post- mortems on the average young adult killed in an accident ? Kershaw: The difficulty lies in getting comparable objective reports. RELATION OF LIFESPAN TO BRAIN WEIGHT AND BODY WEIGHT IN MAMMALS* George A. Sacher Division of Biological and Medical Research, Argonne Natioiial Laboratory, Lemont, Illinois My interest in the comparative study of lifespans stemmed from the reahzation that the zoological literature contains a wealth of data that can be used to analyse the factors govern- ing ageing and longevity in mammalian species. In this paper I shall first describe a statistical analysis of the relations of life- span to brain and body weight, and then enter into a discus- sion of the theoretical issues. These considerations have been alluded to briefly in previous communications (Sacher, 1957, 1958). The objective of the empirical analysis is to establish the quantitative dependence of the lifespans of mammalian species on the body weights, brain weights, and metabolic rates of adult representatives of these species. This is in effect a study of allometric relationships, in which lifespan is regarded as a physical dimension of a species on the same footing as the linear or mass dimensions. In this paper the species lifespan is defined as the maximum documented longevity for that species. With a few exceptions the lifespan data come from domesticated animals or from zoo animals. There were two reasons for using the lifespan rather than the life expectation. The first is that satisfactory estimates of the life expectations even now are available for only a few domesticated species and for an even smaller number of * This work was performed under the auspices of the U.S. Atomic Energy Commission, 115 116 George A. Sacher captive species (Comfort, 1956, 1958). Second, when data come from animals kept under very different environmental conditions the lifespan is a more stable longevity parameter than is life expectation. This is clearly seen in the life-tables 1 0,000 c — r o o o o" q: ui Q. 1000 100 10 Thoroughbre,d MarftS A. Comfort, J. Ge^otitol 13:342-350, 1958 X 12 16 AGE, 20 years 24 28 32 Fig, la. Logarithm of age-specific rate of mortality plotted versus age (Gompertz plot). Data on thorough- bred mares from Comfort (1958). A life-table for which the Gompertz plot has a straight-line relation to age conforms to the Gompertz equation qx = Ae'^^. of human populations in different countries or in the same country in different historical periods. Instances can be found in which life expectations vary by more than a factor of two, but even in these extreme cases the lifespans do not differ by Relation of Lifespan to Brain and Body Weight 117 as much as 20 per cent. There is a good reason for this, as will be discussed below. One objection that has been raised to the use of lifespans is that the estimated lifespan will increase as the sample on which the estimate is based increases. Until recently there was no comprehensive statistical treatment of this question. With the publication of Gumbel's (1958) treatise on the Statistics of Extremes, we now have a statistical theory that is adequate to deal with most questions that arise. The characteristic oldest age (the age attained when one survivor remains of the initial cohort) is an easily computed statistic in terms of which we can discuss the dependence of lifespan on cohort size. If the survivorship curve is of the Gompertz type, in which the age-specific rate of mortality increases exponenti- ally with age, the characteristic oldest age increases as a double logarithmic function of the cohort size. This is illustrated in Fig. 1, where the characteristic oldest age is shown in relation to cohort size for an actual life-table of the Gompertz type, drawn from Comfort's analysis (1958) of the life-table of thoroughbred mares. It can be seen that increasing the cohort size by a factor of 10^ would increase the characteristic oldest age by only one-third. The lifespan itself, i.e. the age at death of the oldest survivor, would vary more slowly than this. The typical mammalian life-table can be adequately des- cribed by the Gompertz-Makeham equation, in which the relation of rate of mortality to age contains a constant term in addition to the Gompertz term: q^ =Ae^ + B (1) The Makeham term, B, is markedly influenced by environ- mental conditions, whereas the Gompertz term is influenced to only a small degree. However, the Gompertz term will always dominate at advanced ages, and the lifespan therefore tends to behave like an extremum statistic of a Gompertz life- table. Since the inherent ageing parameters characteristic of the species are embodied in the parameters A and a of the 118 George A. Sacher 40 12 35 o lU o < to UJ o o o I- o < cr < X o 30 - 25 - 20 - 15 - 10 - 5 - 1 1 I ■ ■ ! 1 1 ^^..-^"""^"^ - ^ yE — X - / - X - . / / G- Gompertzion, Bq =22yr.. / /e 1 1 E- Exponential, Sq = 2 yr. 1 1 1 1 10 10' I0> 10^ 10= 10^ SIZE OF SAMPLE Fig. 16. Relation of characteristic oldest age to size of initial cohort. The curves drawn are for a Gompertzian life-table (see text) and for an exponential life-table. The curve for the Gompertzian populations is repre- sentative of the amount of variation of lifespan with sample size that might be expected in populations of domestic or zoo animals kept under good conditions. The line for the exponential population is illustrative of the relation of lifespan to sample size that would hold for populations under very heavy environmental pressure, such as small birds or rodents in the wild. Gompertz term* it follows that the lifespan is the preferred statistic for the characterization of species longevity. * It can be argued that the life expectation should be preferred precisely because it measures the response to environmental as well as intrinsic factors. There can be valid reasons for preferring the life expectation in certain contexts, and especially in the discussion of evolutionary or ecological adaptations. However, the present paper is addressed to the question of intrinsic limitations on length of life, and these questions are appropriately discussed in terms of measures that are as nearly as possible invariants for the species, independent of environmental influence. Relation of Lifespan to Brain and Body Weight 119 The theory of extremes can be used to compensate for the bias introduced by very great differences in cohort size. This has not been attempted in the present study. The data on brain weight and body weight were taken from the extensive compilations of these values by Count (1947), BODY WEIGHT -GRAMS 10301^3X10^ lo' 3XI0' 10* 3X10* lo' 3Xlo' 10* 3X10* 2.0 - -. 1.0 LOG|o BODY WEIGHT Fig, 2. Relation of lifespan to body weight for 63 species of mammals. Data plotted on double logarithmic grid. The symbols denote groups of species as follows : open circles — primates and lemurs ; solid circles — rodents and insectivores; crosses — carnivores; solid triangles — ungulates and ele- phants ; star in circular field — man. von Bonin (1937), and Quiring (1950). The body weights and brain weights are in some instances averages of values re- ported by two or more investigators. Lifespans are the maxi- mum records encountered in the sources examined, and were obtained principally from the compilations by Bourliere (1946), Comfort (1956), Flower (1931), Walker (1954), and Yerkes and Yerkes (1929). 120 George A. Sacher There are 63 species for which body weight, brain weight and Hfespan values were tabulated. Most orders of placental mammals are represented, the most important omissions being the bats and marine mammals. Common logarithms of the numbers were used in the statistical analysis. Simple and multiple linear regressions were computed. The statistical 1.0 BRAIN WEIGHT - GRAMS 10 30 100 300 1000 3000 LOG,, 2 0 BRAIN WEIGHT 3.0 Fig. 3. Relation of lifespan to brain weight for the same 63 species shown in Fig. 2. Symbols as defined in legend to Fig. 2. formulae are presented in lucid fashion by Hald (1952). Appendix 1 contains the means and standard errors of the variables, the regression coefficients and their standard errors, the total variances of the variables, and their residual vari- ances after removing the variance in regression. With the information provided the total, partial and multiple correla- tion coefficients can also be computed. Relation of Lifespan to Brain and Body Weight 121 The relation of lifespan to body weight is exhibited in Fig. 2, with the points plotted on a log-log grid. The species are divided into four groups, each comprising one or more orders, and each group has a distinguishing symbol. No use is made of this subdivision in the statistical analysis since the data were not sufficient to justify a separate analysis by orders; such analysis will be undertaken when more extensive data are collected. It can be seen that there is a highly significant relation between lifespan and body weight. The relation of log lifespan in years (x) to log body weight in grams (y) is found by least squares to be (see also Appendix lb) X = 0-198?/ + 0-471 (2) This regression accounts for 60 per cent of the variance of lifespans. It can also be seen that the different groups are stratified in the diagram, with the rodent lifespans lying below the regression line, those for carnivores and ungulates more or less evenly distributed around the line, and those for primates lying almost entirely above. A similar diagram showing the relation of lifespan to brain weight is displayed in Fig. 3. The least squares regression of log lifespan (x) on log brain weight in grams (z) (Appendix Ic) X = 0-3252 + 0-684 (3) accounts for 79 per cent of the lifespan variance. This is significantly greater than the variance reduction brought about by regression on body weight, so it can be concluded that brain weight by itself is a better predictor of lifespan than is body weight. The superiority of brain weight over body weight as a predictor is manifested by a reduced scatter between the groups (shown by their clustering closer to the regression line) and also by a decreased scatter within groups (shown by the smaller dispersion of the individual deviations from the mean deviation for the group). There is still evidence of stratification, however, so that brain weight does not by itself account for all the extractable lifespan variance. 122 George A. Sacher Next we may ask whether Ufespan can be predicted more accurately by a combination of brain and body weight than by either of them alone. The answer to this question is given by the multiple regression of lifespan on body weight and brain weight (Appendix Id), X 0'6S6z - 0-2252/ + 1-035 (4) it! 1.0- .5 1.0 1.5 MULTIPLE REGRESSION VARIABLE 2.0 Fig. 4. Relation of lifespan to multiple regression variable defined by Equation (4) or Equation (8). Symbols as defined in legend to Fig. 2. This regression removes 84 • 4 per cent of the lifespan variance. This is not a significant increase over the 79 per cent removed by regression on brain weight alone. However, the scatter diagram (Fig. 4) suggests that the multiple regression has further reduced the differences between groups. It is possible that a more detailed statistical analysis, in which a separate allometric relation is obtained for each of the major taxonomic Relation of Lifespan to Brain and Body Weight 123 subdivisions of mammals, will further increase the goodness of prediction, since Figs. 2, 3 and 4 give evidence that the relation of brain weight to body weight for the different groups cannot be described by a single allometric relation. Count (1947) is only the last of a number of authors to point this out. Discussion of the independent roles of brain and body weights will be facilitated by the use of a transformed variable. Brain weight and body weight are closely related variables, for the regression of brain weight on body weight (Appendix le) z= 0-666?/ - 0-888 (5) accounts for 91 • 7 per cent of the brain weight variance. The deviation of an individual brain weight value from the regres- sion line is the logarithm of that fraction of the brain weight of the species that is independent of the overall regression of brain weight on body weight. This deviation is defined to be a new variable, w, which is given by the equation w = z - 0-6662/ + 0-888 (6) This quantity is called the index of cephalization. It should be understood that this is by definition a measure of brain de- velopment that is orthogonal to body weight. The definition contains none of the a priori considerations that have fre- quently entered into the definition of this quantity since the time of Dubois (1924). Von Bonin (1937) has urged that the index of cephalization be defined in this objective fashion. The regression of lifespan on index of cephalization (w) is found to be (Appendix If) X = 0-636r£; -i- 1-283 (7) By the definition of w, the regression coefficient for x on w is numerically equal to the partial regression oi x on z in Equa- tion (3). However, the variance of w is but 8 - 3 per cent of the variance of z. In consequence the sampling error of ft^.^ is larger than that for b^^y (Appendix If), and the variance removed by the regression of a; on w is 23 per cent rather than 124 George A. Sacher 84 per cent (Appendix 1). However, the coefficient of regres- sion of X on w is more than four times larger than its standard error, so the regression is highly significant. The multiple regression of lifespan on body weight and index of cephalization is found to be (Appendix Ig) X = O'QSQw + 0-198Z/ + 0-471 (8) It will be noted that the partial regression ofxony in Equation (8) is numerically equal to the coefficient of total regression of cconyin Equation (2). This follows from the fact that y and w are orthogonal variables, so that the regression of cC on i/ is completely independent of, and unaffected by, the regression of X on w. One further dimension of mammalian constitution that has been measured for a large number of species is that of meta- bolic rate. The great amoimt of data accumulated by many investigators, and especially by Rubner, Benedict and Brody, has been masterfully organized in Brody's treatise on Bio- energetics and Growth (1945). Brody has shown that the relation between basal or resting metabolic rate and body weight for warm-blooded vertebrates (including birds) follows a power law relation with great precision. The regression of logarithm of specific metabolic rate, m (in calories per gram per day), on log body weight is (Appendix Ih) m= - 0'2Q6y + 1-047 (9) The correlation coefficient is over 0-99 (Brody, 1945). In view of this high correlation, the partial regression of specific metabolic rate on index of cephalization must necessarily be small. We can therefore assume tentatively that this cor- relation is zero and substitute m (given by Equation (9) ) for ?/ in Equation (8). The resulting equation for the regression of lifespan on metabolic rate and index of cephalization is X = 0-636i£; - 0-744m + 1-252 (10) Sampling errors and residual variance for this relation cannot be given. Relation of Lifespan to Brain and Body Weight 125 It can be concluded that two independent factors are re- quired to account for the observations. One of these may be identified by the highly correlated pair of variables — metabolic rate and body weight. The other factor may be identified by the index of cephalization, which is orthogonal to body weight and must also be nearly orthogonal to metabolic rate (see above). Both of these factors are represented in the brain weight, so that brain weight alone is almost as good a predictor of lifespan as brain weight and body weight combined. Although the existence of two independent factors is very reliably established, their resolution into the two orthogonal terms specified above is not unique. The extension of these procedures of multivariate analysis to larger sets of anatomical and physiological variables may make it possible in future to identify these factors more confidently, and perhaps also to establish the existence of one or more additional factors governing length of life. In spite of the coarseness of the measures used, this first statistical investigation of the allometry of lifespan has been rewarding. Much more remains to be learned from a more detailed examination of these relations within individual orders of mammals, and also in other vertebrate classes, birds in particular. It is to be expected that the numerical values of the coefficients will differ in these different groups, for brain function can be specified by a single number such as total mass only to the degree that brain structure is describable in all its anatomical details by a single allometric coefficient. These same considerations apply to the somatic dimensions. This completes the discussion of the statistical analysis and the implications of these findings for the theory of ageing in mammals will now be considered. First, let us consider the above findings in terms of a theory of ageing put forth by Rubner almost exactly a half- century ago (1908). He took note of the fact that several species of domestic animals with markedly different body sizes and life- spans all had lifetime basal energy expenditure of about 126 George A. Sacher 200 kilocalories per gram. The evidence adduced by Rubner was essentially that in Table I. The lifespans he attributed to some species in the table are out of line with currently accepted values, and introduction of the more accurate values would considerably weaken his evidence. Although Rubner' s own evidence is hardly adequate to support his thesis, the results of the present analysis indicate that his conjecture nevertheless has considerable merit. His hypothesis may be reformulated to say that the lifespan of a species varies inversely as its basal metabolic rate or, in the notation employed above, X = — 1-OOm + constant (11) Equation (10) shows, however, that the empirical relation between these variables is numerically X = — 0'75m + constant (12) The significance of the difference between the theoretical and observed coefficients cannot be tested rigorously, so we cannot say whether this difference is in fact significant. However, Rubner's hypothesis that lifespan varies inversely as the first power of the metabolic rate is in any event subject to quantita- tive revision, for he considered only the basal metabolic energy. There is no good reason to distinguish between the resting and active energy expenditure with respect to their effects on length of life. Furthermore, the relation of active energy expenditure to body size is not accurately known. Hence, it can only be concluded at present that the empirical findings are in accord with the general hypothesis that the attainable length of life of a mammalian species is dependent in part on its rate of energy dissipation. Rubner's original discussion of the energetic theory called attention to the fact that the lifetime energy expenditure for man is seriously out of line with the values calculated for domestic animals (Table I). This discrepancy was in fact one of the considerations that inspired me to undertake this Relation of Lifespan to Brain and Body Weight 127 Table I Lifetime energy expenditure (cal./g.) for several, species OF DOMESTIC ANIMALS AND MAN (FROM RUBNER, 1908) Species Horse Cow Dog Cat Guinea pig Man Body Wt. kg. 450 450 22 3 0-6 70 Length of life (years) 30 26 9 8 6 100 Lifetime energy expenditure (cal./g.) 170,000 141,000 164,000 224,000 266,000 800,000 allometric analysis. I had been working on a theory of mortaHty and ageing which took particular cognizance of the role played by physiological fluctuations (Sacher, 1956, 1958). The essential points of this theory are : (a) mortality is essentially a random process in that the circumstances leading to death in an individual case cannot be predicted with certainty ; (b) the physical basis for this uncertainty is to be found in the fluctuations of physiological state that are inevitably present in living systems; and (c) the magnitude of the physiological fluctuations is determined by the interplay between the random impinging disturbances (of external and internal origin) and the regula- tory mechanisms that act to limit the magnitude and duration of their effects on the organism. A central problem in the mathematical development of the theory is that of deriving the quantitative dependence of the probability of mortality in given circumstances on the para- meters of the physiological fluctuation process. The important implication of the theory in the present context is that the attainable length of life for a species depends on the precision of physiological regulation that the species is capable of. The logical or mathematical developments that lead to this 128 George A. Sacher inference will not be discussed, because in qualitative terms it is readily evident, and a quantitative mathematical statement is not testable at present owing to lack of appropriate data on precision of physiological regulations in different species. In the absence of data that would permit a direct compari- son of theory with experiment it was finally resolved to approach the problem indirectly, by introducing the mediating assumption that the precision of physiological regulation is directly related to the degree of encephalization. Several lines of published evidence suggested that such an approach might be fruitful. First, there was the great discrepancy between the human lifespan and that of other species, and also Rubner's statement about the discrepancy between man and other mammals in lifetime energy expenditure (see above). Second, Flower's tabulation of lifespans of mammals brought out the important point that man's long lifespan relative to that of other mammals is not unique, for primate species in general live longer than non-primates (Flower, 1931). Finally, there was the massive literature on the index of cephalization which indicated that this varies over wide limits (von Bonin, 1937), thus making possible an evaluation of the contribution of brain weight independent of body weight. The statistical analysis presented above strongly supports the hypothesis that lifespan is specifically related to brain development. To establish my thesis completely I should next justify the mediating assumption that the overall pre- cision of physiological regulations is governed by the central nervous system. To do so would take me far beyond the bounds of this paper, for it would entail a review of the vast and rapidly increasing literature on the control of vital physio- logical functions by the central nervous system, and by the cerebral cortex in particular {Ciba Found. Symp. Neurological Basis of Behaviour, 1958). The developments of the last few years have done much to dispel the dichotomy between "higher" and "lower" forms of behaviour in so far as their representation in the nervous system is concerned, and we Relation of Lifespan to Brain and Body Weight 129 tend more and more to the view that the central nervous system participates in the vegetative processes continually. I therefore conclude that my assumption about the close relation between brain development and overall precision of physiological regulation is justified by our present knowledge. Moreover, I am confident that this relation will find concrete support when a quantitative comparative physiology comes into being which will make possible direct tests of more specific forms of this general assumption. Identification of the factors which determine the character- istic ageing rates, and hence the lifespans of mammalian spec- ies, is one of the central problems of gerontology. In recent years several authors have proposed theories of ageing which posit a relation of ageing to the spontaneous mutation rate (Szilard, 1959) or to loss of information content (Yockey, 1958). Whatever other utility these theories may have, they contribute little to our understanding of ageing, for their authors fail to grasp the essential point that the spontaneous mutation rate and the ageing rate are concomitant species characters, so that to account for one in terms of the other is merely to restate an observed fact. It would undoubtedly be widely acknowledged that ageing must ultimately be accounted for in terms of irreversible alterations in the molecular make- up of organisms, and that gene mutations are an especially important class of such irreversible molecular changes. Given this consensus, the basic question is: why does the species mutation rate tend to be proportional to the mean death rate, or inversely proportional to the lifespan? I shall next discuss this question in the light of the energy dissipation and fluctua- tion hypotheses. The occurrence of irreversible molecular changes in bio- logical systems can be exhaustively discussed under four headings, as follows: (a) thermal denaturation — alterations of molecular struc- ture that are essentially due to thermal energy, and occur at a rate that is primarily dependent on the temperature; AGEING — ^V — 5 130 George A. Sacher (b) errors of synthesis and denaturations that occur in the steady state of metabohc activity — the probabihty of produc- ing incorrect molecules per unit time arising from inherent imperfections of the metabolic process (inadequate specificity of enzymes, presence of by-product metabolic poisons, etc.) in the ideal steady state for the species as this is established by natural selection operating on the genetically controllable thermodynamic properties of enzymes ; (c) errors in synthesis and denaturations occurring as a result of deviations from the steady state ; (d) adventitious poisoning of environmental origin — radi- ation, poisons, etc. (this class will not be discussed here). These classes of determiners are distinguishable by the way in which the observed ageing rate (mutation rate, etc.) depends on certain environmental and system variables. Thermal inactivation would operate equally on almost all mammalian species since they have (except for bats and some other forms) essentially the same constant body temperature. Thus thermal denaturation can be ruled out as a factor determining the different rates of ageing in different species. The error rate in the metabolic steady state should have a direct, essentially proportional relation to metabolic rate and in consequence a determinate relation to body mass. Hence, if this is an important factor in determining the rate of accumulation of ageing events one would expect lifespan to vary as an inverse function of metabolic rate. This is the Rubner hypothesis, which was seen above to be in general accord with the data. The error rate due to fluctuations away from the steady state would tend to be smaller in species possessing superior physiological regulatory mechanisms, for in these species the mean square deviation from the steady state is smaller. The evidence on this point, deriving from the relation of lifespan to index of cephalization, is indirect but nevertheless persua- sive, as was discussed earlier. Thus the comparative, allometric approach has shown us Relation of Lifespan to Brain and Body Weight 131 that the rate of accretion of ageing injury in mammahan species increases as the metabohc rate increases and decreases as precision of physiological regulation increases. Therefore the irreversible changes that underlie ageing are not necessarily and inevitably determined by the rate of dissipation of energy. The intervention of improved physiological regula- tions, by maintaining the average value of the milieu interieur nearer to the biochemically ideal, and by reducing the magnitude of the fluctuations away from this average, acts to reduce the probability per unit time of irreversible changes, and thus to reduce the rate of ageing. This has been a highly significant feature in the evolution of mammals, and it remains to be seen whether man can take conscious advantage of this principle to bring about a real decrease in his rate of ageing. Appendix 1. Relation of log lifespan to log body weight, log brain weight, index of cephalization and log specific metabolic rate. Also mean values, variances, and standard errors of the coefficients*. a. Mean values and standard errors of the variables, and their variances. Log lifespan (years) ^ = 1 -283 ± -042 SI = -10900 Log body weight (grams) ^ = 4-099 ± -163 SI = 1 -67158 Log brain weight (grams) 2 =1-841 ±-113 SI = -80666 Index of cephalization w= 0+ -032 S^ = -06570 b. Log lifespan (x) on log body weight (y). See Equation (2). b^ = -198 ± -021 Slj, = -04402 * The total regression of variable x on variable y is denoted by bxj,. The corresponding regression equation is X — X = bxy(y — y) The partial regression of x on y, independent of the regressions of x and y on a, third variable, z, is denoted by bxp.z. The multiple regression equation is of the form X - X = hxy.z{y -y) + h^y.ziz - z) The total variance of variable x is denoted by S^. The variance of x that is not accounted for by regression on the other variable or variables in the regression equation is denoted by S^y, S^^y etc. 132 George A. Sacher c. Log lifespan {x) on log brain weight (z). See Equation (3). b„ = -325 ± -022 SI, = -02403 d. Log lifespan (x) on log body weight {y) and log brain weight (z). See Equation (4). b^^^ = --225 + -045 Sl^ = -01730 b^^^ = -636 ± -065 e. Log brain weight (z) on log body weight {y) ; definition of index of cephalization {w). See Equations (5) and (6). h^ = -666 ± -025 Sly = -06678 w = z- -666?/ + 0-888 f. Log lifespan {x) on index of cephalization (w). See Equation (7). b^,^ = -636 ± -143 Sl„ = -08381 g. Log lifespan (x) on log body weight (y) and index of cephalization (w). See Equation (8). K^o = -198 ± -013 Sl^ = -01730 K.., = -636 ± -065 h. Relation of log specific metabolic rate (m) to log body weight {y) as given by Brody (1945); regression of log lifespan (x) on index of cephalization (w) and log specific metabolic rate (m). See Equations (9) and (10). m = - -2661/ + 1-047 h,„„ = -636 6^.„ = --745 'xw.m REFERENCES BoNiN, G. VON (1937). J. gen. Psychol, 16, 379. BouRLiERE, F. (1946). Annee biol., 22, 249. Brody, S. (1945). Bioenergeties and Growth. New York: Reinhold Publishing Co. Ciba Found. Symp. Neurological Basis of Behaviour (1958). London: Churchill. Comfort, A. (1956). The Biology of Senescence. New York: Rine- hart. Comfort, A. (1958). J. Geront, 13, 342. Count, E. W. (1947). Ann. N.Y. Acad. Sci., 46, 993. Dubois, E. (1924). Proc. kon. med. Akad. Wet., 27, 430. Flower, S. (1931). Proc. zool. Soc. Lond., 145. GuMBEL, E. J. (1958). Statistics of Extremes. New York: Columbia University Press. Relation of Lifespan to Brain and Body Weight 133 Hald, a. (1952). Statistical Theory with Engineering AppUeations. New York : John Wiley and Sons. Quiring, D. P. (1950). Functional Anatomy of Vertebrates. New York: McGraw-Hill. RuBXER, M. (1908). Das Problem der Lebensdauer und seine Bezie- hungen zurn Wachstum und Ernahrung. Munich: Oldenbourg. Sacher, G. a. (1956). Radiology, 67, 250. Sacher, G. a. (1957). Anat. Rec, 128, 616. Sacher, G. A. (1958). In Symposium on Information Theory in Biology, ed. Yockey, H. P., et ah, p. 317. London and New York: Pergamon Press. Szilard, L. (1959). Proc. nat. Acad. Sci. {Wash.), 45, 30. Walker, E. P. (1954). The Monkey Book. New York: Macmillan. Yerkes, R., and Yerkes, A. (1929). The Great Apes. New Haven: Yale University Press. Yockey, H. P. (1958). In Symposium on Information Theory in Biology, ed. Yockey, H. P., et al., p. 297. London and New York: Pergamon Press. DISCUSSION Danielli: The simplest physical interpretation of your regression equation, Mr. Sacher, is that it is an advantage to have a brain, and a disadvantage to have a body ! Wigglesworth: Barcroft's thesis (Barcroft, J. (1934). Features of the Architecture of Physiological Function. Cambridge University Press) was that a constant internal environment was primarily essential for the functioning of the brain, and that for higher development of the brain you needed a more constantly regulated internal environment. As I understand your theory, that same constancy will also be favourable to longevity, so that cephalization and longevity should go hand in hand. Verzdr: Do you not think that lifespan as a measurement of ageing trends is a bad measurement, Mr. Sacher ? Lifespan is the most un- certain of all our records. The mortality curves are not straight ; they are always an "S " shape, and the right side of that "S " shape, especially, is extremely long, so that exceptionally long lives are particularly noticed. Wouldn't it be much better to relate all our age theories not to the maximal lifespan, but to something like a 50 per cent survival of a certain population ? That would make the whole thing experimentally much more certain. Sacher: I agree that maximum lifespan is a gross measure. That it is a bad measure I would not agree. Lifespan is an extremum statis- tic, and can be handled in a perfectly rigorous fashion by statistical 134 Discussion reasoning, as is shown in Gumbel's treatise (1958). Moreover, life expectations or the median survival times are available for only about six of the 100 or so species that ideally we would like to use in this sort of comparative analysis. Finally, as I pointed out in my paper, maximum lifespan is a better estimator of the intrinsic ageing characteristics of the population than is the median or average survival. However, I do not advocate its use in the analysis of laboratory data, or in any situation where life-tables are available. Incidentally, I have been concerned about the problem of what is proper to use as a lifespan for man. Zoo animals are not kept under ideal conditions, as they get little individual attention or medical care. Therefore perhaps some earlier state of human culture would be more comparable. It is interesting that you can get data on the ages of fossil skulls back to the Palaeolithic (Vallois, H. V. (1937). Anthropologic, Paris, 47, 499; Weidenreich, F. (1939). Chin. med. J., 55, 34). Peking man, in a sample of only six skulls, yielded one that probably had an age of over 50 years. They lived under far poorer conditions than our zoo animals today. Neolithic man lived to more than 70 years, even according to rather small samples. Comfort: I think you have been taking much more plausible maximum ages than Rubner did, Mr. Sacher. Rubner gave the life- span of a dog as nine years and of a cat as eight years, which is quite arbitrary. Cats can quite possibly live for 30 years and the extreme credible record for a dog is between 20 and 25. There are much greater discrepancies between the modal and maximum records for animals, and the acceptable mode and maximum for man. I think most people would agree that 110 to 120 is the extreme limit for which there is any good evidence in man, in spite of the 140- year-old Russians. Most cats die before they are 16, but a few have lived very nearly twice that time. One has also to beware among animals of the possibility of very long-lived genotypes and of the differences between hybrid and inbred strains. Sacher: The extreme error in the individual lifespan records is probably a factor of 2 or so, but if the errors are random, their only effect is to weaken the degree of order observed. Holt: Have you looked at data for aquatic mammals ? They have some exceptionally large body sizes, without corresponding increases in brain size, and there are many published age determinations for them, but perhaps not maximum lifespans. Sacher: When I did this work I did not have enough data on aquatic mammals, but I want to study them in future. Maynard Smith: There is a possibility of bias in estimates of this kind since most of the small mammals in your sample are rodents, Discussion 135 and most of the large ones are ungulates. It would be interesting to know whether, if you calculate your coefficients just on rodents, just on primates, and just on ungulates, you would get results which are at least approximately consistent with those you get on the whole sample combined. Sacher: I only know this qualitatively and graphically. The values of the allometric coefficients vary considerably from order to order. The goodness of fit would have been greatly improved if I had omitted the ungulates, because they have quite different allometric relations of body weight to brain weight, and in effect they made the overall relation poorer. Rotblat: I was very glad to see that you describe life processes in terms of numbers. I was a little bit disturbed, however, when you ended up by introducing a term which cannot be expressed in numbers, namely stability or adaptability of physiological function. Can you put this in some kind of quantitative relation to the index of cephalization, or the mitotic activity ? Have you any indication that there may be some ways of increasing lifespan by increasing stability ? Sacher: The term " stability " in my thinking is a general term that <^ subsumes all the properties that have to do with mortality, stress resistance, length of life, etc. In this sense, an animal that lives longer has greater stability, and a species that has a lower mortality rate for a given disease or stress has greater stability. In other words, stability is a general physical character of organisms. If one holds the point of view that all these things are fated, determined in advance by the genotype, then there is no reason for talking about stability. But if one thinks of organisms as dynamic functioning systems whose probabilities of failure arise from their function (Sacher, 1956, 1958), then stability is a natural term and it becomes reasonable to think of improving the stability characteristic of organisms. We cannot replace any of our bodily elements with better ones, as an engineer can replace vacuum tubes, but it might be possible (remembering that the central nervous system is im- plicated in every physiological activity, and that these responses are conditionable) eventually to evolve a kind of ontogeny and con- ditioning that would make for more stable physiological functioning in the given environment. Rotblat: This seems to be going around in circles. You have ex- plained the span of life in terms of stability and vice versa. It seems to me that mutation rate is a quantity outside the circle, because you can say that there is something which goes on all the time inde- pendently of us. 136 Discussion Sacher: If mutations are considered to be molecular events, then one has to ask why the molecules in the mouse mutate some thirty times faster per unit time than they do in man. I have shown that two physical characteristics of mammals, namely the metabolic rate and the goodness of physiological regulation, account for most of the lifespan variance. Some fraction of the remaining variance is undoubtedly under specific genotypic control, but this is a small part of the total. Thus it would appear to be true that species tend to have the maximum attainable lifetimes permitted by their body size and complexity of organization. Danielli : You suggested that there might be a standard amount of metabolism which was permissible per gram of tissue before it deteriorated beyond hope. Could any information about this be obtained, perhaps in fish, by using a poison such as dinitrophenol, which causes a good deal of useless metabolism to go on ? This might enable one either to discover that the life expectancy was a function of the amount of oxygen utilized in respiratory processes, or else to distinguish between one type of metabolism which has an ageing effect, and other types of metabolism which have not. It is probably easier to keep up a constant concentration of dinitrophenol in fish than it is in many other animals. Sacher: I have not yet had the opportunity to do such experi- ments ; it would certainly be extremely productive to use metabolic poisons. In warm-blooded mammals one also could replace part of the basal energy production by producing heat internally with radiofrequency heating. Anything that would produce a dissocia- tion between the amount of metabolism and the other physical characteristics of the organism would be extremely valuable. I would also like to determine whether one could systematically yet diffusely decrease the general regulatory ability, perhaps by destroying the brain to a certain degree, with sonic radiation or diathermy. Can anyone suggest how a uniform controllable deterior- ation could be produced which could be followed in terms of its effect on survival ? Lindop : Could one use colchicine as a mitotic inhibitor in different doses ? Danielli: It might have some effect, but I should have thought it might be better if you could use something of the nature of a cholinesterase inhibitor, or strychnine. Comfort: An experiment with dinitrophenol was done upon mice by M. L. Tainter (1936. Proc. Soc. exp. Biol. (N.Y.), 31, 1161). Mice treated over a period of time did not seem to have their lifespan very much affected — certainly not in proportion to their metabolic rate. Discussion 137 Danielli: That would definitely mean that it was not just a respiration effect. Comfort: With many of these animals you may increase their metabolism and at the same time decrease their appetite, or some- thing like that. One has to be careful. Sacher: If it did not have any effect on body temperature it would not greatly increase the overall metabolism would it ? Comfort : I do not know whether there was that effect or not — the paper may say. Danielli: Dinitrophenol would normally increase the metabolism a lot, unless there were some compensatory mechanisms, and one would expect appetite to increase rather than diminish. Bourliere: Long ago T. B. Robertson (1928. Aust. J. exp. Biol, med. Sci., 5, 69) found that continuous treatment of the mouse with desiccated thyroid, in quantities that stimulated growth, also shortened the lifespan. Berg: We are studying the effect of thyroidectomy on the lifespan of the rat. Maynard Smith : Another possible way of studying the effects of metabolic rate on longevity is to use different genetic strains of the same species. The most striking differences in longevity you can get in flies are between inbred and outbred ; outbred flies will live about twice as long as inbred ones. This certainly is not associated with a lower rate of metabolism in the outbred flies. They are not animals in which it is easy to measure the basal metabolic rate, but if one judges it by rate of eg^g production, for example, the hybrids are laying eggs at about twice the rate of the inbreds, as well as living twice as long. Also the hybrids are much more active. Their greater longevity is much more easily explained in terms of the other con- cept that Sacher used, namely that the hybrids in many respects, both physiologically and developmentally, seem to have far better stabilizing mechanisms than do inbreds, and that what is wrong with inbred animals is that they are just not good at regulating against anything. Sacher: Yes, that is a view which is put forward systematically in the concept of genetic homeostasis. Maynard Smith: I think it is a true one. Sacher: In general terms I think it is true also. We have the same phenomenon in mice but not to the same degree. Verzdr: Thyroxine treatment of the tadpole and axolotl, which — as you know — leads to transformation from larval to adult forms, always leads to a shortening of life. The transformed animals never survive long. 138 Discussion Nigrelli: Is that true of well-fed and of starved tadpoles ? Well- fed tadpoles should not metamorphose as rapidly as starved ones. Verzdr: In the cases which I know of the tadpoles were well fed. Lindop: Mr. Sacher, you were trying to put something forward mathematically, using your concept of lifespan. I tried to relate your criterion to what we were doing, and I found that I could not apply it. In a discussion group like this, where there are so many different disciplines, and we are each putting forward our own specialist information, a short discussion on what is the most useful concept of lifespan for us to be able to compare our different groups might be helpful. People who have experiments which are half-way through would then know what others want them to do for the rest of them. Can you use your mathematical interpretations to get a mathe- matical correlation between lifespan in naturally occurring popula- tions and in populations where we ha^^e altered the lifespan ? By irradiation we have altered lifespans both by your definition and ours, but we have not changed the index of cephalization. Sacher: There are certainly many factors affecting length of life. The thing has to be put into perspective. I have used the lifespan I defined above — the maximum attained life — only in the present context of doing a comparative study on a very broad scale. When I am working on laboratory data I usually use the life expectation, and specifically the after-expectation of life from the beginning of exposure. In order to characterize the effects of radiations on popul- ations, we have found that it is particularly convenient to discuss these in terms of the log rate of mortality (Gompertz) curve (Sacher, 1956; Brues A. M., and Sacher, G. A. (1952). In Symposium on Radiobiology, ed. Nickson, J. J., p. 441. New York: John Wiley). Before discussing this I wish to point out that any one of the life- table functions contains the same amount of statistical information as any other, as long as you have not lost information by rounding, setting up large class intervals, etc. The reason for preferring some particular analytical function of the basic data is that it seems to give the clearest insight into underlying mechanisms. Cohorts of mice kept under laboratory conditions have life-tables such that the plot of logarithm of rate of mortality (Gompertz transform) either is a straight line or shows a moderate amount of curvature. In various mouse strains, the slope of the best-fitting straight line at advanced ages (omitting the mortality primarily due to infectious disease in young mice) does not vary significantly. In hybrids showing marked hybrid vigour the slope again remains unchanged, and the increased survival is due to a decrease in the Discussion 139 intercept values. A single dose of X-rays given in early adult life is followed, after a latent period of 100-200 days, by a displacement of the Gompertz transform parallel to itself (Fig. 1). In cases where the Gompertz transform shows some curvature, it is possible to infer that this displacement is a translation to the left on the time axis. The interpretation is that the Gompertz transform is a linear measure of the amount of ageing injury present. A single X-ray dose produces a residue of permanent injury, and this is manifested by a displacement of the Gompertz transform by a constant amount. 1000- o o o o* IT. UJ Q. o UJ O o 200 400 600 800 TIME FROM BEGINNING OF EXPOSURE Fig. 1 (Sacher). Schematic representation of the long-term effect of single or repeated exposure to ionizing radiations on the Gompertz function (logarithm of the age- specific rate of mortality) for mammals. The age-trend of mortality is considered to conform to the Gompertz equation (compare with equation 1 and text) '°s(x-f) = "'g^ + OCX where A' is the number living at age x. O — unirradiated population; S^, Sg, — populations given single exposures at time zero; R^, Rg — populations given repeated or continuous exposure beginning at time zero. Early portion of lines S^ and Sg dotted to indicate time needed for displacement to attain its steady value. 140 Discussion If each X-ray dose produced a constant displacement of the Gompertz function, and if they added with one another as well as with the basic ageing trend, then a cohort exposed daily throughout adult life should show a constant divergence from controls, and a set of different daily doses should produce a fan of Gompertz curves. This is in fact seen (Sacher, 1956). The parameters of the daily dose response are consistent with those of the single dose response. o > cc If) I- o < a: .00 — =^ ■\ N s ^ - .80 \ \ .60 " \\ \ \ \ \ \ \ V \ V \ .40 - \ \ \ \ \ \ V \ \ \ \ \ \ \ \ \' \ "■ .20 n 1 r\ \ \^ \ xS^rX^S, \p 200 400 600 800 TIME FROM BEGINNING OF EXPOSURE Fig. 2 (Sacher). Survivorship curves corresponding to the schematic Gompertz functions in Fig. 1. The "single dose" curves can be superimposed on the "control" curve by sliding them to the right. The "repeated exposure" curves can be superimposed on the "control" by multiplications of their time scales by scaling factors. This mode of analysis therefore gives a parsimonious description of radiation mortality, and relates it to the ageing process in terms of a life-table function, the Gompertz function, where properties are consistent with the hypothesis that the function is a linear measure of the amount of ageing injury present in the population. A theoreti- cal justification of this hypothesis can be given in terms of the con- cepts of physiological fluctuations and probability of failure that were introduced in the text (Sacher, 1956). In view of the earlier point, that any one of the life-table functions contains the same information as the others, it follows that any other desired quantities, such as expectations, medians, deciles, Discussion 141 etc., can be computed in turn. My chief concern was to validate a theory of radiation mortahty and ageing. However, empirical analysis and theoretical analysis should have the same goals of parsimonious description. Thus the Gompertz function, which is a theoretically meaningful one, should also be best for empirical analysis. Verzd?': Could you describe the same thing with survival curves, Mr. Sacher ? Sacher: An animal following a single radiation dose acts at a given age like a control animal at somewhat greater age. The irradiated population tends to show shallower survivorship curves which can be translated and scaled so that they can be superimposed on a con- trol population of a later starting age (Fig. 2). This can be accom- plished without changing the time scale, and corresponds to the fact that single X-ray doses displace the Gompertz function parallel to itself without change of slope. If we give daily doses of irradiation, the effect is not as if we had set the clock forward but rather as if we had changed the regulator, so that the clock runs faster. Thus, con- comitant with the decrease in survival time there is a steepening of the survival curve in the daily dose condition. That corresponds to the fan of lines on the log rate of mortality plot. Maynard Smith: In comparing life-tables based on wild popula- tions and on laboratory populations, I think what both Dr. Comfort and Mr. Sacher have had in mind here is that what such distributions are most likely to have in common is the maximum lifespan; the oldest individuals in wild populations may correspond roughly in age to the oldest individuals in laboratory populations, but the two distributions have little else in common. Sacher: I can agree with that. It is not yet possible to reach a meaningful correlation between life-tables in the field, and life- tables in controlled environments. These conditions are so far apart that we cannot discuss the respective life-tables in terms of common parameters. It would seem that there have to be intermediate grades of environment between the wild and the laboratory. Lindop: Is there any one method of investigation in which, instead of going through the changes gradually, we could correlate them more rapidly ? For instance, one might take the causes of death in wild animals and the causes of death in laboratory animals, exclude from each group the causes which are not in common, and see how the survival curves fitted for the causes of death which are in common. Sacher: That certainly could be done if they had enough causes of death in common. A REVIEW OF THE LIFESPANS AND MORTALITY RATES OF FISH IN NATURE, AND THEIR RELATION TO GROWTH AND OTHER PHYSIOLOGICAL CHARACTERISTICS R. J. H. Beverton and S. J. Holt Ministry of Agriculture, Fisheries and Food, Fisheries Laboratory, Lowestoft; and Fisheries Biology Branch, F.A.O., Rome Studies on the dynamics of fish populations have received a major impetus in recent years owing to the need to provide an adequate scientific basis for conservation. One aspect of these studies is the measurement of fish longevity and the force of natural mortality in fish populations. In this contri- bution we attempt to review the present state of knowledge on these questions. In so doing we have two objectives in mind. One is to present the data on longevity and mortality in fish for com- parison with what is known for other animals and presented at this symposium; the other is to see to what extent these characteristics are, in fish, associated with size, growth, maturation and certain other physiological factors for which data are available. It has not been possible for us to search through the widely scattered literature as thoroughly as we would have wished. The paper is therefore perhaps best regarded as a progress report from which certain tentative conclusions can be drawn at this stage. Natural mortality and lifespan Many of the fish populations which have been intensively studied are those supporting a major commercial fishery and are therefore ones in which the effect of fishing has profoundly 142 Longevity and Mortality Rates of Fish in Nature 143 influenced the shape of the survival curves and the maximum age recorded in samples. There are, nevertheless, a certain number of instances in which the natural age composition, or ■ t V**^ CLUPEOIDS \ \\ \ ^arenqus \ \ '" \ I '1 • "l I sprattus \ 1 1 1 pallasi \ "^ 1 1 \ 2- O 5 lO 15 20 T 1 1 ■ % ? ''1 \ \ Collionymus \ \ lyra , .\ Fig. 1. Some examples of survival curves in relatively unexploited fish populations. something fairly close to it, has been determined, and some examples of these are shown in Fig. 1. It will be noted that in no case do the data cover the whole lifespan from birth on- wards ; this is because representative sampling of the fry and 144 R. J. H. Beverton and S. J. Holt juveniles is seldom possible. It is known, however, that in many species, and especially in those which lay large numbers of eggs, there is a very high mortality during the first weeks of life ; in North Sea plaice, for example, only about one in ten thousand survive the first few months. Thus the survival curve for fish characteristically descends very rapidly at first and then flattens out, though in viviparous species, and species which lay a small number of eggs but afford the newly- hatched fry some degree of parental care, this initial descent is probably less marked. Even after the early phase of heavy mortality some con- siderable time may elapse before the fish have grown large enough to be retained by the fishing gear, so that a representa- tive survival curve has to begin at some later age when the individuals are first properly represented in the samples. The survival curves shown in Fig. 1 therefore start at the age group which is most abundant in catches, and for comparison all the data have been adjusted to a peak number of 1,000. The numbers are plotted on a logarithmic scale, so that a linear survival curve indicates a constant natural mortality rate independent of age, whereas a downward curve shows that the mortality rate is increasing with age. The broken lines are drawn purely to assist the eye in detecting linearity or departures from it, and where the survival curve is not a straight line the broken lines are drawn through the first and last points. In the long-lived species, of which the examples shown in Fig. 1 are sturgeon (Acipenser fulvescens ; Probst and Cooper, 1954), whitefish (Coregonus clupeaformis ; Hart, 1931) and perch {Perca fluviatilis ; Aim, 1952), the raortality ra,te- seems to be effectively constant over a considerable span of age at about 5 to 10 per cent per year, although in the age groups of sturgeon beyond about 30 years the mortality rate appears to increase. The fluctuations in the data for sturgeon and whitefish are partly due to the fact that sampling was possible for a limited period only and that the age groups refer to Longevity and Mortality Rates of Fish in Nature 145 different year classes of fish whose initial abundance varies considerably. The perch data are unique because they show the survival_of a known number of fish introduced into experimental ponds — in the one case as fry {a) and in the tJtheF'asrfive-year-old fish (b) — and then sampled regularly, for 15^ years and 17 years respectively; within these spans of age there is ho evideiice of a varying age-specific mortality rate. Of the shorter-lived species shown in Fig. 1, nearly all have a survival curve with some degree of downward curvature over nearly the whole range. This is seen particularly clearly Tnjthe herring data (Clupea spp.), of which those for the Norwegian herring (C. harengus; Lea, 1930) are the combined data for a period of twenty years in which year-class fluctu- ation has been largely eliminated. The data for the bullhead (Cottus gobio; Smyly, 1957) and the dragonet (Callionymus lyra; Chang, 1951) are included to show the difference be- tween the survival curves for males and females. In both of these the males have a higher mortality rate and a shorter lifespan, and this is indeed what is usually found where there is any difference between the sexes; we have, however, come across one or two exceptions which are referred to later. The tendency for the natural mortality rate to increase with age, which is noticeable in several of the examples shown in Fig. 1, is, indeed, found more often than not, and for other instances the reader is referred to papers by, for example, Kennedy (1954) on the Lake trout {Cristivomer namaycush), by Wohlschlag (1954a) on the Alaskan whitefish (Leucichthys sardinella), and by Ricker (1949) on several species. This variation of the mortality rate with age reaches an extreme form in species where all or nearly all individuals die at, or soon after, spawning for the first time. The best known instance of this is in the Pacific salmon (Oncorhynchus spp.), which migrates up-river from the sea at between three and five years of age, spawns and then dies. The immature phase of the life-cycle spent in the sea has only recently been studied, 146 R. J. H. Beverton and S. J. Holt but there is no reason to believe that the mortahty rate during that time is abnormally high. Other examples of a catastrophic mortality at, or shortly after spawning include the Tasmanian whitebait {Lovettia seali; Blackburn, 1950); the capelin {Mallotus villosus; Templeman, 1948); the small freshwater atherinid Labidesihes sicculus (Hubbs, 1921) which spawns at about one year of age and then dies off within a further two or three months; and, probably, the dragonet {Callionymus lyra; Chang, 1951). In most of these it is the males which suffer the most severe mortality, the evidence being that a proportion of the females spawn more than once, even though that proportion may be small. Survival curves in fish thus range from effective linearity over the whole of the observed range of age to sharp dis- continuity at the onset of maturity j^ with a wide range of intermediates in which the mortality rate increases steadily with age without obvious discontinuity. This makes it dif- ficult to adopt any single numerical index as an index of life::, span, or of force of mortality, for all species. Thus the maxi- mum age recorded in samples is satisfactory for the species in which the mortality rate increases fairly sharply with age, but is less so in the long-lived species, where it becomes rather critically dependent on the size of the samples and on the accuracy of the age-determination technique. Conversely, the average mortality rate is not a particularly useful measure where the mortality rate is highly age-specific, but is satis- factory in the long-lived species with nearly linear survival curves. For the time being we have therefore tabulated wherever possible both the maximum age recorded in the sample (T^^ax) and the average instantaneous coefficient of natural mortality (M) over the range of age groups which, as far as we could judge, were fully represented in the same samples. These are given in Table I, from which it will be seen that the lifespan of fish can range from little more than a year in several quite unrelated species including Labidesihes, of the mullet family, Longevity and Mortality Rates of Fish in Nature 147 and Lovettia seali of the salmonoid family mentioned above, Hypomesus olidus, the pond smelt, another of the salmonoids, and the dwarf sea-horse {Hippocampus hudsonius), to over 80 years in the Lake sturgeon {Acipenser fulvescens). The maxi- mum recorded age we have found is, in fact, for this latter species (Anonymous, 1954), a specimen 206 cm. long taken in an Ontario lake having been assigned an age of 152 years by examining the structure of the pectoral fin-ray. While it is quite possible that the precise age of such a fish cannot reliably be determined in this way, the work of a number of authors on the longevity of this and related species of sturgeon is con- sistent in showing that they can live to a great age, and it is indeed not unlikely that the occasional truly centenarian sturgeon is still to be found in the more remote water basins as yet untouched by man. Apart from sturgeon and the whitefish mentioned above, other long-lived species include the Arctic char {Salvelinus alpinus; Grainger, 1953) and the halibut (Hippoglossus spp.); whitefish and char are both salmonoids, but sturgeon and halibut are of different sub- orders, so that neither of the extremes of lifespan in fish are confined to a particular taxonomic group. In compiling the data on maximum age in Table I we have not used the records for fish in captivity, of which a recent summary is included in the longevity data given by Brown (1957) and further data are presented to this colloquium by Nigrelli. There is nevertheless a broad agreement between the records from the two sources, and the few instances in which there is reason to believe that the entry in Table I may be substantially below the true maximum age of the species are noted in the legend to that Table. Size and growth Longevity and body sizes are known to be associated in higher vertebrates, especially in mammals (e.g. Sacher, this volume), so that it is of interest to see to what extent the same is true of fish. The growth cycle in fish is, however, more 148 R. J. H. Beverton and S. J. Holt Table I COLLECTED DATA ON GROWTH, MORTALITY, LONGEVITY AND Notes: (i) Loo and K are the parameters of the growth equation (1) given on p. 157, Lm = mean length at maturity. (ii) In a few instances the value of Tmax obtained from the age composition samples longevity of the same or closely related species in captivity, of which a recent sum- ages recorded by Brown are as follows: Gadus virens, 14 years; Melanogrammus Dasyatis pastinaca, 21 years. Species Common Name Locality Clupeoidei Clupea harengus Atlantic herring North Sea C. harengus Atlantic herring Norwegian Sea C. harengus C. harengus Atlantic herring Atlantic herring Lusterfjord (Norway) New Brunswick C. pallasii Pacific herring Canada (west coast) C. sprattus Sprat North Sea Sardinops caerulea California sardine California S. neopilchardus Australian sardine Australasia Gadiformes Boreogadus saida — Arctic Ocean Gadus callarias Cod North Sea G. callarias Cod Barents Sea G. minutus Poor cod English Channel G. minutus Poor cod Mediterranean G. virens Coalfish Norwegian Sea Melanogrammus aeglefinus Haddock North Sea Merluccius merluccius Hake Marmora Sea Pleuronectoidei Citharichthys sordidus Sand dab California Hippoglossus vulgaris Halibut N. Atlantic Longevity and Mortality Rates of Fish in Nature 149 SIZE AT MATURITY OF FISH IN NATURAL POPULATIONS M = natural mortality coefficient, Tmax = maximum age recorded in samples, (indicated by an asterisk) is somewhat below that recorded from records of the mary has been compiled by Brown (1957). For the fish in question, the maximum aeglefimis, 14^15 years; Salmo trutta, 18 years and more; Anguilla anguilla, 55 years; Author Sex (cm.) K M -t^max (sample) (cm..) LmlLoo Burd (unpub.) 30 0-38 0-25 12 24 0-80 /Lea (1930) \Runnstrom (1936) 34 0-27 <0-2 22 28 0-82 Aasen (1952) 21 0-65 0-78 10 — — Tibbo (1957) 34 0-36 — 19 25 0-74 /Tester (1937) \Ricker (1958) 23 0-29 0-56 11 — — Robertson (1938) 13 0-70 11 — — (^Beverton (unpub.) /Rollefsen (1954) \ Taylor (1958) 134 01 — 23 85 0-64 Menon (1950) {T 20 24 0-42 0-40 11 0-9 5 5 11 13 0-55 0-54 Vives & Suau 21 0-97 >2-3 2 — — (1956) Gottlieb (1957) 107 019 015 10* 71 0-66 r Beverton & Holt ^ (1957) 53 0-20 >0-2 10* 26 0-49 tRaitt (1939) Akyuz (1959) {T 44 60 0-13 0 10 0-6 0-5 10 10 23 27 0-52 0-45 Arora (1951) /m \f 30 >30 0-3\ <0-3/ <0-3 7 8 19 <0-63 Devoid (1938) /m \f 170 250 004 0 02 — 30 30 95 132 0-56 0-53 150 R. J. H. Beverton and S. J. Holt Table I — continved Species Common Natne Locality H. stenolepis Hahbut N. Pacific Isopsetta isolepis Butter sole Canada (west coast) Pleuronectes platessa Plaice North Sea Pseudopleuronectes americanus Winter flounder Canada (east coast) Solea vulgaris Sole North Sea Salmonoidei Argentina semifaxiata Coregonus clupeaformis C. clupeaformis C. clupeaformis C. clupeaformis Argentine Whitefish Whitefish Whitefish Whitefish Japan L. Nipigon, Canada Shakespeare Is. Lake, Canada L. Opeongo, Canada Trout Lake, Wisconsin C. clupeaformis Dwarf Whitefish L. Opeongo, Canada Cristivomer namaycush Lake Trout Gt. Slave L., Canada Hypomesus olidus Pond smelt L. Suwa, Japan Leucichthys artedi Cisco Wisconsin, Trout Lake ,, Muskel- lenge L. ,, Silver L. - ,, Clear L. L. kiyi Chub U.S.A. L. sardinella Whitefish L. Ikroavik, Alaska Tasmania Lovettia seali Tasmanian whitebait Mallotus villosus Capelin Labrador Oncorhynchus nerka Sockeye salmon Cultus L., Canada Longevity and Mortality Rates of Fish in Nature 151 Author Sex L(X) (cm.) K M ^max [sample (cm.) LmlLoo 'Thompson & Herrington (1930) Similar to <0-3 ^ Thompson & Bell (1934) //. 5 mlgaris Hart (1948) {T 38 42 0-36 0-26 <1-41 <102 10 18 21 0-47 0-50 Beverton (unpub.) {T 45 70 015 0-08 0-22 012 13 22 25 28 0-56 0-40 Dickie & McCracken (1955) 44 0-4 0-3 >10 32 0-73 f Beverton & Holt < (1957) (^Margetts (mipub.) 39 0-4 :^0-25 >8 — ■ — Hanyu (1956) 19 1-2 2 — — Hart (1931) 50 013 0 17 24 ~27 ^0-55 Hart (1931) 49 009 015 27 ^27 ::^0-55 Kennedy (1943) 70 0-06 <0-5 12 — — Hile & Deason 44 009 — 14 >23 >0-52 (1934) /Kennedy (1943) \Rieker (1949) 14 0-43 1-3 5 — — Kennedy (1954) 56 007 0-6 25 18-4 0-33 rShiraishi (1957) \Sato (1950) 11-12 1-5-1 -8 11-3 -8 1-3 10 2::0-9 Cm V 19 0-65 1 -1 11 6 11 12-5 0-66 J m If 1 TTi 21 0-36 1-2 1-2 1 • 1 3 4, 150 0-72 >Hile (1936) J If 32 006 J. X 0-9 7 14-0 0-44 Jm \f 39 0-27 0-4 0-3 9 11 130(?) 0-33(?) Deason & Hile (1947) {T 28 0-51 <0-9 <0-8 7 10 <18 <0-64 ' Wohlschlag (1954a, b) (and 38 0-40 0-6 11 1 personal comm.) Cohen (1954) Blackburn (1950) 6-7 — — 2 5(?) 0-8(?) Templeman (1948) {T 20 19 0-48 0-48 1-3 5 5 18 17 0-90 0-90 Foerster (1929) 69 0-58 — 6 60 0-87 152 R. J. H. Beverton and S. J. Holt Table I — continued Species Common Name Locality Columbia R., Canada 0. keta Chum salmon 4-year spawners 3-year spawners Salmo salar Atlantic salmon Scotland S. trutta Trout L. Windermere, England Salvelinus alpinus Char Baffin I., Canada ACIPENSERIFORMES Acipenser fulvescens Lake sturgeon Wisconsin J f medirostris '\transmontanus A. nudiventris ^\liite sturgeon Sturgeon California Europe Anguilloidei Anguilla anguilla Eel Windermere Blennioidei Blennius pholis Blenny Welsh coast Callionymoidei Callionymus lyra Dragonet English Channel COTTOIDEI Coitus gobio Bullhead Windermere R. Brathay Cyprinodontiformes Gambusia holbrookii Top minnow Portugal Cyprinoidei Phoxinus phoxinus Minnow Windermere Gasterosteiformes Gasteroteus aculeatus 3-spined stickleback 10-spined stickleback Cheshire >> Longevity and Mortality Rates of Fish in Nature 153 Author Marr (1943) Nail (1927) Frost & Smyly (1952) Grainger (1953) Probst & Cooper (1954) Pycha (1956) Paeeagnella (1948) Frost (1945) Qasim (1957) Chang (1951) ■Smyly (1957) Da Franca (1953) Frost (1943) Jones & Hynes (1950) Sex { (Mostly f ) { /m { ni f Lao (cm.) 120 105 106 102 125 30 150 140 178 300 250 165 17 250 17-5 7-2 7-3 6-5 6-5 0-36 002 003 005 0-06 0 04 0 02 0-30 0-43 0-55 0-7 0-4 0-9 0-5 1-2 0-8 0-55 0-64 1-6 M (3-0) (1-2) l-l(f) 0-94 0-24 0-24 001 003 ^max (sample) 6(f) 5(m) 8* 24 + 24 + 82 30 30 V. small (17)* 0-9 0-96 0-86 11 0-9 0-9 0-8 16 8-12 13-19 ^0-33 2^0-53 Aim (1952) Radhakrishnan 34 37 0 13 0-4 016 >22 4 >13 >0-35 (1957) /Hart (1928) \Ricker (1949) 40 0 14 0-44 13 > 32(f) >0-8 Yokota (1951) {T 105 150 01 01 l-8(?) 0-4r-0-5 4(?)* 7(?)* 40 44 0-38 0-29 Holt (1959a) Fitch (1951 and 1956) Holt (19596) 22 40 46 0-7 0-4 0-25-0'4 <2 0-8-1 0 2 9 4^5 17 32 28-33 0-77 0-80 C^0 67 Appelget & Smith (1951) 119 006 <0-8 12 36 0-30 Herald & Rakowicz 14 2-5 ^1 1 7 0-50 (1951) Strawn (1958) 2 12 2-3 1 2 10 Moore (1951) Tiews (1957) de Sylva (1957) 190 270 236 0-5 0-6 11 0-8 5 13 3-5 — — 156 R. J. H. Beverton and S. J. Holt protracted than in most higher vertebrates, and as a conse- quence the maximum size is often not reached within the range of age covered by the data. This makes it necessary to examine in more detail the actual pattern of growth in fish in order to arrive at suitable indices to correlate with longevity. Fig. 2 gives a few examples of the growth in length of fish. We use length as the measure of body size rather than weight because, as can be seen from Fig. 2, growth in length nearly always follows a simple curve without an inflection.* This is true whether the species is one which can grow to a large or to a small size, and whether it completes its growth pattern rapidly or slowly. Examples of all these are included in Fig. 2. In the upper part of the diagram is shown the growth of sturgeon {A. nudiventris), which both grows slowly (i.e. com- pletes its growth pattern slowly) and also attains a large size, that of sockeye salmon (0. nerka; Foerster, 1929) which grows to a fairly large size but does so rapidly, and that of whitefish {Coregonus clupeaformis) which grows slowly to a rather smaller size. In the lower part of Fig. 2 are some examples of the smaller species, and for these the scales of both length and age are increased roughly fivefold; to aid comparison, the growth of Lusterfjord herring {Clupea harengus; Aasen, 1952) is shown in both parts of the diagram. It will be seen that although the smaller species usually develop their growth pattern more rapidly than do the larger species, there is still quite a range of variation. Thus, for its size, the blenny {Blennius pholis; Qasim, 1957) is relatively slow-growing, whereas Labidesthes has virtually reached its maximum size in little more than a year; and the 10-spined stickleback (Pygosteus pungitius; Jones and Hynes, 1950), although growing to little more than half the size of Labidesthes, takes several years to do so. * Since the growth of most fish is closely isometric after the juvenile phase, the curve of growth in weight is approximated to by cubing that of growth in length. This produces a weight-growth curve which has an inflection at about one-third of the asymptotic weight. Longevity and Mortality Rates of Fish in Nature 157 It is found that all the growth curves shown in Fig. 2 and, indeed, those for a great many other species of fish, can be adequately represented by the mono-molecular or inverse exponential equation which, in its simplest form, is i, = L„(l-e-«) (1) ISO lOO L(cm.) Acipenser nudiventris L(cm) Oncorhynchus y nerka .oreqonus T clupeaformis Clupea harenqus 25 30 4 5b ACE (yr.)- Fig. 2. Some examples of curves of growth in length of fish. where l^ is the length at age t, L^ is the maximum or asympto- tic length and ^ is a constant which determines the curva- ture of the growth curve, that is, the rate at which the asymp- totic length L^ is approached (see, e.g. von Bertalanffy, 1938; 158 R. J. H. Beverton and S. J. Holt Beverton and Holt, 1957). There is evidence that both the constants L^ and K have a physiological significance, as will be mentioned later; at this stage we need only regard equation 200 r I50- 100 Fig. 3. Growth data of Fig. 2 plotted as length at age t against length at age < + !. The slope of the line drawn through the points is e-^ and the inter- section with the bisector (shown as a broken line) gives an estimate of L. Longevity and Mortality Rates of Fish in Nature 159 (1) as a means of representing mathematically the general growth pattern of fish in terms of two parameters to provide a simple means of relating size and growth to mortality and lifespan. It is a property of equation (1) that it can be transformed to a linear function relating length at age t to length at age ^ + 1, namely: /,^i = L„(l-e-^) + /,e-^ (2) Fig. 3 shows the growth curves of Fig. 2 plotted in this way. From equation (2) it will be seen that the slope of the line drawn through the points provides an estimate of e~^, and hence of K; and that the intersection of the line with the bisector drawn through the origin (shown by broken lines in Fig. 3) gives an estimate of the asymptotic length L^. Esti- mates of L^ and K for all the species under consideration are listed in Table I. Apart from providing a means of estimating the two para- meters of the growth equation (1), plotting Z^ against Z^^^ in this way is a valuable technique for the comparative analysis of growth curves (Walford, 1946). For example, it can be seen from Fig. 3 that male plaice not only have a lower L^ than do females, but also grow towards it rather more quickly, i.e. they have a higher K. In the case of Lahidesihes sicculus (insert in lower part of Fig. 3) the lengths are at monthly instead of yearly intervals, but when plotted one against the next they nevertheless give a close approximation to a straight line; in this case, however, the slope is e"^^^, and so in reality is very much flatter than the other graphs of Fig. 3. The method is also useful for detecting departures from the simple growth pattern which sometimes arise because of special en- vironmental conditions, of which lack of uniformity in the supply of food to fish of difiPerent sizes is usually the most important (see below and also papers by Aim, 1946, and Decider, 1951). 160 R. J. H. Beverton and S. J. Holt Interspecific relations between maximum age (Tj^ax)* mortality rate (M), asymptotic length (L^) and growth rate (K) Table I lists, for each species, values of a pair of parameters defining lifespan and the force of mortality {T^^^ and M) and a pair defining the asymptotic size of the organism and the rate at which that size is attained {L^ and K). The para- meters T^iax ^^d ^ ^r^' of course, closely linked on purely 50 lOO mathematical grounds ; there is no a priori reason why L^and_ K should be, but it appears from the data that they ar'e fairly, closely correlated (inversely), although there are some im- portant exceptions. In this paper we therefore consider only two of the possible relationships, that between T^^ax ^^^ -^oo concerning the extremes of age and size, and that bfijtw£^n M and K which, in effect, refer to the course of events within the lifespan. Other possible relationships which might give a better interpretation of the available data are under investiga- tion. Longevity and Mortality Rates of Fish in Nature 161 30 25 20- 100 I50 L^(cm) 200 • • SALMONOIDEI / true salmon ISO 200 250 Fig. 4. Relation between maximum age (Tmax) and asymptotic length (Loo) in the Clupeoidei, Gadiformes, Pleuronectoidei and Salmonoidei (from Table I). Because the correlation between Tmax and M, and Loo and K, is inverse the species appearing in the bottom left-hand corner of Figs. 4 and 5 tend to occur in the top right-hand corner of Figs. 6 and 7 ; this point should be borne in mind when comparing the two sets of diagrams. AGEING — V — 6 162 R. J. H. Beverton and S. J. Holt In four of the main groups of fish there is a sufficient range of values to enable each group to be examined separately. These are the herring and related species (Clupeoidei), the cod family (Gadiformes), the salmonoids (Salmonoidei) and the flatfishes (Pleuronectoidei). Fig. 4 shows the relations between L^ and T^^^ax i^ these groups. In each case there is a well- defined trend, especially so in the Clupeoidei which are per- haps a more homogeneous group than the others. The lines drawn through the points have no statistical significance, since the precise accuracy of the individual points is largely unknown, and in some cases the values recorded are certainly over- or under-estimates of the true values ; this is particularly so with the parameters Tj^^ax ^^^ ^'> which are more difficult to determine accurately than are the growth parameters L^ and K, and the lines have been drawn with these considera- tions in mind. Despite these qualifications, it does seem that the line for the Clupeoidei differs from that for the other groups, the larger members of the herring family appearing to live to a greater age than do fish of the other groups of a comparable size, the contrast being most noticeable with the gadoids. The scatter of the points is most marked in the salmonoids, which may be a reflection of the heterogeneity of this group and of the varied environments in which members of it are found, since they include marine, freshwater and anadromous species. The true salmon, ringed by a broken line, fall outside even the considerable variation of the rest of the salmonoids, since for their size they have a very short lifespan indeed. The pleuro- nectoids form a compact group, with a closely linear relation between L^ and T^^^^ with the exception of the halibut ; the maximum age recorded for this species (30 years) may, how- ever, be somewhat below the real maximum owing to diffi- culties of determining the true age of the oldest fish, and the fact that there was some fishing on the populations in question. The lines drawn for the four groups shown in Fig. 4 have been reproduced in Fig. 5, together with the data for all other Longevity and Mortality Rates of Fish in Nature 163 species. Most of these fall somewhere near the lines for one or other of the first four groups, with the sturgeon in the top right-hand section of the diagram having the highest values of both Tjj^ax ^^d ^ooj ^^^ ^ cluster of the small and short- lived species near the origin (see enlarged panel). The only 50 - 40 O' '-^ ' ' ' ' /SALMONOIDEI O lO 20 30 40 bO / PLtURONECTOIDEI 3CO Fig. 5. Relation between maximum age (Tmax) and asymptotic length (Loo) in various species not included in Fig. 4 (from Table I). The lines are those for the four groups shown in Fig. 4. species which, from the data we have examined so far, appear to be exceptional are the Thunniformes — with their large size and relatively short life they occupy a position similar to that of the true salmon — and possibly the Rays (e.g. Dasyatis akajei; Yokota, 1951), but age determination is difficult in the 164 R. J. H. Beverton and S. J. Holt cartilaginous fish and it may well be that the values of T^^^ recorded for this species in Table I are too low (see legend to Table I). Figs. 6 and 7 show the relations between M and K in the Fig. 6. Relation between natural mortality coefficient (M) and rate of curva- ture of growth curve (K) in the Clupeoidei, Gadiformes, Pleuronectoidei and Salmonoidei (from Table I). same way as do Figs. 4 and 5 for Tj^^x ^^^ ^oo- Again there is a fairly definite trend within most groups, although the scatter is rather greater than before and a trend in the case of the pleuronectoids is hardly detectable. Part, at least, of this greater variation is due to inaccuracies or uncertainties in the Longevity and Mortality Rates of Fish in Nature 165 values of ikf , and in several instances it has been necessary to draw an arrow indicating the direction in which the true value of M is thought to lie. However, when the remaining values are superimposed on those for the four main groups (Fig. 7), a rather more ordered picture is produced than that of Fig. 5. Of the previously aberrant species, the Thunniformes now fall 2-5 20 SALMONOIDEI PLEURONECTOIDEI 175 Fig. 7. Relation between natural mortality coefficient (M) and rate of curva- ture of growth curve {K) in various species not included in Fig. 6 (from Table I). The lines are those for the four groups shown in Fig. 6. into line, because although they grow to a relatively large size they approach it rapidly, i.e. they have a high K, The same is true for the Atlantic salmon (Salmo solar \ Nail, 1927) since this also has a relatively high value of ^; so indeed have the Pacific salmons {Oncorhynchus spp.), although the trend of mortality with age is so abruptly discontinuous in these species that a single value of M cannot be assigned to all of them. 166 R. J. H. Beverton and S. J. Holt Because both T^^^ and M, and L^ and K, are themselves fairly closely correlated (inversely), it is not unexpected that if a relationship is found between Tj^ax ^^^ -^oo ^^^ should also appear between M and K. What is perhaps significant is that the clearly established exceptions to the general positive correlation between longevity and size (which undoubtedly emerges from the above analysis) are those species which have a much higher value of K than would be expected from their L^. This suggests that the "rate of completion of the growth pattern" is more closely related to longevity than is size and some further evidence on this point is discussed below. Be that as it may, we have not yet come across a slow-growing species which dies from natural causes when only a small fraction of its growth pattern has been covered, or the con- verse— a species whose natural lifespan extends far beyond the point at which the limiting size is reached (as in man); and it seems that for a wide range of fish species the natural span of life is nicely adjusted to the time needed to complete, or nearly to complete, the growth pattern. Some intra -specific comparisons We have so far looked at the relations between growth and mortality in different species, but intra-specific comparisons also throw some light on this question. If, firstly, we consider differences between sexes we notice that, when the growth rates are clearly different, L^ for males is usually less than for females in the same population {Dasyatis akajei, Gadus minutus, Ganibusia, Coitus, Isopsetta, Pleuronectes). In these cases K for males is greater than for females, and the male mortality rate is higher. When the growth rates do not differ, or differ only slightly, the natural mortality rates are also similar (Leucichthys artedi, L. kiyi, Salvelinus, Mallotus, Leuresthes, Citharichthys). The chum salmon (Oncorhynchus keta) of the Columbia River is exceptional : the male natural mortality is greater than that of the female, and the male has a higher L^. In Callionymus, also, the males approach a Longevity and Mortality Rates of Fish in Nature 167 greater size; they have a lower K than the females, but a higher mortality rate. In^loAce {Pleuronectes platessa) of the /f North Sea, and perhaps in other species, the sexual difference in mortality rates is not simple; thus in both sexes the rnortalr ity rate tends to vary with the age of fish, but whereas in i "rnales it increases with increasing age — at least from the fifth To about the fourteenth year — in the females the evidence is that it may even decrease. A species of mosquito fish, Gambusia holbrooki, gives evidence that males are more susceptible than females to adverse conditions of temperature, oxygen, ion concentration and presence of cyanide. The females also overwinter more successfully and are less severely affected by catastrophies due to unknown causes (Geiser, 1924). It would be interesting to know whether similar sexual differences have been noticed in the many tests which have been made with several fish species of the toxicity of various substances, especially those found in polluted water; we have not, however, found information of this kind in the published reports of such experiments. Analysis of growth rates within populations of the same or closely related species living in different areas suggests that two factors account for most of the variation found : food and temperature. The asymptotic size is greatly modified by the supply of food available, but this does not affect the para- meter K. Differences in environmental temperature, however, / / aftect both K and L^; thus w^th an increase in water tempera- ture K increases roughly proportionally with the logarithm of temperature and L^ decreases, but to a lesser extent (see Taylor, 1958; Holt, 1959a). This temperature relation at least partly explains the statement often repeated in fisheries literature that in warmer waters the fish tend to be smaller than in cooler waters but, equally, that they grow faster in the former (see e.g. d'Ancona, 1937; Gunter, 1950); the size distribution that is actually observed at any time depends, however, on the mortality rate as well as the growth pattern. There are rather few data 168 R. J. H. Beverton and S. J. Holt which can be used to examine this question, but those we have seen suggest that in this case also a high value of K is associated with both a low L^ and a high mortality. This can be seen for Gadus minutus in Table I, and there is other scattered — but usually incomplete — evidence pointing in the same direction. Thus the grayling {Thymallus signifer) has a higher K and lower L^ in Michigan lakes (warmer) than in the Great Bear Lake (colder) and it apparently lives about twice as long in the latter locality as in the former (Brown, 1943; Miller, 1946). It is said that in France, where it grows fast, the stickleback (Gasterosteus aculeatus) lives only 14-18 months, whereas in northern Europe it lives much longer, and indeed does not mature until it is several years old (Bertin, 1925) ; according to Flower (1935), sardines (Sardinapilchardus) grow more slowly and live longer in the English Channel than in the south of the Bay of Biscay; and so on. Jenkins, Elkin and Finnell (1955) studied the growth of six species of sunfish (Lepomis spp. and Chaenohryttus) in over one hundred water bodies in Oklahoma and noted for each species that the oldest individuals were always in the populations having the slowest growth rates. We have to be careful in interpreting data of this kind, however, because a general observation that the maximum age attained is lowest in areas where growth is fastest may sometimes be due to effects of fishing coupled with a density-dependent growth rate, the fishing causing a reduced survival and population density and so permitting a better supply of food per fish with a consequent increase in the growth rate (see, for example Fry, 1936, for populations of Hesperoleiicus venustus in Calif ornian streams). It is interesting to note that the same associations we have recorded above between growth and longevity in related species, or even in populations of the same species which have become established as independent units in different water basins, do not necessarily hold when growth is modified experimentally. There is not much information on this, but the studies of Aim (1946) on perch populations with stunted Longevity and Mortality Rates of Fish in Nature 169 growth did not indicate any marked difference in longevity compared with those in which growth was normal. The ex- perimental studies being carried out by Comfort (personal communication) on growth and longevity in guppies {Lebistes reticulatus) appear so far to be giving the same result, although severe underfeeding during the early life of trout kept in tanks has been shown to delay maturity and actually prolong life (McCay, Dilley and Crowell, 1929). Again, the association between high values of M and of K noted above may not hold for comparisons between populations of the same species in closely adjacent waters, as in the case of the bullhead {Coitus gobio; Smyly, 1957) in Lake Windermere and the River Brathay (see Table I). Mortality, growth and metabolic rate To understand the relations tentatively identified above, it is necessary to extend our studies to include comparative physiology and behaviour, and at this stage w^e can do little more than indicate the lines of comparison that might profit- ably be pursued. One of these follows from the fact that the growth parameter K is predictable from the rate of endogen- ous nitrogen excretion by a starved animal (von Bertalanffy, 1938), and it would be expected that this is also closely related to metabolic rate and to activity, as Edmonds (1957) has shown in a comparative study of some invertebrate groups. For fish, the available data seem to confirm the relation of K to metabolic rate. Thus the oxygen consumption of Gadus callarias at 7-11° is 0-33-0-35 O2 ml./g.^^^/hr. and that of Gadus virens at the same temperature is similar, 0-36-0 -47 O2 ml./g.2/3yhr. (Sundnes, 1957). These two species have the same K values (0-2) though the natural mortality of G. callarias is possibly rather higher than that of G. virens. On the other hand, Leucichthys sardinella has, at about the same temperature (7-9-4°), an oxygen consumption of 0-55- 0-75 O2 ml./g.2/3yhr., corresponding with a higher K value (0-4) and much higher M (0-6). The cyprinids Labeo rohita, 170 R. J. H. Beverton and S. J. Holt Catla catla, and Carassius carassius, all of which have rather low values of K, have a low oxygen consumption of about 0-2 ml./g.^^^/hr. (Blazka, 1958). Over the temperature range 5-35°, oxygen uptake by another cyprinid, the goldfish, Carassius auratus, ranges from 0-05-0-46 ml./g.^^^/hr. (Fry and Hart, 1948); this species has a K value of about 0-3. Metabolic rate has been estimated, in connexion with ex- periments on the transport of live fish in closed containers, from the rate of carbon dioxide accumulation; in one such case Tilapia mossambica, which has a higher K value than Cyprinus carpio, respired faster than the latter, though kept at the same temperature (Vaas, 1952). Further evidence of relative metabolic rates comes from studies of the rate of uptake and loss of radioactive substances by fishes. Thus, in a review of this subject. Boroughs, Chipman and Rice (1957) quote results indicating that the exponential loss coefficient of orally administered strontium 89 from the body of Tilapia mossambica is two and a half times that of skipjack {Euthynnus yaito), yellow-fin {Neothunnus macropterus) and "dolphin" (Coryphaena hippurus). These latter fishes are more active species than Tilapia and, from the scanty data available, appear to have higher K values. We have not found any published data to indicate whether, in fish for which K is higher in males than in females, the respiratory rate of males is also higher, as might be expected. Natural death and reproduction The last line of evidence we shall mention is that concerning the connexion, in fish, between natural death and reproduc- tion. We have previously mentioned that in the short-lived species where there is an abrupt end to the lifespan, death usually occurs at or soon after spawning. What has been called "reproductive drain" may also become apparent in other ways. For example, the ratio of the weight of a fish to the cube of its length (called the "condition factor" or "ponderal index" in fisheries literature) varies seasonally, Longevity and Mortality Rates of Fish in Nature 171 being highest just before spawning. In the plaice it is apparent that this variation is very much greater in older than in younger individuals, and it seems that as the fish gets older (or, perhaps, merely bigger) the strain of meeting the repro- ductive demand increases to a point at which recovery is not possible. It seems that this kind of effect is most evident in species with high K, which mature at an early age but at a size which is rather large in relation to the asymptotic length, L^. Fish, such as sturgeon, with a low K, which mature when relatively rather small, do not show a decline in their repro- ductive capacities; Gamhusia affinis, on the other hand, exhibits an absolute decrease in brood size w^ith increasing age (Krumholz, 1948) and indeed this species seems to have a true post-reproductive phase, which is rather rare in fish. Further evidence on the decline in reproductive powers with age in fish is presented in the contribution by Gerking to this colloquium. The complexity of the relations between the growth and reproduction parameters and mortality rates is illustrated by Svardson's (1943) review of data for the guppy, Lehistes reticulatus. Male guppies mature before females, and die younger. Under experimental conditions of low food supply they grow slowly to a low asymptotic size, and mature late; with a medium food supply the final size is greater and maturity somewhat earlier, but with an abundant food supply the final body size is again lowered although the onset of maturity is still further accelerated. Other observations concerning the relations between growth, reproduction and death have to be fitted into the picture outlined in this paper, but pending a detailed survey of the known facts, can only briefly be mentioned here. It has been thought for many years that the onset of sexual maturity in fish is a function of their size rather than of their age but like most such generalizations this is only partly true since, within a species population, individuals reach maturity over a considerable range of both age and size. There is, in 172 R. J. H. Beverton and S. J. Holt fact, a considerable amount of data on the size at which fish first reach maturity, and we have investigated whether the average length at which maturation occurs (L„J in relation to the asymptotic length (L^) has any bearing on longevity. Thus, in the last column of Table I are given the ratios LJL^ and in Fig. 8 these are plotted against maximum age, T The points are very much scattered, although there is perhaps 00> max* Lm L CO lOO -o • • o • • ( 3 X X 07S oO o Xo ^ X • •• o» X °<^ o°*J o« 0 0*' o 08 0 °1 • • 0 0 -0 0 0 i i 0 § 0 0 0 1 0 0 0 1 1 10 20 30 — T max (yr ) — 40 50 Fig, 8. Size at maturity (L^) and longevity. Plot of ratio LmjLco against Tmax- • = Salmonoidei, x = Clupeoidei; other species shown as Q • just a hint that the shortest-lived species are those which have, on average, the highest value of LJL^, that is, mature at a relatively late stage in their growth cycle. This tendency is a little clearer within the salmonoid group (solid circles), but in the clupeoids (crosses), with one doubtful exception, the values of LJL^ are consistently high and have no trend at all. It may, of course, be that this way of relating maturity and longevity is too crude; maturation size might better be ex- pressed, for example, as a function of both K and L^, but this needs further examination. Longevity and Mortality Rates of Fish in Nature 173 The undoubted association between reproduction and death in many species of fish makes it difficult, merely on the kind of evidence we have considered in this paper, to detect whether true senescent processes play a part in determining longevity in fish, as they do, for example, in mammals. The mathematical representation of the characteristic growth pattern in fish which we have adopted here does, of course, imply that growth proceeds towards a finite limiting size, and so is not "indeterminate" in the sense used by Bidder (1932). When the growth of fish is plotted as in Fig. 3 the impression gained is not only that this is a valid interpretation but also that the growth pattern of the long-lived species (including plaice) does not differ qualitatively from that of the short- lived ones in which a limiting size is efPectively reached within the observed range of age. As Comfort (1956) has pointed out, however. Bidder's hypothesis of immortality in fish does not necessarily depend on whether there is a finite limit to their growth or not, and can equally well be maintained if that limit can be shown to be approached asymptotically — and hence reached only after an infinite span of time — as opposed to abruptly, with growth ceasing beyond a certain specific age. To test this latter alternative directly, at least in the slow- growing species, is as difficult as it is to prove whether or not there is a finite Umit to the lifespan of a species which has a constant mortality rate within the observed range of age. In this connexion, it is perhaps worth noting that although the growth equation we have adopted requires an asymptotic approach to a limiting size, high enough values of ^ (as are found, for example, for Labidesthes sicculus) can produce a theoretical growth curve in which the approach to L^ is so rapid that it would be indistinguishable in practice from an abrupt approach, especially when it is remembered that there is usually a seasonal periodicity of growth superimposed on the general pattern. Thus we are inclined to the view that further speculation along these lines is unlikely to contribute much to the solution 174 R. J. H. Beverton and S. J. Holt of the question as to whether ageing in fish differs funda- mentally from that in higher vertebrates. A more profitable approach would seem to lie in a better understanding of the intrinsic causes of natural death in fish, about which relatively little is yet known. A recent study of the European eel {Anguilla anguilla) by Tucker (1959) suggests that the debility of these fish at the inception of gonadal and other hormone activity, which causes them to drift passively downstream, is due to demineralization of the starving body. That this process is reversible, at least in the early stages, is shown by the fact that silver eels imprisoned in fresh water can survive by regression of the gonads and consequent remineralization of the body fluids ; and it is also known that recovery of the Atlantic salmon (Salmo solar) after spawning can be hastened by placing them in salt water. It is true that both these species, and more especially the eel, have a highly atypical life history, but this kind of explanation of certain behavioural patterns in physiological terms would appear to be an essen- tial step in the solution of at least some aspects of the problem of longevity in fish. The other line of investigation that would seem to be of special significance is a comparative study of the physiology of growth and reproduction in species which have a post-reproductive phase. We would hope that an under- standing of the beginning of the reproductive phase of the life history in relation to growth processes would help inter- pretation of those events at the end of the reproductive life- span that lead to death. REFERENCES Aasen, O. (1952). Fiskeridir. Skr. Havundersok., 10, (2). Akyuz, E. (1959). Unpublished data, filed at FAO, Rome, and Et ve Balik Kurumu, Istanbul. Alm, G. (1946). Rep. Inst. Freshw. Res. Drottning., 25. Alm, G. (1952). Rep. Inst. Freshw. Res. Drottning., 33, 17. d'ANCONA, U. (1937). Rapp. Comm. int. Mer Medit., 10, 162. Anonymous (1954). Comm. Fish. Rev., 16, 28. Appelget, J., and Smith, L. L. (1951). Trans. Amer. Fish. Soc., 80, 119. Arora, H. L. (1951). Calif. Fish Game, 37, 3. Longevity and Mortality Rates of Fish in Nature 175 Berdegue, J. (1955). Rev. Soc. mex. Hist, nat., 16, 45. Bertalaxffy, L. von (1938). Human Biol, 10, 181. Bertin, L. (1925). Ann. Inst. Oceanogr., 2, 1. Bevertox, R. J. H. Unpublished. Beverton, R. J. H., and Holt, S. J. (1957). Fish. Invest., Lond.y Ser. 2, 19. Bidder, G. P. (1932). Brit. med. J., 2, 583. Blackburn, M. (1950). Aust. J. Mar. Freshiv. Res., 1, 155. Blazka, p. (1958). Physiol. ZooL, 31, 117. Boroughs, H., Chipman, W. A., and Rice, T. R. (1957). U.S. Nat. Acad. Sci.JNat. Res. Counc., Publ. 551, 80-87. Brown, C. J. D. (1943). J. Wildlife Mgmt, 7, 353. Brown, M. E. (Editor). (1957). The Physiology of Fishes. New York: Academic Press. Burd, a. C. Unpublished. Chang, Hsiao-Wei. (1951). J. Mar. hiol. Ass. U.K., 30, 281. Clark, F. N. (1925). Fish Bull., Sacramento, No. 10. Clark, F. N. (1940). Calif. Fish Game, 26, 39. Cohen, D. M. (1954). Staiif. ichthijol. Bull, 4, 167. Comfort, A. (1956). The Biology of Senescence. London: Routledge and Kegan Paul. Deason, H. J., and Hile, R. (1947). Trans. Amer. Fish. Soc, 74, 88. Deelder, C. L. (1951). Hydrobiologia, 3, 357. DE Sylva D. (1957). Bull Mar. Sci. Gulf Caribb., 7, 1. Devold, F. (1938). Fiskeridir. Skr. Havundersok, 5, (6). Dickie, L. M., and McCracken, F. D. (1955). J. Fish. Res. Bd Can., 12, 187. Dymond, J. R. (1933). Contr. Canad. Biol, 8, 1. Edmonds, S. J. (1957). Aust. J. Mar. Freshw. Res., 8, 131. Fitch, J. E. (1951). Fish. Bull., Sacramento, No. 83. Fitch, J. E. (1956). Calif. Fish Game, 42, 143. Flower, S. S. (1935). Proc. zool. Soc. Lond., 265. Foerster, R. E. (1929). Contr. Canad. Biol., 5, 1. Franca, P. da (1953). Bull. Soc. portug. Sci. nat., 4, 198. Frost, W. E. (1943). J. Anim. EcoL, 12, 139. Frost, W. E. (1945). J. Anim. EcoL, 14, 106. Frost, W. E., and Smyly, W. J. P. (1952). J. Anim. EcoL, 21, 62. Fry, D. H. (1936). Calif. Fish Game, 22, 64. Fry, F. E. J., and Hart, J. S. (1948). Biol. Bull., Wood's Hole, 94, 66. Geiser, S. W. (1924). Biol. Bull., Wood's Hole, 47, 175. Gerking, S. D. (1959). This volume, p. 181. Gottlieb, E. (1957). "On the dynamics of the population of coal-fish, Gadus virens L., in Norwegian waters." MS. Grainger, E. H. (1953). J. Fish. Res. Bd Can., 10, 326. GuNTER, G. (1950). Copeia, 298. Hanyu, I. (1956). Bull. Jap. Soc. sci. Fish., 21, 991. Hart, J. L. (1928). Publ. Ont. Fish. Res. Lab., No. 84. 176 R. J. H. Beverton and S. J. Holt Hart, J. L. (1931). Contr. Canad. Biol, 6, 427. Hart, J. L. (1948). Trans, roy. Soc. Can., 42, (Ser. Ill), 65. Herald, E. S., and Rakowicz, M. (1951). Aquarium J., 22, 234. HiLE, R. (1936). Bull. U.S. Bur. Fish., 48, 211. HiLE, R., and Deason, H. J. (1934). Trans. Amer. Fish. Soc, 64, 231. Holt, S. J. (1959a). J. Cons. int. Explor. Mer, 24, 374. Holt, S. J. (19596). Draft Report of International Training Centre on the Methodology and Techniques of Research on Mackerel (Rastrelliger). Rome: FAO, Mimeo. HuBBS, C. L. (1921). Ecology, 2, 262. Jenkins, R., Elkin, R., and Finnell, J. (1955). Oklahoma Fish res. Lab. Rep., 49. Jones, J. W., and Hynes, H. B. N. (1950). J. Anim. EcoL, 19, 59. Kennedy, W. A. (1943). Publ. Ont. Fish. Res. Lab. No. 62. Kennedy, W. A. (1954). J. Fish. Res. Bd Can., 11, 827. Krumholz, L. a. (1948). Ecol. Monogr., 18, 1. Lea, E. (1930). Rapp. Cons. Explor. Mer, 65, 100. McCay, C. M., Dilley, W. E., and Crowell, M. F. (1929). J. Nutr., 1, 233. Margetts, a. R. Unpublished. Marr, J. C. (1943). Stanf. ichthyol. Bull, 2, 157. Menon, M. D. (1950). J. Mar. biol. Ass. U.K., 29, 185. Miller, R. B. (1946). Copeia, Til. Moore, H. L. (1951). Fish. Bull., U.S., 52, 132. Nall, G. H. (1927). Fisheries, Scotland, Salmon Fish., II. NiGRELLi, R. F. (1959). This volume, p. 212. Paccagnella, B. (1948). Arch. Oceanogr. LimnoL, Venezia, 5, 141. Phillips, J. B. (1948). Fish. Bull, Sacramento, No. 71, 5. Probst, R. T., and Cooper, E. L. (1954). Trans. Amer. Fish. Soc, 84, 207. Pycha, R. L. (1956). Calif. Fish. Game, 42, 23. Qasim, S. Z. (1957). Proc. zool. Soc Lond., 128, 161. Radhakrishnan, N. (1957). Indian J. Fish., 4, 254. Raitt, D. S. (1939). Rapp. Cons. Explor. Mer, 110, 65. RiCKER, W. E. (1949). Trans. Amer. Fish. Soc, 77, 114. SL RiCKER, W. E. (1958). Bull. Fish. Res. Bd Can., 119. ^ Robertson, J. A. (1938). Fish. Invest. Lond., Ser. 2, 16, (2). RoLLEFSEN, G. (1954). Rapp. Cons. Explor. Mer, 136, 40. RuNNSTROM, S. (1936). Fiskeridir. Skr. Havundersok., 5, (2). Sacher, G. (1959). This volume, p. 115. Sato, R. (1950). TohokuJ. agric Res., 1, 87. Shiraishi, Y. (1957). Bull. Freshw. Fish. Res. Lab., Tokyo, 7, 33. SiLLiMAN, R. P. (1943). Spec sci. Rep. U.S. Fish Wildl. Serv., No. 24. Smyly, W. J. P. (1957). Proc. zool. Soc Lond., 128, 431. Strawn, K. (1958). Copeia, 16. Sundnes, G. (1957). Fiskeridir. Skr. Havundersok., 11 (9). SvARDSON, G. (1943). Rep. Inst. Freshw. Res. Drottning., 21. Longevity and Mortality Rates of Fish in Nature 177 Taylor, C. C. (1958). J. Cons. int. Explor. Mer, 23, 366. Templeman, W. (1948). Bull. Newf. Govn. Lab., 17. Tester, A. L. (1937). J. biol. Bd Can., 3, 108. * Thompson, W. F., and Bell, F. H. (1934). Rep. int. Fish. Comm., 8. Thompson, W. F., and Herrington, W. C. (1930). Rep. int. Fish. Comm., 2. TiBBO, S. N. (1957). Bull. Fish. Res. Bd Can., Ill, 85. TiEWS, K. (1957). Ber. Dtsch. Komm. Meeresforsch., 14, 192. Tucker, D. W. (1959). Nature (Loud.), 183, 495. Vaas, K. F. (1952). Proc. Indo-Pacif. Fish. Coun., 3, 119. VivES, F., and Suau, P. (1956). Invest. Pesq., 5, 17. VNIRO (1949). Ministry of Industrial Fisheries, Moscow. Walford, L. a. (1946). Biol. Bull, Wood's Hole, 90, 141. WoHLSCHLAG, D. E. (1954a). Ecology, 35, 388. • WoHLSCHLAG, D. E. (19546). Stanf. ichthyol. Bull., 4, 189. YoKOTA, T. (1951). Bull. Jap. Soc. sci. Fish., 16, 188. DISCUSSION Rotblat: Have you also plotted 1/jK" against T^jax? IjKhas the dimen- sions of time and is proportional to the time it takes to grow to half size. This may also be a linear function of the span of life. Beverton: No, we have not yet done that; so far we have examined only the relationships between K and M and between Loo and T,^^. Rotblat: The combination of these two factors is of special interest because you relate, then, time to time. This also agrees with what Sacher said about low metabolic rate. BeveHon: In fish the natural mortality coefficient and the maxi- mum age are closely correlated, because most of the logarithmic survival curves tend to be rather straight. We do not have curves of nearly the same slope which suddenly dip to very diff'erent maximum ages. Danielli: These growth-rate limitations may be entirely due to differences in natural conditions. Have you any data from fish which have been exposed to toxic substances in the water ? Beverton : The only paper on exposure of fish to toxic substances that I can recall offhand is on Gambusia (Geiser, S. W. (1924). Biol. Bull. {Wood's Hole), 47, 175). There the investigation was to show that males not only died off more quickly than the females after reproduction but were also more susceptible to a range of toxic sub- stances. In other words, their balance with the environment seemed more precarious than that of the females, with respect to artificially induced hazards as well as natural ones such as reproduction. On the first point, growth is undoubtedly very flexible in fish. Nevertheless, 178 Discussion our impression is that looking over the whole range of growth data for fish it is possible to see fairly characteristic sorts of growth. However much food you give a small fish it never grows very much bigger than its characteristic size in nature. This is a pretty general statement, but despite the extent to which growth can be varied experimentally, there does seem to be an overall pattern in nature which is fairly consistent. Holt: This sexual difference in susceptibilities is found rather commonly in experiments on the effects of water pollution and in studies to improve the transport of live fish in closed containers ; the males are usually more sensitive. Unfortunately I have no good data with which to compare respiration rates of the sexes. Spawning plays a more important part as a factor associated with death in fish, than, for example, food supply. One can vary the growth rate tremendously by adjusting the food supply, without changing the mortality rate at all. But where death is caused through spawning or is associated with it and maturity, it seems that the males are more drastically affected than the females. Thus in one salmon species all the males and most of the females die after spawning, but some of the females return to the sea. Nigrelli: In aquarium management we constantly find, when we autopsy fish, that there are about three females to one male, which means that there is some sort of selection in the collection. Danielli : If the fish are dying as a result of spawning perhaps one can change the situation. Death following spawning in salmon is said to be due to some syndrome involving the pituitary. This, surely, could be modified by appropriate hormone treatment. Nigrelli: Are there any comparative figures on population or growth studies on salmon or trout in hatcheries and those under natural conditions ? Holt: I do not know. Nigrelli: I think that in large hatcheries there is a lower infant mortality rate than is found under natural conditions. Comfort: In these natural populations, is there always a tendency for the growth to be smooth in outline ? Or does the same effect occur in wild fish as I have produced by keeping them small arti- ficially and then increasing the food supply ? Beverton: Yes; perch is the most notable example. You get that in the wild, particularly where there is a marked change in feeding habits as the fish grow bigger. For example, perch up to about 15 cm. feed primarily on animal plankton such as freshwater shrimps, etc. Above that size they change to exclusively carnivorous habits. If the pond or lake has no supply of small fishjthey^ just stop growing. Discussion 179 But the data also show that this stunted growth does not seem to be Associated with any marked increase in mortality rate. ~ '■ ■^ Comfort: I think that is what I am going to find. Another point is that H. J. Van Cleave (1934; 1935. Ecology, 15, 17; 16, 101) suggested that the apparent senescence of some mol- luscan species which are more or less indeterminate in growth is actually a size effect, because there is no accumulation of animals in the older age groups. In one of the freshwater limpets the question arose whether, when they got beyond a certain size, their holding-on mechanisms became inefficient or whether they were taken selectively by predators (Hunter, W. R. (1953). Proc. zool. Soc. Lond., 123, 623). Have you any instance of an adverse size effect in fish ? Beverton: I should have thought it usually worked the other way in fish. For instance, they escape predation as they get bigger, rather than the reverse, I would say. Comfort: What about catching them ? Beverton: That depends on the gear. Usually fishermen are after the bigger fish and take steps to catch them, but not always. Holt: It is usually in the smaller fish, having rather high K values, that the effect of reproduction on lifespan seems to be greatest, as they die off quickly after reproduction. In fish like sturgeons which grow to large sizes rather slowly there is no noticeable effect of re- production on their mortality, even though they mature at a re- latively small size. In middle-sized fish, such as the plaice, there is a certain effect but not such a drastic one. Thus in large spawning plaice there is a relatively great seasonal variation in the relation of length to weight (what we call the "condition factor"), suggesting that attainment of spawning condition becomes an increasing strain on the fish as they grow older. Rockstein: Do land-locked salmon reproduce year after year ? Holt: Some land-locked populations do. Beverton: There are the "residual" and the "Ko Kanee" salmon {Oncorhynchus nerka). Both are non-migratory, but whereas the "residuals" are the progeny of anadromous parents, the "Ko Kanee" is a self-maintaining stock which has no connexion with either of the other two. Gerking: There are some land-locked Atlantic salmon that repro- duce year after year. Rockstein: In a large reservoir in New York State we have brown trout, also called salmon trout, and these can be caught in all sizes depending on how successful the first year stock is in eluding the angler. It appears from their annual movements into the lake in spring and out again in the fall that they are spawning each year. 180 Discussion something like the salmon. If they do spawn each year, however, then spawning may have no appreciable effect on their longevity, as it is said to have in the case of the salmon. Gerking: There has been a very good study on those lines in Scotland, but it does not concern the question of ageing. It is a migration pattern. PHYSIOLOGICAL CHANGES ACCOMPANYING AGEING IN FISHES* Shelby D. Gerking Department of Zoology, Indiana University ^ Bloomiyigton, Indiana This paper is a review of efforts that have been made to show the relation between age and two important Ufe func- tions of fishes, nutrition and reproduction. Only a few studies have been made on the effects of age on food conversion, but feeding experiments demonstrate clearly that the ability to convert protein to body flesh declines as size and age increase. Other vertebrates also conform to this pattern, although they achieve a specific size relatively early in life in contrast to the prolonged period of growth in fishes. The rate of decline in the ability to utilize food for growth seems to be a matter of degree, rather than a basic difference between animals of specific and non-specific size. Studies of the effect of age on reproduction in fishes have produced no general conclusions. Nevertheless, the subject deserves attention because changes in reproductive capacity with age are commonly used as a criterion of senescence in other vertebrates. The reproductive capacity of live-bearing fishes of the family Poeciliidae declines with age and there may be a period of sterility before death. Neither of these two facts apply unequivocally to egg-laying fishes, however. There is no period of reproductive senility, and it is an open question whether or not age has an effect on reproductive capacity. Individual variation in fecundity is so great that age effects cannot be detected by refined statistical techniques. A more * Contribution No. 668 from the Department of Zoology, Indiana Univer- sity, Bloomington, Indiana. 181 182 Shelby D. Gerking subtle change in the ovary has been discovered that has not been appreciated before, however. Egg number does not increase in proportion to ovary weight. Either eggs become larger and fewer or the relative amount of connective tissue increases as the ovary grows. No critical evidence is available to support either contention. If the latter is true, ageing changes in the gonads of fishes would be similar to those in higher vertebrates. The lack of conclusive proof of the effect of age on reproduction is disturbing because it has not been possible to accept or refute without question that portion of Bidder's argument (1925a) that fish are immortal because old individuals show no decline in reproductive capacity. Knowledge about fishes has not progressed to the point w^here the relative importance of the influence of rate of growth and chronological age on physiological function can be distinguished. This constitutes a great w^eakness in the analysis of differences in nutrition and reproduction related to size and age. Svardson (1951) has expressed the opinion that age of fishes should be expressed as "physiological age" based on nutrition and rate of growth. Larkin, Terpenning and Parker (1957) have demonstrated that growth of rainbow trout {Salmo gairdneri) in different British Columbia lakes can be dealt with more effectively by comparing growth rates of fish of the same size rather than of the same age. This point of view reflects a tendency to depart from traditional age and growth studies because of dissatisfaction wdth chronological age as an adequate unit on which to base physiological change. Critical experiments are clearly needed to separate the effects of rate of growth from those of chronological age. In view of the lack of information it has been necessary to refer in the ensuing discussion to age and size indiscriminately without distinguishing which of the tw^o factors is the more important. Efficiency of Protein Utilization for Growth A series of studies on the protein metabolism of sunfish, family Centrarchidae, has been done in order to learn whether Physiological Changes with Age in Fish 183 or not size and age have an effect on the abihty to convert protein to body flesh (Gerking, 1952, 1954, 1955a, b). Longear sunfish (Lepomis megalotis), green sunfish {Lepomis cyanelhis), and bluegill (Lepomis macrochirus) gave similar results. The methods used in these experiments were essentially the same as those used to study food conversion in other animals (May- nard, 1951). A weighed quantity of food was fed each day to a group of fish of various sizes kept in separate aquaria at temperatures of about 25° for a period of 30 to 50 days. They were fed at maximum or near-maximum rates. At the end of the period the fish were killed, weighed, and analysed for protein. At the beginning of the period their weight was known, and their protein composition was estimated by averaging protein determinations by a micro-Kjeldahl method on several fish collected at the same time and place as the ones used in the experiments. The food was mealworms, Tenehrio molitor larvae, which had also been analysed for their protein content. Thus the efficiency of protein utilization for growth could be determined for fish of various sizes by comparing the amount retained with the amount absorbed by the gut. Absorption of protein was measured by subtracting the amount of nitrogen in the faeces from that consumed. Ab- sorption was practically complete in every fish. Menzel (1957) has duplicated these experiments on two Bermuda reef fishes, angelfish (Holocanthus hermudensis) and red hind (Epinephalus guttatus). The former species is a herbivore and the latter a carnivore. They were fed as much as they would eat at three temperatures, 19, 23, and 28°. The angelfish were fed algae (Enteromorpha salina and Monostroma oxysperma) which were kept in the tanks with them, and the red hinds wxre fed three species of small fish, Harengula callolepis, Sardinella anchovia, and Anchoa choerostoma. Efficiency of protein utilization for growth was determined over a 21 -day period in the manner described above. The weight of Menzel's fish varied from 50 to 763 g. and the sunfish from 7 to 184 g. Protein accumulation was used as an 184 Shelby D. Gerking index of the efficiency of the growth process because protein synthesis is the most characteristic feature of growth in animals. The fat content of fishes varies considerably from one individual to another and during the seasons of the year. Such large variations in fat complicate precise growth measure- ments based on body weight, dry weight, or caloric content. LONGEAR SUNFISH olO-DAY EXPTS. • 50-OAY EXPTS. BLUEGfLL SUNFISH DRY WEIGHT IN GRAMS Fig. 1. Relationship between efficiency of protein utilization and dry weight of four species of fish. Sources of information given in text. With the exception of the angelfish, the other four species fit a common pattern. The efficiency of protein utilization for growth decreases as the fish increase in size (Fig. 1), indicating that either size, age, or both affect the ability of the fish to utilize their food for the synthesis of new protoplasm. The angelfish is a special case because of its herbivorous feeding habits. They were unable to grow on a diet of algae and generally had a negative nitrogen balance. Menzel concluded that angelfish cannot grow on a diet of plants alone unless they Physiological Changes with Age in Fish 185 eat extraordinary quantities or algae with a much higher protein content than he used. Probably the "herbivorous" fish depends to a large extent on the invertebrates living in association with plants for the protein required for growth. The results of the experiments on angelfish are. therefore, not comparable with the others. The experimental results on the four carnivorous species are similar. In every case the efficiency of protein utilization decreased as the size of the fish increased. Efficiency was very high in the smallest fishes. For example, green sunfish (body weight = 7'1 g.) were 39*7 per cent efficient in using protein for growth; longear sunfish (9-1 g.) were 33-3 per cent efficient; bluegills (7-7 g.) 38-0 per cent; and the red hind (about 227 g.) 32-1 per cent. In contrast, utilization among the largest specimens was 20-0 per cent (48-5 g.), 4-7 per cent (103-3 g.), 23 • 6 per cent (184-0 g.), and 22 • 7 per cent (612 g.), respectively, for the four species. The value for the largest longear sunfish departs considerably from the others. This specimen was as large as any ever observed in the creek where it was captured near Bloomington, Indiana, and it was probably over six years old. The other three species are not represented by individuals of maximum size or age. This may indicate that extremely large individuals have a very low efficiency of protein utilization for growth. More evidence is required to establish this point, however. The shape of the graphs is not consistent. A linear relation- ship expresses the relation between protein efficiency and dry weight in the longear sunfish, but it is curvilinear in the other species. The inconsistency is due to individual variation and the difficulty in establishing the initial protein content of the experimental fish from analyses of sample fish. Menzel answered an important question with respect to the effect of temperature on protein utilization. The red hind ate only slightly more at 23° than at 19° but ate about twice as much at 28°. This great difference in feeding rate did not alter the efficiencies of protein utilization, however. Thus all 186 Shelby D. Gerking protein utilization experiments in this 9° temperature range are comparable. A large amount of work has been done on food conversion by fishes, but most of it relates weight gain to the amount of food consumed. Large variations have been encountered and are due to variable fat deposition, unknown organic composi- tion of the food, differences in the size of experimental fish, and other factors. Most of the research has been done on a variety of foods fed to young trout in order to obtain maxi- mum growth rates in hatcheries. This material is not applic- able to the present discussion since the fish used in the experi- ments were nearly uniform in size and age. Ivlev (1939a, h, c), Karzinkin (1939) and Morgulis (1919) have studied fish nutri- tion by detailed analyses of the organic constituents of the fish and their food, but none of these workers compared the efficiency of food utilization by fish of different sizes. Although the effect of age and size on protein utilization for growth cannot be separated, it can be stated definitely that larger fish are less efficient in this respect than smaller ones. Chronological age may play some part in this phenomenon. It is universally true that the rate of growth declines as age and size increase. In this respect fishes conform to the com- mon vertebrate pattern. The protein metabolism studies demonstrate that this decline is due to a decreasing ability of the fish to utilize its food for growth as it grows larger and older. The growth of most vertebrates stops relatively early in life while that of fishes is prolonged. Although this dif- ference cannot be explained at the present time, the loss of growth efficiency with age is clearly exhibited by both. Fecundity Live -bearing fishes There are several groups of fish which give birth to well- developed young, but reproduction in relation to age has been studied in only one family, the Poeciliidae. In this family the Physiological Changes with Age in Fish 187 male is much smaller than the female. His growth stops soon after sexual maturity is reached, but the female continues to grow until death. Fertilization is internal and the sperm transfer is accomplished by a greatly modified anal fin, the gonopodium, which acts as the male copulatory organ. The sperm penetrate the ovarian wall and fertilize the eggs while they are within the ovarian follicle. Development proceeds, and the young are born in various stages of development according to the species. Reproduction and senescence have been studied only in the western mosquitofish, Gambusia a. affinis. Krumholz (1948) has provided an unusually good series of observations on the number of young in successive broods of females from the onset of sexual maturity until death. Females were collected from ponds in southern Michigan and in the vicinity of Chicago during the summers of 1939 to 1944, and the number of embryos in the ovary was counted. The number of young produced by a single female depends on: (1) the number of broods liberated during a season, (2) the length of the mother fish, (3) the time of season at which the individual broods are cast, and (4) the locality in which the mother fish lives. The first two factors are pertinent to the present discussion. The age and size at first maturity are correlated with the time of year when the female offspring are born and their rate of growth. Faster-growing individuals generally reach maturity at an earlier date than slower-growing individuals. Mosquitofish born late in the summer do not reach maturity until the following spring while those born early in the summer may reproduce within a month or two. The number of young in a brood increase^ as the length of the female increases, but at large sizes the rate of increase is drastically reduced. The rate of increase in fecundity in relation to length of the female is greater in the second brood than in the first or subsequent ones. In one case the number of embryos in the first brood increased as the 1 • 3 power of the length ; the second brood as the 1 • 4 power, and the third brood as the 0 • 8 power. 188 Shelby D. Gerking Numbers of embryos in the fourth brood were so variable that no relationship with length of the female could be demon- strated. The influence of age on fecundity was proved by both field and laboratory observations. Females of similar size col- lected at an earlier date in the summer contained a greater Table I The effect of age on fecundity of female mosquitofish {Gambusia affinis) of the same size. Modified from Krumholz (1948) Size of female Date of collection No. of females Average no. of in mm. embryos Argonne Woods Pond 46-55 June 9 63 210-4 July 13 194 152-9 38-44 July 13 179 153-3 Aug. 14 3 42-7 Sanitary — District Lake 35-44 June 9 67 30-7 July 13 236 28-4 50-55 July 13 228 28-8 Aug. 14 449 8-8 Parr's Lake 27-45 July 31 330 47-4 Aug. 14 324 34-3 35-46 Aug. 14 167 49-9 Sept. 19 25 17-0 number of embryos than those collected later (Table I). The decrease in embryo production was greatest during late summer, near the end of life. Among fish born at approxi- mately the same time there was a decrease in fecundity with age despite an increase in size. One group of females averaged 49 mm. in length and yielded a mean of 205-4 embryos on June 9, 1939; on July 13 the females were 52-6 mm. long and gave 155 • 7 embryos; and a few remaining fish of the same age Physiological Changes with Age in Fish 189 group measured 52-3 mm. on August 14 and contained 42-7 embryos. Under laboratory conditions four fish gave birth to a decreasing number of young. They averaged 30 • 8 young in the first two broods, 10-0 in the third, and 7-0 in the fourth. There may be a period of reproductive sterihty in Gamhusia late in life. In one collection the largest female was not gravid while other sexually mature individuals were. The ovaries of this female were examined microscopically, but no ova were found. The only other pertinent observation in this connexion was made by Fraser and Renton (1940) on Heterandria formosa, the dwarf top-minnow. Successive broods were followed in a single female from April 1934 until death in May 1936. She grew to a remarkably large size (40 mm.), and produced a total of 170 young, 150 of which were born during the first eight months. She showed diminishing fertility during the last few months of life. At the time of death the ovary was examined microscopically and found to be a "mass of de- generating tissue". Egg-laying fishes Bidder (1925a, 19256, and 1932) raised an extremely provo- cative question when he proposed that fish and certain other aquatic animals are potentially immortal. He based his argument on reproduction and growth in the plaice (Pleuro- nectes platessa). He noted that these fish continue to grow throughout their lives and that the oldest individuals retain their reproductive capacity. His idea was immediately challenged by Wallace (1925) who pointed out that male plaice had a higher mortality rate than females and that "this apparently implies natural death". Bidder answered by making a distinction between "senile death" and "parental death". Parental death refers to that which occurs as a result of the reproductive act. He supported this definition by draw- ing upon the same plaice data which Wallace had used to show that six-sevenths of the males die after the first spawning 190 Shelby D. Gerking Senile death in Bidder's opinion should be reserved for those animals which exhibit "negative growth" after full sexual maturity and specific growth have been achieved. In his last paper Bidder took his stand on the basis of correlating ageing with specific growth and ceased to make an issue of reproductive performance. The issue of reproduction was soon raised again by Orton (1929) in a somewhat different manner. He asked the question whether or not fishes might die as a direct or indirect result of expending themselves in reproduction. Russell's (1914) data were cited, which suggested that the reproductive organs of large haddock (Melanogr animus aeglefinus) make up a larger proportion of the body weight than they do in smaller speci- mens. Metabolism is concentrated overwhelmingly on repro- duction, and although the expenditure of energy may not kill, the animal may become so unstable that otherwise sublethal factors might be brought into play and cause death. Orton dramatized this effect by terming it "over-reproduction". Orton's viewpoint turns up in another connexion. Svardson (1949) considered the effect of natural selection on the egg number of fishes and concluded that there must be opposing selection pressures for decrease as well as an increase in egg number. Clearly, a mutation causing an increase in egg number would have a selective advantage and spread through a population unless there were factors opposing such a change. In his words: "There would' be an anatomical and physio- logical limit for the females' capacity of producing more eggs. When the egg number has been brought up to this limit, only those individuals not exceeding the limit could spawn, while the others died." Svardson later rejects this as a major factor in evolution and concludes that egg number would be limited by the ability of fish to protect the young among those fishes which behave in this way. More importantly, he thinks, egg number is limited by the premium on large eggs which produce large fry. He postulates that the large fry are in a better competitive position in the population. Physiological Changes with Age in Fish 191 Bidder's theory that fish are potentially immortal has never attracted much attention among fishery workers. No doubt its lack of popularity is due to the fact that those who work with fish know that they die and that the lifespan of most species is short compared with our own. Among the host of workers who have aged fish by inspecting the growth rings on the scales or other hard parts, none has observed a specimen which spanned the centuries. Direct observation by aquarium curators throughout the world has led to the same conclusion (S. Hinton, unpubHshed). Only 21 of 328 species reported lived longer than 20 years. The oldest fish was a sturgeon (Acipenser ruihenus) in the Royal Zoological Society Aquarium in Amsterdam, Holland, which lived for 69 years and 8 months. The cause of death was not reported for any species in the list. On the other hand, Orton's views have led to some hard thinking. There have been several attempts to learn whether or not fishes "over-reproduce" to the point where the effort interferes with other life functions. As we shall see, some studies affirm while others negate this point of view. Variability. There is a tremendous variation in fecundity, not only in different species but within the same species from place to place and among individuals of the same size. This variation immeasurably complicates studies on the effect of age on fecundity, and as a result the critical studies are of a statistical nature. The following sources of variation should be kept in mind while reviewing the evidence. 1. Size variation. Fecundity increases as the length and weight of the fish increases. 2. Individual variation. Every fecundity study has demonstrated a great individual variation in fecundity for fish of the same length and of the same weight. 3. Geographical variation. The fecundity of individuals of comparable size varies considerably from one locality to another. Maar's (1949) work on the char (Salmo alpinus) of the Faxalven Water System, Sweden, has shown that the egg 192 Shelby D. Gerking number per female may vary as much as fivefold in different lakes of the same watershed. In fact, geographical variation has been so prominent that "races" of some species have been delineated partly on the basis of fecundity. Davis (1957) has shown that the ova production of the American shad {Alosa sapidissima) varies considerably from one river system of the Atlantic Coast drainage to another, and McGregor (1923) could distinguish certain river races of the king salmon (Oncorhynchus tshawytscha) on the Pacific Coast on the basis of egg counts. 4. Year-to-year variations in the same locality. Individuals of the same size may bear significantly different numbers of eggs from one year to the next. This is true for the long rough dab {Hippoglossoides platessoides) off the coast of Scotland (Bagenal, 19576) and for the lake trout {Salvelinus namayciish) of Lake Opeongo, Ontario, Canada (Fry, 1949). These wide variations in fecundity are usually explained on the basis of genetic and dietary differences. The racial studies imply that heredity is very important in determining fecundity. Year-to-year variation has been attributed to dif- ferences in the availability of food. Both sources undoubtedly account for the fluctuations in the egg number of individuals of the same size. Because of the complex nature of the information on fecundity, evidence of the ageing process from a few species in which fecundity has received considerable study will be reviewed here. Methods of counting eggs. The usual methods of deter- mining fecundity vary from counting the entire number of eggs from both ovaries to counting samples from various sections of one or both ovaries and computing the total number of eggs on the basis of the sample. Three sampling methods have been used: volumetric, dry weight, and wet weight. In the first method the total volume of eggs is determined in a graduated cylinder and the number in a Physiological Changes with Age in Fish 193 known sample volume is counted. The fecundity is deter- mined by direct proportion. The dry- weight and wet- weight methods are essentially the same except that the counts are made either on eggs which have been dried to constant weight or on eggs taken directly from the preserving fluid. Like those of other animals, the ovaries of fishes contain eggs in various stages of development. Early workers debated the issue as to which eggs would be spawned. By measuring the diameter of the ova in fish just ready to spawn, by des- cribing their external appearance, and by examining histo- logical sections of the ovary, various categories of eggs were established. A comparison of these observations with those on the ovaries of spent fish was the basis for deciding which eggs to count for an estimate of the fecundity of an individual. These observations have been made in great detail, and they have resulted in learning that not all ripe eggs are shed dur- ing the spawning season. Usually the numbers retained are insignificant compared to those that are liberated. Ripe ova which are retained in the ovary quickly degenerate and are resorbed. Such observations have also been responsible for discovering that some species, like the long rough dab, do not spawn every year (Bagenal, 1957a) while in others, like the yellowfin tuna {Neothunnus macropterus), several batches of eggs may mature during a single, long spawning season (June, 1953). Herring. Wynne-Edwards (1929) immediately set about testing Orton's hypothesis about a possible disproportionate growth of the gonad in relation to body weight in the herring {Clupea harengus). Both ovary and testes reach maximum weight during the spawning season and decrease to an insignificant size immediately after the sexual products have been shed. The cycle of gonad growth then begins again in preparation for the next spawning. Since the reproductive tissue makes up about 20 per cent of the weight of the body at maturity, a considerable amount of the anabolic process is concentrated on the maturation of the sex organs. AGEING — ^V — 7 194 Shelby D. Gerking Wynne-Edwards argued that if the gonads of mature fish increase in weight each year at a rate greater than that of the body, then the increasing tax of spawning may bring about the animal's death. Otherwise, death must be due to other factors if the development of the reproductive organs is in harmony with the rest of the body. The main study was made on a sample of herrings, called "calf herrings", from the Irish Sea. Each individual was measured, the fish and gonads were weighed separately, and the age was determined by examining the number of annual rings on the scales.* A linear relationship was found between body weight and gonad weight. The testes were somewhat heavier than the ovaries in herrings of comparable size, but the rate of growth of the gonads was practically the same. The relative size of the gonads in relation to the body increased during the first four spawnings but remained constant thereafter. Wynne-Edwards also compared the ratio of gonad weight to body weight for fish of the same weight but of varying ages and learned that the ratio remained constant. He concluded that age was not responsible for any significant change in the weights of the gonad. He states, "There is no indication of an increasing tax which the fish cannot make up, the effects of which con- tinue to pile up until ultimately they cause its death, in a way that has sometimes been suggested". The growth of the gonads was in harmony with that of the other organs. The question appeared to be settled until Farran (1938) undertook a further study of the Irish herring. He was pri- marily concerned with the difference in ova diameter between autumn and spring spawners, the latter having larger and less numerous eggs than the former. During the course of the analysis, he related the weight of the ovary and the number of ova in 435 herring to the 4-5 power of the length of the fish, a value greater than the relation of total body weight to * An age group consists of all fish in a population sample which have the same number of annual rings. A Roman numeral is conventionally used to indicate the number of annual rings. Physiological Changes with Age in Fish 195 length. No ages were reported. This relationship was des- cribed without rigid statistical treatment, but it is apparent from his diagrams that more refined methods would have produced little change from his value. Farran's formula described the situation except for the largest fish, and it is these that interest us most. The ovary weights and number of ova for these individuals fall below the values predicted by Farran's equation. He recognized this departure and con- ceded that very large fish show a smaller rate of increase in the size of the ovaries than smaller fish. He interprets this growth pattern in the following way: "... that the rate of increase in number (of ova) in fish over 32 cm. in length ceases to cor- respond to the increase in length and either falls off consider- ably or ceases altogether." Contrary to Wynne-Edward's conclusions, Farran's results might indicate that reproduction places an increasing strain on the larger female herring, leading to a reduction in number of eggs. On the other hand, Farran may have simply described the normal course of ovary growth in relation to the body. Sensitive to both of the above workers' findings, Hickling (1940) took up the question, this time using the herring of East Anglia. Both of the previous workers had used only weight of gonads in their analyses. Farran had counted the eggs of only three specimens and used the ratio of number of eggs to ovary weight to calculate egg numbers in the remainder of his sample. Hickling counted the eggs of 136 herring of known age and length, and observed the sex, length, weight, and age of 475 additional individuals. Hickling agreed with Farran that the rate of gonad growth was greater than the rate of body weight gain in relation to length, and concluded that reproduction became an increas- ingly greater burden to both male and female as they grew larger and older (Table II). Even more interesting is the fact that the weight of the ovaries increased at a more rapid rate in relation to length (L^^^) than the rate of increase in the number of eggs (L^'*^). This was consistent with his analysis 196 Shelby D. Gerking of the weight of the gonad and egg number as related to age. The rate of increase in gonad weight with respect to age was greater than the rate of increase in egg number. Therefore Hickhng was forced to conclude that the permanent tissue of the ovary increases disproportionately as the herring grows larger and older. This suggests a gradual degeneration of the ovary, the reproductive tissues being replaced by connective tissue in much the same manner as the testis changes in Astyanax mexicanus in relation to age (Rasquin and Hafter, 1951). Unfortunately, there have been no observations of this sort on the histology of the fish ovary. Table II Regressions of body weight in grams (TF), gonad weight in grams (GW), AND fecundity {F) on body length in centimetres (L) in the ENGLISH HERRING {Clupca harengus). Taken from Hickling (1940) Males Females Body weight Gonad weight Egg number W = 0-0661 L2.312 GW = 2-41 X 10-5L4-237 W = 1-1471 Li-^se GW = 5 94> X 10-5L3-9'3 F = 0-2954 L3* 465 Research on the fecundity of the Pacific herring (Clupea pallasii) agrees with the conclusions of Hickling and Farran. Katz and Erickson (1950) analysed the relationship of fecundity and length in different age groups. This is a log-log relation described by the formula: F = CL^, where F = fecundity, L = length, and C and n are empirically deter- mined constants. The values of the exponent differed con- siderably among separate age groups. Those herring that were spawning for the first time (age II) were the least effec- tive egg producers {n = 3-46). Age groups III and IV were most efficient (n =3-89 and 3-87, respectively). The rate of increase in fecundity was considerably less among ages V to VIII (n =3-52). This result suggests that the relation of Physiological Changes with Age in Fish 197 fecundity to length is not a simple one and can be broken down into three parts, possibly a sigmoid curve. Eschmeyer (1950, 1955) has described a sigmoid fecundity /length re- lationship in the walleye {Stizostedion vitreum) and lake trout. The correlation of a decrease in the rate of egg production with age was regarded by Katz and Erickson as a criterion of ageing in the herring. A ballot on whether or not the strain of reproduction in the Atlantic and Pacific species of herrings upsets the homeostatic mechanism to the point where the fish dies as a direct or indirect effect of its reproductive efforts yields three votes affirmative, one vote negative. The affirmative votes should be scrutinized carefully, however, because none of them were cast after having taken individual variation into account. Salmon. Some interesting information has been accumu- lated which indicates that age influences egg production in two species of salmon. The Atlantic salmon (Salmo solar) spawns in rivers after spending either two or three years in the sea and some may spawn more than once. Belding (1940) studied the fecundity of this species from the Gulf of St. Lawrence and learned that the youngest spawners, those which had lived two years in the sea, produced a greater number of eggs in relation to their weight (834 eggs per lb.) than either three- year sea-life individuals (723 eggs per lb.) or those which had previously spawned (738 eggs per lb.). In actual numbers the two-year salmon produced 8,850 eggs per female and the three-year salmon produced about 14,000. He attributed this decline in relative egg production in part to the fact that large salmon usually have larger eggs than small salmon, but he did not discount age as a factor influencing egg production. Rounsefell (1957) has made a more detailed study of fecund- ity of the sockeye salmon {Oncorhynchus nerka) in the Karluk River, Alaska. In this part of the world the young sockeye spend either three or four years in freshwater lakes before migrating to the sea and remain there either two or three years before returning to freshwater streams to spawn. 198 Shelby D. Gerking Classifying the salmon on the basis of the same freshwater age but different ocean age, Rounsefell pointed out that the older salmon produced a significantly smaller number of eggs than younger ones. The average for the former was 2,987 per female and the average for the latter was 3,285, based on about 150 specimens in each group. The difference in fecund- ity for sockeye with identical ocean histories but different freshwater ages also favoured the idea that age has an effect on salmon fecundity since the older females produced 118 fewer eggs than the younger ones. The difference was not statistically significant and the data were more variable. The increased variation was explained by the variable freshwater environment as opposed to more stable ocean conditions. Rounsefell reviews the literature of the fecundity of the family Salmonidae, and of primary interest here is his demon- stration that generally the rate of increase in number of eggs declines as the fish increase in size. This may indicate that age influences egg production, or it may be simply a description of the way in which the ovary grows in relation to the rest of the body. It does explain, however, why trout hatcheries discard their old brood stock. The number of eggs in relation to the weight of the fish declines as the fish increase in size, and it is to the hatchery's advantage to have on hand a greater number of smaller breeders than an equal weight of older ones. Haddock. One of the most penetrating analyses on the effect of age on the fecundity of egg-laying fishes has been done by Raitt (1933). He was also influenced by Orton's writings. The mature eggs of 169 haddock {Melanogrammus aeglefinis) of Scottish waters were counted and fecundity was related to length, weight, and age. Four main comparisons were made: (1) rate of increase in fecundity with length, (2) rate of increase in ovary weight with length, (3) rate of decrease in body weight in relation to length during ovarian development, and (4) rate of increase in body weight with length. Comparison of these relationships indicated that fecundity increases with age up to age V, but at older ages Physiological Changes with Age in Fish 199 egg production declined (Table III). The values were based on regressions calculated separately for the different age groups. At similar lengths regular increases in fecundity with age occur among ages II, III, IV, and V. Ages VI and VII were combined in the calculations because individuals of Table III Fecundity of haddock {Melanogrammns aeglefinis) of the SAJME SIZE AT VARIOUS AGES. MODIFIED FROM RaITT (1933) Length Age II Age III Age IV AgeV Ages VI in cm. and VII 20 11,495 25 34,255 58,185 30 88,305 105,150 35 172,950 181,500 189,410 178,650 40 266,450 283,850 290,900 276,950 45 415,150 424,250 50 591,600 595,550 55 809,850 796,600 these ages were scarce in the population. Their fecundity was about five per cent lower than that of age V. At extreme lengths the latter group has a slightly lower fecundity than age IV of the same length. The same result was found when the fecundity of haddock of the same weight were compared at different ages. The effect of age was shown to be due to the relationship between fecundity/length and body weight/length. Raitt calculated the first of these comparisons separately for the different age groups. Fecundity increased very rapidly in the youngest spawners of age group II {F = 0-005187 X L*'^^), but was considerably lower in the remainder of the popula- tions, as the remaining equations show : age III {F = 1- 788 X L^-^% age IV (i^ = 1-527 X L^'^^), age V (i^ = 2-069 X L3-21), and ages VI and VII combined (2^ = 1-546 X L^^^). Raitt used Russell's extensive data (1914) on the haddock to 200 Shelby D. Gerking compute the rate of increase in body weight with length. Here he found the relationship to he W = 0-0044 L^'^^ based on the mean yearly weights as recorded by Russell. At all ages fecundity increased more rapidly than body weight. The relationship between fecundity and body weight was: F = 196 X PF^'^* (all ages). Since the exponent is greater than one, the equation confirmed the above interpretation that fecundity increases at a rate greater than body weight. The rate of increase in ovary weight with length was consistent with this result. The rate of decline in body weight with length during ovarian development showed that somatic tissue was being converted to gonad from November to June, the spawning season, and that somatic tissue increased from July to November. The consistency of Raitt's results arouses serious suspicion that reproduction is a drain on the individual in later life. In his words, "One cannot but regard the above evidence as hinting at an end point to reproduction, and inviting postula- tion of stress of egg production, ultimately overbalancing ability to recover within the annual cycle. It would seem that an affirmative answer is suggested to Orton's question of whether 'over-reproduction' may be regarded as a general predisposing cause of death in fishes." Long rough dab. Bagenal (1957a) has recently studied the fecundity of the long rough dab of Scotland by detailed statistical procedures. His fish were caught in one locality by a small mesh cotton trawl. Length, sex, gutted weight, gonad weight, and age were determined for a large series of specimens taken from October 1933 to May 1955. Egg counts and the foregoing measurements were made on two samples, totalling 119 females, one caught in February and the other in March 1954, just before spawning. By an analysis of covariance he was able to show that there was no effect of age on the length/ weight regression so all ages were pooled for the calculation of the regression coefficient describing this relationship. Simil- arly, age had no significant effect on the ovary weight/body Physiological Changes with Age in Fish 201 weight regression. There was a significant difference due to age on the ovary weight/length relationship in the March sample, but on good grounds Bagenal considers this to be an anomalous result. Weight increases at a power 3-11 of length in maturing females while ovary weight increases at about the 3 • 5 power of length. The latter figure is an estimate because the entire data would have to be recalculated in order to obtain an overall coefficient for the two samples. Since the ovary increases at a more rapid rate than body weight, atten- tion is again directed to Orton's theory of" over-reproduction". Whether or not there is a statistical difference between these coefficients would require a separate analysis. There was no effect of age on fecundity /length or fecundity/ weight relationships according to a covariance analysis. Thus differences in these relations could not be attributed to age but to the individual variations that occur in fecundity at any given length or weight. This result makes us more cautious in accepting the rather small decline in fecundity of six- and seven-year-old haddock which Raitt found. He presented no measure of deviations from the regressions, and since the individual measurements were not given the computation cannot be performed. Even though Bagenal was satisfied with statistical proof that age played no significant role in determining ovary weight or fecundity, he was disturbed about two features in his data which did not coincide with this interpretation. Milinsky's (1944) very large dabs from the Barents Sea did not produce the egg numbers expected of their size, based on the Scottish population. The difference might be explained by geographical or by racial variation, which is often very great. Also, the number of eggs did not increase in proportion to the weight of the gonad. The regression coefficients between these two variables are 0-6907 and 0-8117 in his two samples computed from a regression of the logarithm of fecundity on the loga- rithm of ovary weight. The latter result was, as Bagenal says, "... unexpected since the larger gonads will have a 202 Shelby D. Gerking proportionally smaller surface area and so, not only should carry less surface moisture when they are weighed, but also less ovarian tissue should be found surrounding the eggs in the larger gonads. We can only suppose that the heavier gonads produce fewer eggs per gram than do the lighter ones, so the eggs are presumably larger and heavier." It is also possible that the amount of connective tissue of the ovary may in- crease disproportionately as it grows larger. After comparing the information on the haddock and long rough dab, we have mixed feelings. Raitt's data were con- sistent throughout and left the impression that there was a slight but definite effect of age on fecundity although statis- tical tests were lacking. Bagenal, on the other hand, offers statistical proof to the contrary but cannot explain satis- factorily at least one important feature of his data from a purely biological point of view. Plaice. A long history is associated with studies on the fecundity of the plaice. Just before the turn of the century Reibisch (1899) completed a detailed study of the histology of the ovary and performed many egg counts on plaice from the Baltic Sea. He was acquainted with the fact that egg production declines with age in higher vertebrates and was puzzled to find that this was apparently not true in fishes. One case was pointed out where an older and larger individual produced the same number of eggs as a younger one. This observation was by no means consistent, and Reibisch was unable to reach a definite conclusion about the effect of age on fecundity. Soon after, Franz (1909) duplicated Reibisch's work and again was unable to answer the question. Franz admitted that he had insufficient material from older age groups to judge whether older plaice had a greater or lesser egg number than younger ones. Individual variation was very great in the specimens above age X in his sample. The lack of sufficient old specimens has plagued all the studies to date and con- stitutes the chief source of difficulty in settling the problem. Physiological Changes with Age in Fish 203 The older age groups are represented by so few individuals in the population that it is virtually impossible to collect enough material on which to base critical judgments. A thorough review of plaice fecundity has been done recently by Simpson (1951), who added a considerable number of egg counts from the Southern Bight and Flamborough regions of the North Sea. Ovaries were gathered from a total of 256 females taken just before spawning, from mid-October to mid-February, in 1948 and 1949. By inspecting graphs of the fecundity of plaice of similar lengths against age and graphs of fecundity of fish of the same age against length, Simpson was convinced that age, apart from its relation to size, plays an insignificant part in determining fecundity. On the chance that statistical analysis might show up dif- ferences that a graphical inspection would not, a covariance analysis of Simpson's 1948 sample from Southern Bight (Table IV) was performed. Ages ranging from II to XVI are represented, although it was necessary to lump together ages II and III and ages XII through XVI in order to have sufficient numbers of observations in the younger and older categories. Regressions of body weight on fecundity at each age were calculated and compared. Simpson had found that fecundity bore a linear relation to weight in plaice, thereby simplifying the computations. The regression coefficients ranged from 0-054 to 0-273 and showed no trends with age. No significant diff'erences were found between the regression coefficients or between the adjusted means. Thus we conclude, as Simpson did, that age has no detectable eff'ect on the fecundity of the plaice. Simpson also measured ovary weights and found a rather high correlation between those and egg number (r = 0-890). He provided a complete tabulation of his data, and it was possible to make a more detailed study of this relationship. The Southern Bight information was again used, consist- ing of 163 pairs of observations. Egg number and ovary weights were converted to logarithms and the regression was 204 Shelby D. Gerking calculated to solve the formula : F = COJ^ ; where F = fecun- dity in thousands of eggs; 0^^ = ovary weight in grams. C and n are empirically determined constants. The advantage of this computation is that it is possible to judge whether fecundity is increasing in a linear fashion in Table IV COVARIANCE ANALYSIS OF REGRESSION OF BODY WEIGHT IN GRAMS AND FECUNDITY IN THOUSANDS OF EGGS IN PLAICE (PleurOliecteS platCSSO) OF VARIOUS AGES. Calculated from data of Simpson (1951) Source of Degrees of Regression Mean square variation freedom coefficient from regression Age II-III 9 0-099 68-06 IV 35 0-183 238-09 V 8 0-273 146-99 VI 5 0 054 102-19 VII 17 0-117 116-81 VIII 25 0-113 470-47 IX 17 0-148 13648-19 X 9 0 138 1017-94 XI 11 0-098 828-72 XII-XVI 11 0-164 453-74 Within age groups 147 481-98 Due to regression 9 899-11 Common to all age groups 156 0-140 506-04 Due to adjusted means 9 605-99 Total 165 511-50 F for regression coeff icients: 899-11 = 481-98 1 • 87 (not significant) F for adjusted mean 3: 605-99 = 506-04 1 • 20 (not significant) relation to ovary weight or whether it is increasing more rapidly or less rapidly. The exponent, n, would not deviate significantly from 1 in the first case, would be < 1 in the second, and > 1 in the third. The solution of the equation was: F = 7'14 OJ^'^'^^^ with 95 per cent confidence limits of Physiological Changes with Age in Fish 205 the exponent lying between 0-5553 and 0-6035. Thus, the number of eggs does not increase in proportion to the weight of the ovary (Fig. 2). Plaice is the fourth species to show this peculiar relationship. It had also been reported by Hickling in the English herring, by Raitt in haddock and by Bagenal in the long rough dab. SOOr 240 180 120 60 • ^ •• • • •• PLAICE 60 120 180 OVARY WEIGHT IN GRAMS 300 Fig. 2. Scatter diagram of the relation between fecundity and ovary weight in the plaice (Pleuronectes platessa). From Simpson (1951). The explanation for it is still in doubt, but a clear-cut problem has emerged. A choice can be made between Bagenal's inter- pretation that larger ovaries may produce larger but fewer eggs, and Hickling' s tentative conclusion that connective tissue may increase disproportionately in the ovaries of older fish. The decision can be made by studying the histology of the ovary, provided a method can be devised for measuring 206 Shelby D. Gerking the relative amounts of connective tissue. This would have to be done in maturing ovaries where maximum egg diameters could be measured at the same time. There seems to be no reason to continue studying fecundity of egg-laying fishes with respect to age for gathering evidence for or against the ageing process. Individual variation masks any effect that age may have. Great variation plus the difficulty in collecting a sufficiently large number of old individuals makes it very improbable that this line of research will ever become profitable. Summary Nutrition and reproduction of fishes are reviewed in relation to age. A decline in the ability to utilize protein for growth is exhibited as fish grow larger and older. This conclusion is based on laboratory feeding experiments on the longear sunfish (Lepomis megalotis), green sunfish {Lepomis cyanellus), bluegill sunfish {Lepomis macrochirus), and red hind [Epine- phalus guttatus). Other vertebrates also conform to this pattern, although they achieve a specific size relatively early in life in contrast to the prolonged period of growth in fishes. The reproductive capacity of live-bearing fishes of the family Poeciliidae declines with age and there may be a period of sterility before death. These results were obtained by field observations and laboratory experiments on the fecundity of the western mosquitofish, Gambusia affinis. Scattered ob- servations among other species in the family agree with this viewpoint. The effect of age on fecundity in egg-laying fishes is not yet clear. The number of eggs in three-year sea-life sockeye salmon (Oncorhynchus nerka) is significantly lower than in two-year individuals. The same general phenomenon has been said to be true of haddock {Melanogr animus aeglefinis), but there is some doubt about this conclusion since statistical Physiological Changes with Age in Fish 207 procedures fail to show any effect of age on the fecundity of either the long rough dab (Hippoglossoides platessoides) or plaice {Pleuronectes platessa). Individual variation in fecundity is very great and masks any effect that age may have. Ovary weight and fecundity increase more rapidly in rela- tion to length than does body weight in the long rough dab, haddock, and herring (Clupea harengus and Clupea pallasii). This result strengthens Orton's hypothesis that reproduction becomes an increasing strain on the metabolism of fish as they grow larger and older, thereby causing death either directly or indirectly. The number of eggs does not increase in proportion to the weight of the ovary in the haddock, dab, herring, or plaice. Either larger ovaries produce larger and fewer eggs or con- nective tissue increases disproportionately in the ovaries of larger fish. No critical evidence is available to support either contention. If the latter is true, ageing changes in the gonads of fishes would be similar to those in higher vertebrates. REFERENCES Bagenal, T. B. (1957a). J. Mar. hiol. Ass. U.K., 36, 339. Bagenal, T. B. (19576). J. Mar. hiol. Ass. U.K., 36, 377. Belding, D. L. (1940). Trans. Amer. Fish. Soc, 67, 285. Bidder, G. P. (1925a). Nature (Lond.), 115, 155. Bidder, G. P. (19256). Nature (Lond.), 115, 495. Bidder, G. P. (1932). Brit. med. J., 2, 583. Davis, W. S. (1957). Res. Rep. U.S. Fish Serv., 49, 1. Eschmeyer, p. H. (1950). Bull. Inst. Fish. Res. Univ. Mich.y 3, 7. EscHMEYER, P. H. (1955). Trans. Amer. Fish. Soc, 84, 47. Farran, G. p. (1938). J. Cons. int. Explor. Mer, 13, 91. Franz, V. (1909). Wiss. Meeresuntersuch. (AM. Helgoland), 9, 59. Eraser, E. A., and Kenton, R. M. (1940). Quart. J. micr.Sci., 81, 479 Fry, F. E. J. (1949). Biometrics, 5, 27. Gerking, S. D. (1952). Physiol. Zool., 25, 358. Gerking, S. D. (1954). Ecology, 35, 490. Gerking, S. D. (1955fl). Physiol. Zool, 28, 267. Gerking, S. D. (19556). Phijsiol. Zool., 28, 283. HiCKLiNG, C. F. (1940). J. Mar. hiol. Ass. U.K., 24, 619. IvLEV, V. S. (1939a). C.R. Acad. Sci. U.R.S.S.y 25, 87. 208 Shelby D. Gerking IvLEV, V. S. (19396). Bull. Soc. Nat. Moscou, 48, 70. IvLEV, V. S. (1939c). C. R. Acad. Sci. U.R.S.S., 25, 90. June, F. C. (1953). Fish. Bull., U.S., 77, 47. Karzinkin, G. S. (1939). Arb. limnol. Sta. Kossino, 22, 219. Katz, M., and Erickson, D. W. (1950). Copeia, p. 176. Krumiiolz, L. (1948). Ecol. Monogr., 18, 1. Larkin, p. a., Terpenning, J. G., and Parker, R. R. (1957). Trans. Amer. Fish. Soc, 86, 84. McGregor, E. A. (1923). Calif. Fish Game, 8, 138. Maar, a. (1949). Rep. Inst. Freshw. Res. Drottning., 29, 57. Maynard, L. a. (1951). Animal Nutrition, 3rd ed. New York : McGraw- HiU. Menzel, D. W. (1957). Ph.D. Thesis, University of Michigan. MiLiNSKY, G. I. (1944). Trans. Knipovich polyar. sci. Inst., 8, 388. MoRGULis, S. (1919). Trans. Amer. Fish. Soc, 48, 34. Orton, J. H. (1929). Nature (Lond.), 123, 3088. Raitt, D. S. (1933). Rep. Fish. Bd. Scot., p. 1. Rasquin, p., and Hafter, E. (1951). J. Morph., 89, 397. Reibisch, J. (1899). Wiss. Meeresuntersuch. {Abt. Kiel), 4, 233. Rounsefell, G. a. (1957). Fish. Bull., U.S., 122, 451. Russell, E. S. (1914). Fish. Invest., Lond., Ser. 2, 1. Simpson, A. C. (1951). Fish. Invest., Lond., Ser. 2, 17, 3. SvARDSON, G. (1949). Rep. Inst. Freshw. Res. Drottning., 29, 115. SvARDSON, G. (1951). Rep. Inst. Freshw. Res. Drottning., 32, 79. Wallace, W. (1925). Nature (Lond.), 115, 337. Wynne-Edwards, V. C. (1929). J. Mar. biol. Ass. U.K., 16, 49. DISCUSSION Comfort: We have done some work on regeneration in guppies (Comfort, A., and Doljanski, F. (1958). Gerontologia (Basel), 2, 266) which might have some relevance to what you said about protein utilization, Dr. Gerking. We cut off the tips of the tails and measured the percentage restoration in length at various times. In female guppies up to three years of age the growth curves were typically asymptotic, and the corresponding curve for percentage restoration of an excised regenerate was roughly a mirror image of them. In fish that had been kept without much to eat, the growth curve flattened out and the regeneration rate fell exactly as in freely growing fish approaching full size. Full-size and retarded brood- mates were therefore behaving in almost the same way as regards regeneration. If retarded fish are then allowed to grow, the rate of regeneration rises until growth declines again. In other words, as the asymptote for size under given conditions is approached, the Discussion 209 regeneration rate falls with the growth rate. Growth eventually declines to zero, and the regeneration rate to its basal level, which persists even in starving fish. Danielli: Are these measurements in terms of percentage re- generation of what was removed ? Comfort: Yes. They are not absolute measurements. We had to adjust the size of the amputate to the size of the fish. The general finding seems to be that as somatic growth flattens out, the regenera- tion rate comes down to its basic level; as growth is restarted, so regeneration is restarted. I would predict that this occurs also with nitrogen utilization, as in Dr. Gerking's experiments. As G. V. Samokhvalova has shown (1952. J. gen. Biol., Moscow, 12, 153), in guppies during the first part of life the number of young per brood is a function of the size of the female. I have kept them up to four and a half years of age. Brood size declines fairly rapidly even though body growth continues, and there is quite a long post- reproductive period during which females may produce one or two broods if remated, but generally they do not. Gerking: It seems that my paper would have been much more com- plete if your experiments had been published a little earlier. Holt: Some light could probably be thrown on the regeneration question if the growth could be modified not just by changing the food supply, but also by changing the temperature. You could then see whether the regeneration curves behaved in the same way. In our terminology your food supply is changing the Loo (the asymptotic size), whereas the temperature would change K (the rate of approach to the asymptote). Comfort: I am now engaged in temperature experiments. If two batches of fish are allowed to regenerate at different temperatures the final percentage restoration is identical, but it is reached at a different rate. It is striking that, as far as I have got, quite large shifts in temperature do not alter the final percentage of restoration; they only alter the rate at which it is approached, as you suggested, Mr. Holt. Rotblat: One of your graphs [not printed] appears to be the differential of the other. Comfort: In general but not invariably. If very old retarded fish are kept very long, as is often the case, some of them will not restart growth. They grow only very sluggishly or very little. Nevertheless the regeneration rate still rises. Rotblat: Even when there is a small change in weight, the initial slope appears to increase. Comfort: Some of them do not increase at all or show barely 210 Discussion measurable growth. But even in these the regeneration rate rises, and I would suggest that the nitrogen uptake rises too. Roiblat: Does this mean that the rate of restoration equals the rate of growth? Comfort: The response of regeneration to a growth-promoting stimulus is more sensitive than that of body growth. That is why I would not like to say that regeneration rate is a direct function of growth rate. Gerking: Do your findings apply to the male ? Comfort: No; the male guppy not only stops growing rather suddenly and early, but also it has tail shapes of different kinds. As far as we have got, in males with small, wild-type tails the per- centage regeneration falls in relation to growth cessation, as in females, but the basic restoration rate always stays higher than in the female. You sometimes find male guppies of all ages which for some reason have not quite completed their growth and which have a very high restoration rate. In strains with big tails, it appears that the rate of restoration is based, as it were, on the wild-type tail, but that anything after that is extra. Holt: Dr. Gerking, you have looked at particularly good sets of data and put aside the incomplete oddments. I have had a look at the oddments and have the impression that although no single species shows a statistically significant decline in fecundity (eggs per gram) with increase in size, yet in many cases the points for the larger fish fall below the proportional line. Gerking: That is very true. I am not convinced whether the decrease is significant or not. Holt: In this kind of study you may be seeing the effects of less fecund fish surviving longer. What we cannot do is to follow a cohort. We need a method of determining the fecundity of live individuals and marking them. Rockstein: Was this large longear sunfish that utilized very little protein caught at a great depth, Dr. Gerking ? Gerking: No, it was caught in a shallow stream. This observation needs to be confirmed, and I do not wish to over-emphasize that portion of Fig. 1. It may indicate that extremely old fish utilize very little protein for growth. Nigrelli: In the experiments carried out in Bermuda on the angel- fish you characterize these fish as herbivores. Gerking: We have thought up to this time that they were herbi- vores; it is known that they feed upon algae because algae can be found in the gut. Menzel's experiments (1957), however, indicate that if they are fed exclusively on algae they will not grow, even if Discussion 211 they eat fairly large quantities. I would conclude that these angel- fish are not strict herbivores but that they must have some animal food. Nigrelli: That is true, because we keep small angelfish in our aquarium and feed them with clams, etc., and no algae at all, and they grow to a good size. However, the algae may prevent growth by producing antibiotics. Gerking: We wonder now whether there is a strictly herbivorous fish. V LONGEVITY OF FISHES IN CAPTIVITY, WITH SPECIAL REFERENCE TO THOSE KEPT IN THE NEW YORK AQUARIUM Ross F. NiGRELLI New York Aquarium The New York Aquarium, at one time or another during its history, has exhibited for varying periods of time fishes representing 33 of the 57 known orders. This figure is more spectacular when it is reaUzed that 10 of the 57 orders contain species restricted to fife in the benthic areas of the seas, which have never been brought to the surface ahve long enough to exhibit. The orders, as listed by Berg (1947), contain 425 families, at least 50 of them deep-sea forms. The New York Aquarium has exhibited species representative of 152 families. Other aquaria have probably kept fishes of another five or six orders, involving an additional 25 families. Therefore, it is apparently possible to maintain in captivity fishes represent- ing 38 of the 57 orders and from 175-250 of the 425 families. However, only relatively few of the estimated 25,000 species have ever been captured and kept alive in aquaria; the exact number has never been determined. Jordan, Evermann and Clark (1930) list 4,137 species in the North and Middle Ameri- can waters, and probably only a little more than one-fourth of these have ever been exhibited. Breder (19366) summarizes some of the environmental and physiological barriers that may be limiting factors for the successful maintenance of many species. He concedes, however, that these barriers may be overcome and that some day it may be possible to exhibit species not heretofore shown, including the exotic life in the great depths of the oceans. 212 Longevity of Fishes in Captivity 213 It is generally known that fishes represent a physiologically highly diversified group of vertebrates, and attempts to maintain them in so-called standard aquarium conditions regardless of environmental origin are fraught with danger and will invariably be reflected in a short lifespan in captivity. As environmental origin is recognized as important in aquarium management, it is the practice in the New York Aquarium, in so far as it is possible, to diversify physical and chemical conditions such as temperature, pH, surface-volume ratios, conditioning factors, w^ater movements, salinity and illumina- tion. As a result of such management we have been able to increase our survival rates significantly. Thus, in 1940, our last full year of operation at the New York Aquarium at the Battery, average longevities were as follows: marine fishes 9*58 months, temperate freshwater species 24*50 months, and tropical freshwater fishes 11-22 months. The mortality rate for the year was 169 per 1,000. Infectious diseases are the primary causes of death in aquaria (Nigrelli, 1940, 1943). It is generally known that fishes are susceptible to a large variety of metazoan parasites, but what is not common knowledge is that they are also prone to infections by pathogenic micro-organisms that are similar in many respects to those responsible for diseases in man and other mammals. For example, to mention a few, fishes are susceptible to infections by viruses, Rickettsia, PseudomonaSf Proteus, diphtheroids, tubercle bacilli (Myco- bacterium), Monilia and other mycotic organisms. They are also susceptible to such protozoans as trypanosomes, haemo- gregarines, coccidians, babesioids, and Toxoplasma, in addition to such ubiquitous parasites as flukes, tapeworms, nematodes and acanthocephalans. The most important aetiological agents of fish diseases, however, belong to a subclass of Sporozoa called cnidosporidians. These are truly spore- producing parasites in which transmission is direct, i.e. by ingestion of the spore. During the course of routine autopsies at the New York Aquarium in the last 25 years, the present 214 Ross F. NiGRELLI author has observed more than 150 species of enidosporidians (myxo- and microsporidians) in more than 1,000 species of fishes (see Walford, 1958). These parasites are often tissue- and cell-specific and have been found in all the tissues and organs, including the eyes, brain and heart. The parasites produce a variety of lesions, the extent of which varies with the species of parasite and degree of infection. They may cause no more damage than the development of a simple cyst, or they may cause acute and chronic diseases. For example, some may produce cellulitis, cystitis, nephritis, hepatitis, enteritis, pericarditis and endocarditis; others may induce tumours of the infected organs and/or the surrounding tissues, many of them bordering on true neoplasia; still others may cause hyaline degeneration of muscle and other tissues. It is safe to say that all fish harbour one or more kinds of parasites. The resistance of fish, or the rate at which they can acclimatize to changes in the environment, appears to be related to their parasitic load. Experience has shown that as a rule a 10 per cent mortality can be expected when fish are first netted or trapped randomly, another 10 per cent as the result of handling and shipment, and 10 per cent more will succumb in the first few weeks of life in the tanks. Such fish tend to show a relatively high degree of parasitaemia, and their ability to withstand shock is related to the intensity and the site of infection or infestation, the rates being highest for those fish in which infections are localized in the kidneys, gills and skin, which are important organs of osmoregulation. Selective methods of trapping, handling and shipping usually result in higher survival values. For example, fish caught in traps rather than in nets are less subject to trauma and can be transferred to holding-pens with very little injury. Survival rates are further increased if the fishes are starved for a period of time and before shipping are transferred to waters with slightly altered densities. Relatively young fish, as related to potential age, are better risks than yearling or older (larger) fish. Longevity of Fishes in Captivity 215 Conditions in aquaria at best are still artificial since move- ments of fishes are restricted, and for this reason there cannot be any escape from environmental stress. It is apparent, then, that fishes that survive aquarium conditions are those that can withstand shock stresses and can acclimatize to a variety of exaggerated environmental factors. But even such fish as these are often at the limits of their tolerance, and any sudden change in one or more of the physical, chemical and bio- logical factors often results in death or increased suscepti- bility to infections. Invariably these infective agents are ex- ternal (gill and skin) protozoan and helminthic parasites; only rarely are they bacterial or mycotic organisms. This would indicate that once the fish is acclimatized to its new environ- ment (captivity), its resistance is increased and the parasite load diminished to a point where immunity is maintained by pre- munition. Diseases caused by internal parasites are often self- limiting and in some instances may spontaneously disappear. Once a balance has been established between fish, parasites and environment, other diseases of a non-infectious nature may develop. It may not be surprising that the greatest single cause of mortality is associated with nutrition. Fishes, like other animals, are herbivorous, carnivorous and omnivor- ous and all need an exogenous source of vitamins and other nutriments. The main food source in the New York Aquarium consists chiefly of commercial-grade fresh and frozen fish, clams and Crustacea. The kind of fish used for feeding depends entirely on their availability on the market and in collecting areas. Feeding oily fish (mackerel, herring, etc.) over relatively long periods to species that normally eat invert- ebrates and non-oily fish frequently results in liver damage, commonly referred to as fatty degeneration but properly called fatty "metamorphosis" or "fatty change". The pancreas, kidneys and other organs may also be involved in this type of damage. Fatty changes may be due to a relative anoxia, the result of prolonged passive congestion. Other diseases of fishes indicative of disturbances in carbohydrate 216 Ross F. NiGRELLI and protein metabolism are glycogen storage comparable to von Girke's disease in man, cloudy swelling (albuminous degeneration), hydropic degeneration and amyloidosis of the liver and kidney, to mention a few. The development of melanosis, a common disturbance in fishes generally, certainly indicates profound changes in metabolism involving the amino acids phenylalanine and tyrosine, and the excessive accumulation of guanine crystals in the tissues of certain marine species in captivity is indicative of disturbance in purine metabolism. Further, it should also be emphasized that fishes are sus- ceptible to neoplasia (Nigrelli, 1952, 1954a). With the possible exception of typical leukaemias, all types of tumours and cancers, benign and malignant, occur in fish. The same basic tissues as in mammals are involved, with tumours of mesenchymal origin predominating. Although we do not have definite age statistics, there is ample evidence that such abnormal growths occur more frequently in older fish. This is especially true for sarcomas and lymphomas, basic tumour types that appear more frequently in young persons. This finding, together with the fact that tumours in fish are usually slow-growing, may have some physiological and phylogenetic significance. It serves no purpose here to extend the list of metabolic diseases that we have found in fishes in captivity. The vertebrate fish is no different from vertebrate man in these respects. It is sufficient to say that we have ample evidence that some of the metabolic diseases may be hereditary or congenital in origin, or that, in older fishes at least, they may result from nutritional disturbances and hormonal imbalances. The pathological changes are generally similar to those found in warm-blooded vertebrates; in extreme cases they include hyperaemia, anaemia, haemorrhage, inflammation, sclerosis, atrophy, hypertrophy, hyperplasia, oedema, ascites and necrosis. Basically, such conditions result in disturbances in the electrolyte balance, thus affecting homeostasis. Longevity of Fishes in Captivity 217 The literature concerning growth and senescence in fishes has been reviewed by Comfort (1956) and Brown (1957) and it is generally agreed that many of the larger teleosts, and perhaps certain sharks, continue to grow throughout life. Most fishes, however, reach their maximum growth, often with sexual maturity, within a limited time. Nevertheless, the growth pattern and lifespan of only a relatively few but well-known species have been established by fishery biologists. The lifespan of the more exotic forms is based mainly on longevity records kept by various aquaria. These, and others, have been summarized by Flower (1925, 1935), Bourliere (1946), and more recently by Hinton (1959, personal com- munication). In 1956 Hinton canvassed twenty institutions in Europe and the United States for information concerning longevity of fishes in attempts to bring these records up-to- date. The New York Aquarium's longevity lists for fishes 4 years or over are included in this report as Tables II, III and IV. The data were obtained from published papers by Townsend (1904, 1913, 1928a, b), Mellen (1919, 1925), Breder (1936a), Nigrelli (1940, 19546) and from our own mortality records not previously published. In the most recent listing by Hinton, 325 species had lifespans of 5 years or more and these belong to 88 families. A further analysis of his data shows the following: 64 families contained 93 species with lifespans from 5 to 9 years, 11 months; 43 families contained 108 species that lived from 10 to 19 years, 11 months; and 18 families contained 28 species that lived more than 20 years. The latter group is listed in Table I. Families with the largest number of species that lived for 5 years or more are Charac- idae, 37 species; Cyprinidae, 39 species; Serranidae, 28 species; Sparidae, 12 species and Cichlidae, 17 species. The two orders represented by these families are Cypriniformes and Perciformes. Forty-four of the 88 families are represented by single species and the other families by two to nine species. A casual examination of all published lists of long-lived fishes in captivity shows, with few^ exceptions, the following 218 Ross F. NiGRELLI general characteristics : (1) they are phylogenetically primitive, (2) they are sluggish in their movements, (3) they are bottom inhabitants or live in fairly shallow waters, (4) they have accessory respiratory devices, (5) they can aestivate or hibernate, (6) they live in regions of extreme sunlight, (7) they are adapted to live in environments with low oxygen concen- tration (less than 5 p.p.m.), and (8) they are adapted to environments with extreme fluctuations in temperatures and salinities. The significance of some of these factors may not be apparent. For example, there seems to be some correlation between longer daylight and decreased respiratory rates, especially in fishes kept above 15°. Also, fishes that live on some coral reefs and banks no doubt are acclimatized to great changes in temperature and salinities. Certain "reef" fishes can even survive in fresh water, provided the calcium content is high. Thus, Breder (1934) reports 12 typically marine species that he found living in a freshwater lake (Lake Forsyth) on Andros Island, British West Indies. Further, experience has shown that many marine tropical species survive longer in captivity if the sea water is reduced in salinity to around 30 parts per thousand. Although there is much evidence that fishes (teleosts) undergo reproductive and actuarial senescence, there is very little information in the literature relative to pathological changes associated with ageing. Rasquin and Hafter (1951) and Hafter (1952) reported age changes in the testes and thymus of the teleost, Astyanax mexicanus. From the clinical records of the New York Aquarium, in addition to gonad atrophy we have found the following pathological manifestations to be most frequently associated with ageing: cirrhosis and fatty changes in the liver and kidneys, haemo- chromatosis, hypochromic anaemia, degeneration of the mucus-producing glands of the skin, hyperostosis of the haemal arch bones (see Breder, 1952) and other vertebral abnormalities. The haemochromatosis and anaemia are the direct result of changes in the kidney and spleen, important 2 3 3 3 o" cr o c tJD o 2 o Oh-; 3H •? ^ T3 .;i o > a § s s ■c .2 .2 E3 fc- ti 3 cS as „ era 3 c«< CO" .2 g ^ ^ .2 g^ 3 *u C5 O" C3 rH C j^ o o C3 fe £3 333 .323 3g 3 J3 e 'C ^ t- c Ja 'S .2 "C c 'C 3 rt 2..2 ss.='.2r'^=''^=333rt § o" E cr g" g-'C U cr 3 cr-c cr 3 cr< ^ < < < ^ r^ 5'< -'^'-'(Me0'^C00iOO©M00i-(rHQ0r-(O«0b-OOOO'^CiC5OC0 k^ WCl CO 05 S so '^ IS ^ ? Co CO O CO w 3 o o w^ .£fJ OJ c3 (U ehPh;z;ph O C3 t-i D c o I 2 O 0) •a S »- CO • ^ «0 O c s a O o c3 ■c* CJ i^i O W VJ »i -^ '^ S2 ^ - ^ S5. =0 Coo CO ^ CO ^^ > g ;S§^ P S o ^ ^ ^ ^^^c4&qd;a^ CO o .1 « CO ?^ § « . CO CO O 555 cS c5 4^ eg 0) CO •,-- ex- es D Six! >^ O 1| S 0) c5 D .— VI OJ 05 O 0) 05 • p-l fl) I es tfw:'*O"^O"^O'^o-^a0'^"^ifl«Tf(TjiTfi'ij( ao s: s i (^ o ?• «"« § C3 •a 3 0) eSTS it 0) 35 •=; ^ TS w 03 C CO 0) 03 c8 d +j 73 .5 — C .5 O 5 '^ - c3 CO o hocq OQ O o Q O i I eo*- eOCOlN riri r-( M t < o o g o c o ti o ^ ^ 5 cZ .«- -^ o a> O - , ^ ta 2 Q O 3 2 «^ rt £ Sr; o cj c ^ o.-S 5 •^ -J o o ^ i 0) 3 o 3 S e«9 O S s 05 a S 55 w • • 60 c 65 60 « I S 65 ?: .< ■S^ ^ ^ a, ^ ^ • ;j r» ^ • S ^ J* S o •I: i *5 «-« s ^ vj C» n QQ fi H ^ S Pi PS V 0 p ^ g 09 ^ 1 S H OQ Ph 22 Siq 0 2 H U* Q < 45 03 Pi O Ph t6 M ei O»0t>l>O'*OO>0OOOTjiNrHt>>^'»J>»J(00Tfixcqoi0W5 4) •2 a> C/3 O -S -« J? -^ .S ^ CC :1 ^ 1 1 H o <5 ^ -5 .2 55 'C C/5 5 3 a: .2 C2 TJ C/2 3 +-> T3 5c/2cn« C/2 '^ £ e 1^ w 3 3 O 3 fc- 1> 1/ c3 rt « o _«J — S to c3 o Q s " ? O o e -^ s a «3 o o H c a o Co c3 o CO 2 ^ Sj S 53 o ^ ^ S of ^ 1 < ifefi Ra yin tfisl om: our na Coates' Kni Electric Ee; Three-lined Slender Ra; Malayan Fl Electric Cal Pike-killie Snake-head Orange Chr Croaking G 4) "S CO 1 c O CO -i CO "Si o Is v3 i^j 5 e 00 CO ■£ o 05 -o s. C o 5: 4? S K o o ?: • CO ■»~> CO CO « ^ « CO c '^ e 7 » .^ o 05 rs i» *c 03 a; t3 •t-j a* O o S 4J t-* o c8 t, J2 P^^ V 03 rt o •r' 03 13 OJ OKu t-l 3 c8 T3 +J i^ a ^ ^ o 'V '3 a i^p; 11 o h^u o o . . PS Oh W UOPk Longevity of Fishes in Captivity 225 organs of haemopoiesis. Fishes with these abnormahties may Hve for an exceptionally long time, even when the kidneys are extensively damaged, since the osmoregulation function of this organ can be taken over in part by the gills. Finally, arterio- or atherosclerosis has not been found in fishes. Summary Fishes are susceptible to a variety of parasitic and metabolic diseases, many of which are counterparts of those that occur in mammals. These diseases, together with changes in the physical, chemical and biological characteristics of the environ- ment, are responsible for mortality. Athero- or arteriosclerosis has not been found in fishes but pathological changes indicative of ageing do occur in other organs. These are cirrhosis and fatty changes in the liver and kidneys, haemochromatosis, hypochromic anaemia, degenera- tion of mucus-producing glands of the skin, hyperostosis of the haemal arch bones and other vertebral abnormalities. Some lifespans of fishes in the New York Aquarium and in other institutions are listed. The long-lived fishes kept in captivity have certain general characteristics, and the signifi- cance of some of these is discussed. REFERENCES Berg, L. S. (1947). Classification of Fishes, both Recent and Fossil. Ann Arbor : J. W^. Edwards. BouRLiERE, F. (1946). Annee bioL 22, 249. Breder, C. M., Jr. (1934). Zoologica, N.Y., 18, 57. Breder, C. M., Jr. (1936a). Bull N.Y. zool. Soc, 39, 116. Breder, C. M., Jr. (19366). Bull. N.Y. zool. Soc, 39, 149. Breder, C. M., Jr. (1952). Growth, 16, 189. Brown, M. E. (1957). In Physiology of Fishes, 1, 861. New York: Academic Press. Comfort, A. (1956). The Biology of Senescence. New York: Rinehart. Flower, S. S. (1925). Proc. zool. Soc, Lond., 247. Flower, S. S. (1935). Proc zool. Soc, Lond., 265. Hafter, E. (1952). J. Morph., 90, 55. HiNTON, S. (1959). Personal Communication. AGEINQ — ^v— 8 226 Ross F. NiGRELLI Jordan, D. S., Evermann, B. W., and Clark, H. W. (1930). Rep. U.S, Comm. Fish., No. 1055, 1. Mellen, I. (1919). Rep. N.Y. zool. Soc, No. 23, 101. Mellen, I. (1925). Bull. N.Y. zool. Soc, 27, 61. NiGRELLI, R. F. (1940). Zoologica, N.Y., 25, 525. NiGRELLI, R. F. (1943). Zoologica, N.Y., 28, 203. NiGRELLI, R. F. (1952). Ann. N.Y. Acad. Sci., 53, 1076. NiGRELLI, R. F. (1954a). Trans. Amer. Fish. Soc, 83, 262. NiGRELLI, R. F. (19546). Trans. N.Y. Acad. Sci., 16, 296. Rasquin, p., and Hafter, E. (1951). J. Morph., 89, 397. TowNSEND, C. H. (1904). Bull. N.Y. zool. Soc, No. 14, 163.^ TowNSEND, C. H. (1913). Bull. N.Y. zool. Soc, 16, 1049. TowNSEND, C. H. (1928a). Bull. N.Y. zool. Soc, 31, 11. TOWNSEND, C. H. (19286). Rep. U.S. Comm. Fish., No. 1045, 249. Walford, L. a. (1958). Living Resources of the Sea., New York: Ronald Press. DISCUSSION Rockstein: Which is the longest-Uved freshwater fish known? Nigrelli: In captivity it is this sturgeon which hved 69 years. Beverton: Sturgeons are also the longest-lived according to field observations; the longest we have ever seen recorded was 151 years, an estimate based on the number of rings in the rays of the pectoral fin. Are any age determinations of the conventional kinds done on these fish at death? Nigrelli: We have attempted to make age determinations but we have given it up as hopeless because at these steady temperatures they lose the markings. We do take measurements of the fish though, to give us some idea of growth. Scheidegger : In which organs have you found cancer? Nigrelli: In practically every organ. We have also seen virus tumours, and I have published a paper (Nigrelli, 1952) on virus in relation to cancer. There are two or three interesting "tumours", however, that are not of the neoplastic type, and one is well known in Europe — it is called lymphocystis disease (Nigrelli, 1954a). It was first described in English plaice but it is widely distributed both in freshwater and marine species. This is a hypertrophy of the con- nective tissue cells in which diseased cells blow up from about 10 fx to 0 • 5 mm. or more. This disease was definitely shown to be of virus origin. Holt: Most of the long-lived species are rather large, but there are clear exceptions, just as there are for mammals, where rats and bats of roughly the same size, as Mr. Sacher pointed out, have quite Discussion 227 different lifespans. We find the exceptions among the large fish in that sturgeons and tunas of about the same size have quite different growth rates and lifespans, but I do not know any examples of very small fish which are long-lived. None of those indicated as reaching 20 years or more were small species. By small fish I mean those with low upper limits to Lqq. Are there some such fish which commonly live for ten years or more? Comfort: There is a reported instance of a goldfish being kept till 40, but it was small and had been kept in a barrel. (Hervey, G. F., and Hems, J. (1948). The Goldfish. London: Batchworth Press.) NigrelU: That is an important factor in aquarium management. Fish are restricted in size by the size of the container, though not necessarily by the volume of water. Tarpon have been kept, for instance, for seven or eight years, and some have never grown larger than about 18 inches. An 8-year-old tarpon in the wild might be about 4 feet long or more. Holt: By small fish, I did not mean fish which were artificially kept small, but ones which cannot grow big no matter how much food or space they are given. NigrelU: There are several examples of small fish, hke goldfish, that live for about ten years (see Table III). Comfort: The matter of fatty degeneration is one which has always puzzled me when looking at pathological sections. The normal liver, in the guppy at any rate, looks very fatty but I do not know if it is degenerate. In underfed fish the liver is much more cellular. NigrelU: Glycogen infiltration is also evident in stained sections. Comfort: I suspect that a lot of what appears to be degeneration is reversible, and is actually fat storage. NigrelU: We call it fatty change and it does indicate a disturb- ance of lipid metabolism that is not reversible. Comfort: I believe you showed that fatty degeneration occurs in small aquarium fish fed only with enchytrae and nothing else? NigrelU: That is correct. Bourliere: It is interesting to notice that tuna are said to have a higher internal temperature than most of the other marine fish of similar size. Maybe there is here again a correlation between metabolic rate, growth rate and ageing rate. Has any study ever been made on the relative incidence of the various pathological causes of death in wild fish? Beverton: I do not know of any comprehensive study. A member of the Aberdeen Fisheries Laboratory has taken it up as his special subject in the last year or so. 228 Discussion Nigrelli: No critical work has ever been done to determine the exact meaning of natural mortality. Fishery biologists lump many things under this item. For example, epidemics of myxosporidiosis are quite common but not recognized, and the disease affects young populations much more than the old. Such an infection may be diffuse. All it takes to kill off a population after an epidemic is a slight change in environment, as the fish are already weakened by the disease. These diseases are often self-limiting, and if the fish can survive to, say, a year old, then they apparently can withstand changes in temperature, or other factors including pollution. If the fish die and are not examined for disease the death is then attributed to natural mortality. Holt: There is a current work which is relevant to this question but it is being done for another reason. Quite intensive population stud- ies of parasitization are being carried out on cod and salmon in an attempt to identify races or sub-populations. The results may well help us to interpret the mortality rates of, say, different sub- populations of the same species. Bourliere: Are there any indications about the rate of infestation by the various parasites in wild fish? Is it as great as in captive ones? Gerking: There have been some studies made in Canada of Tri- aenophorus (a tapeworm) infestation in whitefish, because it reduces their market value. There have been attempts made to reduce the incidence of this parasite by eliminating, as much as possible, the intermediate hosts, and there has been partial success. Rockstein: What is the rate of growth of the common carp, and its longevity? It is a fairly large fish, and it must therefore grow rather rapidly. Nigrelli: I cannot tell you about that. Gerking: One of the most recent developments in the study of large fish with rapid growth is some work on the sailfish in Florida waters (de Sylva, D. (1957). Bull Mar. Sci. Gulf and Carib., 7, 1). It is a well-known sport fish because it is large and can be caught on hook and line. It was amazing to learn that these fish grow very rapidly, living only four or five years, yet reaching 100 lb. Holt: Carp grow rapidly, but they have a low K; that is, their increments in length tend to be constant, but they have a particul- arly plastic growth. I think it is one of the few larger species in. which there is some evidence of a post-reproductivx phase. Gerking: Is that well documented? Holt: No — except that I believe in pond culture it is the practice not to keep large carp because their reproductive capacities are beginning to fall off. Discussion 229 Comfort: Goldfish breeders, who are not scientists, say that very often when a goldfish gets to about 16 years of age and is post- reproductive, it improves in condition and size for show purposes, and that a goldfish between the ages of 16 and 20 will often be a very large and particularly glossy specimen for show purposes. Gerking: What was said about the carp is true of trout also. Breeders do not keep old trout around because they are not efficient reproductively. This is not due to a sterility factor, however, but it reflects the growth of the gonad in relation to the weight of the trout. This is a special case that I did not have time to mention. The growth of the gonad is not proportional to the growth of the fish. Therefore, older trout do not produce as many eggs as a younger one does in proportion to the weight of the fish. It is more efficient for the hatchery to keep younger fish for breeding than it is to keep the same weight of older fish. Nigrelli: There are many parasites that affect the gonads. If there is an epidemic of gonadal strigeidiasis in any population, it will affect the reproductivity for that particular time. Holt: Would you suggest, then, that the older fish has had a longer time to be exposed to infestation, and therefore might be more likely to have a low rate of reproduction? Nigrelli: No, it would have to be an acute infection of some kind, which could spread around by contact, particularly in close schools of fish. Tuberculosis of the ovaries and testes will be found in relatively older fish. Beverton: You gave us examples of fish living quite happily under extreme environmental conditions. The opposite occurs as well. We have done some work on the Arctic cod over the last few years, and that is an interesting example of a fish which is living on the borderline of its environmental tolerance. When it goes across that border, in effect into the cold water from the Polar basin, the cold water seems to cause quite a substantial mortality. The immediate cause of death seems to be a breakdown in the osmoregulation, and the blood saline of these fish that are caught in cold water is up to four times the normal level. It is difficult to measure mortality in that case, but the inference is that the cold had quite a substantial effect. Danielli: If fish of the same age are put in different-sized con- tainers, they grow to different sizes. What happens if they are put into a larger container which is separated into two compartments, one large and one small, with a perforated screen which allows the passage of water but not of fish? Comfort: I have tried this with guppies. If you confine a fish in a 230 Discussion perforated netting cage inside a two-litre jar, it grows at a rate not much less than if it had the full swimming space. I do not think restriction of movement affects the growth of these fish. There is a very different effect when you put more than one fish into a tank. There is a social effect on size, which is partly behavioural and partly chemical. Where the fish have been kept in separate perspex com- partments and water comes rapidly through from a large tank, they seem to grow at a not much lower rate than they would have grown if they had been loose in the tank. Gerking: Many fish are very aggressive to others around them. A fish in company with others will fight or nip in an attempt to maintain territory. The social hierarchy is similar to that which has been described for the chicken and many other animals. These social factors play an important role in the rate of growth offish kept together in aquaria. Danielli: Has the experiment been tried of taking fish out of a tank in which a "pecking" order has been established, and putting them into a mirrored tank? Comfort: No, but if you remove the largest fish from a tank in which a size hierarchy has been established, for no good reason everybody "moves up" one place. Danielli: The fish in the mirrored tank would see one exactly the same size as itself. Comfort: You might try an enlarging mirror! This should be done with other fish and not with guppies, because guppies are extremely unaggressive. I have been very fortunate in having guppies which have never shown any sign of eating their young, and the counts from trap tanks are no higher than without traps. This is not true of all breeds of guppy, and it would be a perpetual reservation on all this work if the fish fight and eat each other's young. FACTORS INFLUENCING THE LIFESPAN OF BEES Anna Maurizio Bienendbieilung, Liebefeld-Bern The lifespan of the honey-bee and its dependence on internal and external influences have been the subject of numerous investigations during the last ten years. The problem is a complex one because the honey-bee does not exist as an individual insect but as a member of a colony. The characteristics of the honey-bee must therefore be con- sidered not only from the point of view of the solitary insect, but also from that of the colony. The lifespan of the colony is theoretically unlimited, but that of the individual bee is quite short. The lifespan is also very closely related to the physiological condition of the individual, which in its turn depends on the environment, nutrition and general condition of the colony. The author would therefore like to present this communication in two parts : (1) the lifespan of the bee in a free-flying colony, and (2) the lifespan of the bee when separated from the colony and caged. Lifespan and Physiological Condition of Bees in a Free -Flying Colony In those regions of the Temperate Zone marked by clearly defined seasons, the colony is inhabited by two types of bees : summer bees and winter bees. These represent two entirely distinct generations of bees with regard to their lifespan, their physiological condition and their mode of life. Here we are confronted with one of the most important and interesting 231 232 Anna Maurizio phenomena in the hfe of a bee colony, namely, overwintering, or the bridging of an unfavourable season during which the bees cannot forage for food. More primitive social insects like bumble-bees and wasps solve this problem by dispersing the colony in the autumn, leaving only the inseminated females to overwinter. In the case of the honey-bee, however, the entire colony overwinters with the queen, and the rearing of the first spring brood takes place before the active season begins. The difficulties of overwintering are overcome by producing in the autumn a long-lived generation with big body reserves; the colony is thus enabled to live through the winter and rear the first spring generation. To understand this, one has to consider in greater detail the characteristic habits of the summer and winter bees, and the ageing and seasonal changes which certain organs undergo. In particular, the pharyngeal glands and the fat body are affected; the wax, mandibular, and salivary glands, and the ovaries of the working bees, are also affected, though to a lesser extent (Anderson, 1931; Armbruster, 1931; El-Deeb, 1952; Evenius, 1937; Farrar, 1936, 1949Z?, 1952; Freudenstein, 1924; Gontarski, 1950, 1954; Haydak, 1934, 1937a, h, 1939, 1940a, h; Haydak and Tanquary, 1943; Koehler, 1921; Kratky, 1931; Levin and Haydak, 1951; Lotmar, 1939; Mauermayer, 1954; Maurizio, 1946, 1954; Nickel and Arm- bruster, 1937, 1938; Phillips, 1922, 1928; Ribbands, 1953; Rockstein, 1950a; Zander, 1947, 1951). Let us begin with the summer bee. In Central Europe it is to be found in the colonies between May and August. Its lifespan varies from a minimum of 25-35 days to a maximum of 60-70 days, and it emerges at the height of full colony development. It will find in the colony large patches of open brood, and very often the number of nurses is scarcely suf- ficient to feed the young larvae. During the first 10-14 days of its life the young bee consumes a large amount of pollen, thus stimulating full development and functioning of the Factors Influencing the Lifespan of Bees 233 pharyngeal glands. The newly emerged bee has undeveloped hyaline pharyngeal glands of irregularly angular shape; its fat body consists of a very delicate transparent cell layer with- out inclusions. In the approximately ten-day-old summer bee performing nursery duties, the pharyngeal glands are rounded, milky-white to yellow, with well-filled lobes completely covering the secretory ducts. In its third week of life, when the bee assumes different duties, the pharyngeal glands slowly degenerate but the wax glands attain full develop- ment. In the four-week-old summer flying-bee the pharyngeal and wax glands have already degenerated, and under normal conditions the fat body of the summer bee remains un- developed throughout its life. This form of presentation is somewhat schematic, because we know from studies carried out by Lindauer (1953), Ribbands (1952), Sakagami (1953), and others that the division of labour within the colony varies greatly; thus it is possible that the length of any of the above- mentioned duties can be prolonged, shortened or even omitted, and the degree of development of the pharyngeal and wax glands shows parallel fluctuations. The life of the bee emerging in autumn follows quite a dif- ferent course. The bee makes its appearance in the colony at a time when the brood is greatly reduced, the stores pre- pared and the bees ready for overwintering. The young bee feeds for several weeks on pollen without having to perform nursery duties (maximal pollen consumption is between September and October — Lotmar, 1939). Thus the pharyngeal glands and the fat body attain full development and remain in this state until spring, when the autumn bee takes up the nursing of the newly emerging brood. The pharyngeal glands of the winter bee therefore remain in the nursing state for six to seven months, and the fat body consists of several densely folded layers, the cells of which are milky- white, rounded, and packed full with inclusions. At the same time the bee's expectation of life increases, and it becomes a long-lived winter bee, remaining in the colony six to eight months 234 Anna Maurizio (maximum expectation of life 300-400 days — Anderson, 1931; Farrar, 1949a; Lotmar, 1939; Maurizio, 1954). Many attempts have been made to explain these striking differences between the lifespan of summer and winter bees. The customary conception is that the lifespan is regulated, in particular, by the collecting activity of the bee outside the hive. According to this conception the bee works itself "to death" on its collecting flights, when death is often caused by accidents (Phillips, 1922, 1928). However, exacting experi- ments have proved that where death is due to foraging the shortening of life is only a matter of between four and eight days, and thus is not long enough to explain the difference in lifespan between summer and winter bees (El-Deeb, 1952; Mauermayer, 1954). The genetically conditioned differences in lifespan between bees of different races and cultivated strains are also few. The bee is subject — like any other living creature — to the process of physiological ageing, which is based on the destruction of certain nerve cells. Rockstein (1950a, 6, 1953) and Weyer (1931) estabhshed that in the ageing bee the number of nerve cells in the olfactory lobe and in the lower pharyngeal ganglion decreases by 35 per cent. Factors influencing the bees' lifespan become clear only when one considers the behaviour of bees under experimental conditions. For example, if one forces a young summer bee in the nursing state to undertake continuous brood-rearing, the pharyngeal glands will remain at their full development for several weeks and thus increase its lifespan (Moskovljevic, 1939). This increase, as well as an enhancement of physio- logical condition, accompanies the limitation of brood nests in a colony ready to swarm. Particularly striking is the behaviour of bees when put into a queenless or broodless colony during the summer. Such bees not only possess for several weeks fully developed pharyngeal glands, but also form a winter fat body and thus become long-lived (maximum life-span 166 days, as against 38 days in a colony with brood — Maurizio, 1954). Therefore it is possible to transform a Factors Influencing the Lifespan of Bees 235 summer bee during the active season into a physiological winter bee by preventing it from brood-rearing. Further indications of the interdependent effects of food, care of brood, physiological condition and lifespan were obtained from experiments carried out with bees separated from their colony, and kept in cages or in small experimental colonies. Lifespan and Physiological Condition of Caged Bees In the experiments with caged bees attempts were made to clarify the following questions: (a) What effect has pollen-feeding on caged bees and are there any differences in the effects of pollen from various plant species? (b) Which organs of the bee depend directly on the food- composition? (c) Does a direct relationship exist between physiological condition and lifespan? (d) Which substances in the pollen are decisive in the activation of the physiological condition, and in the lifespan? The first investigations in this direction go back to the time when Czech workers (Soudek, 1927; Hejtmanek, 1933, 1943; Svoboda, 1940) observed that the pharyngeal glands of newly emerged bees could be brought to full development by feeding with pollen, soya flour and albumin. Subsequently, the present author proved (1946, 1954) that pollen-feeding of caged bees not only brought about an activation of the pharyngeal glands, but also caused a building-up of the fat body, the development of the ovaries, and a statistically proved longer lifespan. These findings were confirmed by other workers (Back, 1956, 1959; Bertholf, 1942; Beutler and Opfinger, 1949; Freudenstein, 1958; de Groot, 1953; Hess, 1942; Mussbichler, 1952; Pain, 1951a, b; Wahl, 1956), and it was also shown that pollen- feeding produces development of the wax glands (Freuden- stein, 1958). 236 Anna Maurizio These experiments further confirmed that pollens of dif- ferent plants can vary greatly in their effect on the bees. Thirty-five pollen samples were examined and these can be divided into many different groups. Depending on their biological effectiveness, they range from the very effective (e.g. Salix spp., fruit trees, Papaver spp., Zea mays, Crocus alhiflorus, Trifolium spp., Castanea sativa) to the almost ineffective, e.g. all seven species of conifer examined (Pinus cembra, P. silvestris, P, montana, Picea excelsa, Abies alba, Cedrus libani, Pseudotsuga taxifolia). No relationship could be established between the type of floral biology of any particular plant and the effectiveness of its pollen on bees ; on the other hand, pollen collected by bees w^as more effective than hand-collected pollen from the same plant species. It remains to be mentioned that bees are very selective in their choice of pollen sources, and this, according to Louveaux (1958), is connected with the nitrogen content of the pollen. These observations on bees were recently confirmed by ex- periments carried out with other insects (Tribolium larvae, Koch, 1952; Osmia lignaria, Levin and Haydak, 1958). These authors also established differences in effectiveness between pollen from different plant species and between hand and insect-collected pollen. Further evidence shows that the effect of pollen is time- conditioned, i.e. consumption must take place within the first 10-14 days of the bee's life in order to activate the pharyngeal and wax glands as well as the fat body. Pollen- feeding at a later stage has little or no effect (Beutler and Opfinger, 1949; Freudenstein, 1958; de Groot, 1953; Kratky, 1931; Maurizio, 1954). Data obtained from the author's experiments were statistic- ally interpreted in order to examine in greater detail the relationship between lifespan and physiological condition. On the basis of a multiple correlation (part-regression) it was established that there is a positive relationship between physiological condition and lifespan, i.e. that 56 per cent of Factors Influencing the Lifespan of Bees 237 the variations in lifespan can be traced to the physiological condition. The closest correlation was found between the development of the fat body and lifespan ; the least close was that between ovary development and lifespan. The result of this statistical evaluation confirmed the biological investiga- tions which showed that the main function of the fat body is the storing of protein, glycogen and fat, whereas the pharyn- geal glands not only serve as a store, but perform other important functions such as the secretion of larva food and the enzymes necessary for honey production. The functioning of the ovaries of the workers is mainly controlled by hormones and in the queenright colony the ovaries remain undeveloped (Butler, 1954, 1955, 1956; Pain, 1954, 1958; and Voogd, 1955). Thus, of all the organs examined, it appears that the fat body plays the most important role in the regulation of lifespan of the summer and winter bee. A further problem is posed in deciding which substances in the pollen are responsible for the activation of the physio- logical condition and lengthening of the bees' lifespan. Pollen is rich in carbohydrates, fats, proteins, vitamins and minerals, but the concentration of any one of these substances varies greatly from one plant species to another (Hay dak and Palmer, 1938, 1940, 1941, 1942; Haydak and Vivino, 1943, 1950; Hejtmanek, 1943; John, 1958; Kocher, 1942; Pearson, 1942; Sarkar et ah, 1949; Todd and Bretherick, 1942; Vivino and Palmer, 1944; Weaver and Kuiken, 1951; Weygand and Hofmann, 1950). For instance, the protein content of dif- ferent types of pollen can vary as much as 5 to 35 per cent, and the spectra of amino acids present may be variously composed. It was found that pollen contains the following vitamins: B^ (thiamine), Bg (riboflavin), Bg (pyridoxine), nicotinic acid and biotin. Moreover vitamin K, w^hich is not normally present in fresh pollen, was found in pollen obtained from comb cells. Bees have sufficient carbohydrates at their disposal in honey, thus pollen is their main protein and vit- amin source. Over the last few years many lively discussions 238 Anna Maurizio have taken place on the relationship between lifespan and physiological condition in the adult bee, and to what ex- tent it is influenced by the proteins and vitamins of pollen. It must be pointed out that this concerns only the vitamin requirements of the adult bee, as there are no doubts about the vitamin requirements of the brood. The results of the experiments carried out so far can be summarized as follows (Back, 1956, 1959; Freudenstein, 1958; de Groot, 1953; Haydak and Vivino, 1950, Koch and Schwarz, 1956; Maurizio, 1954; Miissbichler, 1952; Pain, 1951a, b). In the caged young bee lifespan and physiological condition are closely related to the protein content of food. Bees fed exclusively on devitaminized casein develop pharyn- geal glands, fat bodies and ovaries, and a statistically proved increase of lifespan is observed by comparison with bees of the same age fed on a pure sugar diet. An increase in nitrogen and in dry weight occurs in the young bee if it is fed from its emergence onwards on synthetic food — a mixture of sugar and pure amino acids (de Groot, 1953). Addition of any one of the vitamins (B^, Bg, Bg, nictotinic and pantothenic acids) resulted in no statistically proved increase in lifespan, but addition of vitamins to a protein-containing diet influenced the development of the pharyngeal and wax glands. Although young bees can develop their pharyngeal glands on vitamin- free casein food, the degree of development is enhanced if vitamins are added. Back's latest investigations contribute greatly towards a solution of this question. According to her, large doses of vitamins added to the casein food have the effect of shortening the lifespan. A very slight prolongation was observed when very weak mixtures of vitamins were added (mixture: Bj, Bg, Bg, A, C, E, K, nicotinic, pantothenic and folic acids, biotin and inosite). Young bees fed with vitamin-free casein food reared one generation of normal brood; bad rearing occurred only with the second generation of larvae. Young bees reared by nurses fed on a vitamin-free diet are unable to utilize the protein in their food and thus Factors Influencing the Lifespan of Bees 239 cannot rear a brood unless vitamins are added to their diet. In such vitamin-starved bees addition of pantothenic acid or a mixture of vitamins to the diet will restore normal brood- rearing within 24 hours. Apparently the newly emerged young bee is left with large body reserves of vitamins, thus making it possible to develop its pharyngeal glands, lengthen its lifespan and rear one brood-generation in spite of consuming a protein-rich but vitamin-free diet. It seems, however, that this vitamin reserve is not sufficient to pass on to the brood an adequate vitamin surplus; so that the second generation of vitamin- starved bees shows signs of vitamin deficiency, i.e. the development of the pharyngeal glands and the lifespan are affected. Similarly, a colony of bees kept on a protein-free diet are able to raise one generation of normal brood by using up their own protein reserves (Haydak, 1935). Discussion Present-day knowledge of the inter-efFects of nutrition, physiological condition and lifespan of the bee, and of the processes of life in the colony throughout the year, can be sum- marized as follows: (1) Pollen is the main source for protein, vitamins, and minerals, and therefore forms, with the sugar-containing juices, the basis of nutriment and development of the bee colony. Pollen-feeding of the young bee promotes the develop- ment of the pharyngeal and wax glands, the formation of a many-layered fat body and an increase in lifespan. Of all the pollen substances proteins are the most decisive in influencing the lifespan and physiological condition of the bee. For the adult bee vitamins in food are less essential as it still has a large store of vitamins from the larval phase. Absolutely essential is the extra vitamin supply at the time of brood- rearing. There is a statistically comprehensible correlation between physiological condition and lifespan. (2) The yearly cycle of life in the bee colony may be 240 Anna Maurizio described as follows. During the warm season the young bee feeds for the first few days of its life on pollen, developing its pharyngeal glands. With the commencement of brood- rearing the reserves are used up and the bee ages physio- logically, thus becoming a short-lived summer flying-bee. However, if for any cause brood-rearing is limited (e.g. due to a queenless colony, swarming, prolonged rain, etc.), or com- pletely omitted, the pharyngeal glands will remain fully developed, the fat body will be formed and the lifespan thus increased. In the free-flying colony during the active season it is possible to produce experimentally such a generation of physiologically young, long-lived bees. When brood-rearing is naturally limited and at the same time intensive pollen-feeding takes place, a long-lived winter bee with large body reserves, fully developed pharyngeal glands and a many-layered fat body is produced in late summer or autumn ; the bee is thus enabled to overwinter and to rear the first brood in spring. During the winter the bee feeds almost entirely on sugar and therefore it seems that the substances (protein, glycogen and vitamins) necessary to survive winter are drawn from the reserves built up in the fat body. A similar kind of overwintering is known of other solitary insects such as Anopheles, Culex, Lasiopticus, Epistrophe, Ips typographus, etc. (Buxton, 1935; Kuhn, 1949; Schneider, 1947, 1948; Wigglesworth, 1950). (3) It appears that the lifespan and physiological condition of bees in free-flying colonies depend greatly on nutrition and brood-rearing. Long-lived bees always appear in a colony where a rich supply of pollen is available and where little or no brood is in need of nursing. In our climate long-lived bees appear in autumn and winter, but under diff'erent climatic conditions this may occur at different times or fail to take place. Thus the physiological condition of the "summer" and "winter" bee does not depend on a certain season, but can be promoted at any time as a reaction to external conditions essential to life. Factors Influencing the Lifespan of Bees 241 The continuity of life in a bee colony depends upon an unbroken succession of generations. This succession is main- tained by the fluctuations between nutrition, brood-rearing, physiological condition and lifespan, giving the colony adaptability to the prevailing external conditions. The greater the brood nest is in proportion to the number of nursing bees, the larger is the new batch of young bees, but the shorter the lifespan of the individual insect. Restricted brood areas resulting in small numbers of young bees, on the other hand, mean long-lived bees. Both types of colony are well known to the practising bee-keeper. These types are partly genetically conditioned, it being probably less the longevity of the bees which is hereditary than the tendency to the laying down of large or small brood nests. [The editors would like to thank Mrs. M. Silverman for her work in translating this paper, and Mr. H. K. Airy Shaw for his expert advice.] REFERENCES Anderson, J. (1931). Bee World, 12, 25. Armbruster, L. (1931). Handb. Erndhr. Nutztiere, 3, 478. Back, E. (1956). Insect. Soc., 3, 285. Back, E. (1959). Personal communication. Bertholf, L. M. (1942). J. econ. EnL, 35, 887. Beutler, R., and Opfinger, E. (1949). Z. vergl. Physiol., 32, 383. Butler, C. G. (1954). Bee World, 35, 169. Butler, C. G. (1955). Amer. Bee J., 95, 275. Butler, C. G. (1956). Proc. roy. Soc., A., 31, 12. Buxton, P. A. (1935). Parasitology, 27, 263. El-Deeb, a. L. a. (1952). Amer. Bee J., 92, 517. EvENius, C. (1937). D. Imkerfuhrer, 11, 128. Farrar, C. L. (1936). Amer. Bee J., 76, 452. Farrar, C. L. (1949«). Bee World, 30, 51. Farrar, C. L. (19496). In The Hive and the Honeybee, ed. Grout, R.A. Hamilton, Illinois: Dadant. Farrar, C. L. (1952). J. econ. Ent., 45, 445. Freudenstein, K. (1924). Arch. Bienenk., 6, 49. Freudenstein, H. (1958). In XVII Congr. int. Apia. Gontarski, H. (1950). Imkerfreund, 11, 183. Gontarski, H. (1954). Z. Bienenforsch., 2, 161. Groot, a. p. de (1953). Dissertation, Utrecht. Haydak, M. H. (1934). J. agric. Res., 49, 21. 242 Anna Maurizio Haydak, M. H. (1935). J. econ. Ent, 28, 657. Haydak, M. H. (1937a). J. agric. Res., 54, 791. Haydak, M. H. (19376). Ann. ent. Soc. Amer., 30, 258. Haydak, M. H. (1939). J. econ. Ent, 32, 663. Haydak, M. H. (1940a). J. econ. Ent., 33, 397. Haydak, M. H. (19406). Glean. Bee Cult., 68, 615. Haydak, M. H., and Palmer, L. S. (1938). J. econ. Ent., 31, 576. Haydak, M. H., and Palmer, L. S. (1940). J. econ. Ent., 33, 396. Haydak, M. H., and Palmer, L. S. (1941). J. econ. Ent., 34, 37. Haydak, M. H., and Palmer, L. S. (1942). J. econ. Ent., 35, 319. Haydak, M. H., and Tanquary, M. C. (1943). Bull. Minn, agric. Exp. Sta.y 160, 1. Haydak, M. H., and Vivino, E. (1943). Arch. Biochem., 2, 201. Haydak, M. H., and Vivino, E. (1950). Ann. ent. Soc. Amer., 43, 361. Hejtmanek, J. (1933). des. VMaf, 10. Hejtmanek, J. (1943). Die PoUenverdauung der Honigbiene, Ustr. slov. vcel. spolkov, Prievidza. Hess, G. (1942). Beth. Schweiz. Bienenztg., 1, 33. John, M. (1958). Z. vergl. Physiol, 41, 204. Koch, A. (1952). Tijdschr. Ent., 95, 166. Koch, A., and Schwarz, J. (1956). Insect. Soc, 3, 213. KocHER, V. (1942). Beih. Schweiz. Bienenztg., 1, 155. KoEHLER, A. (1921). Schweiz. Bienenztg., 57, 424. Kratky, E. (1931). Z. wiss. ZooL, 139, 120. KuHN, W. (1949). Mitt. Schweiz. ZentAnst. Forstl. Versuchsw., 26, 245. Levin, M. D., and Haydak, M. H. (1951). J. econ. Ent., 44, 54. Levin, M. D., and Haydak, M. H. (1958). X Int. Congr. Ent., 4, 1079. Lindauer, M. (1953). Bee World, 34, 63, 85. Lotmar, R. (1939). Landw. Jb. Schweiz., 53, 34. LouvEAUx, J. (1958). These, Paris. Mauermayer, G. (1954). Arch. Bienenk., 31, 31. Maurizio, A. (1946). Beih. Schweiz. Bienenztg., 2, 1. Maurizio, A. (1954). Landw. Jb. Schweiz., 68, 115. Moskovljevic, V. (1939). Schweiz. Bienenztg., 75, 689, and Bee World, 21, 39. Mussbichler, a. (1952). Z. vergl. Physiol, 34, 207. Nickel, H. K., and Armbruster, L. (1937). Arch. Bienenk., 18, 257. Nickel, H. K., and Armbruster, L. (1938). Arch. Bienenk., 19, 134. Pain, J. (1951a). Arch. inU Physiol, 59, 203. Pain, J. (19516). C. R. Soc. Biol (Paris), 165, 1505. Pain, J. (1954). Insect Soc, 1, 59. Pain, J. (1958). Apiculteur, 103, 1. Pearson, P. B. (1942). Proc Soc. exp. Biol (N.Y.), 51, 291. Phillips, E. F. (1922). J. econ. Ent, 15, 368. Phillips, E. F. (1928). Beekeeping. New York: Macmillan Co. Ribbands, C. R. (1952). Proc. roy. Soc B., 140, 32. Ribbands, C. R. (1953). The Behaviour and Social Life of Honeybees. London : Bee Research Association Ltd. Factors Influencing the Lifespan of Bees 243 RocKSTEiN, M. (1950a). Ann. ent. Soc. Amer., 43, 152. RocKSTEiN, M. (19506). J. cell comp. Physiol, 35, 11. RocKSTEiN, M. (1953). Biol. Bull., Wood's Hole, 105, 154. RoscH, G. A. (1925). Z. vergl. Physiol, 2, 571. SakagAxMI, S. F. (1953). Jap. J. Zool, 11, 117. Sarkar, B. C. R., Wittwer, S. H., Luecke, R. W., and Shell, H. M. (1949). Arch. Biochem., 22, 352. Schneider, F. (1947). Mitt, schweiz. enl Ges., 20, 306. Schneider, F. (1948). Mitt, schweiz. enl Ges., 21, 249. SouDEK, S. (1927). Bull Fac. Agric. Brunn, 10, 1. SvoBODA, J. (1940). Schweiz. Bienenztg., 76, 206. Todd, F. E., and Bretherick, O. (1942). J. econ. Enl, 35, 129. ViviNO, E., and Palmer, L. S. (1944). Arch. Biochem., 4, 129. VooGD, S. (1955). Experientia, 40, 181. Wahl, O. (1956). Sudwestdtsch. Imker, 8, 348. Weaver, N., and Kuiken, A. (1951). J. econ. Enl, 44, 635. Weyer, F. (1931). Z. Zellforsch., 14, 1. Weygand, F., and HofxMann, H. (1950). Chem. Ber., 83, 405. Wigglesworth, V. B. (1950). The Principles of Insect Physiology, 4th ed. London: Methuen. Zantder, E. (1947). Das Leben der Biene, 5th ed. Stuttgart: Ulmer. Zander, E. (1951). Der Bau der Biene, 4th ed. Stuttgart: Ulmer. DISCUSSION Wigglesworth: Have you considered the further complication that building up of these reserves may not be a direct effect of the various nutrients, but an effect through the endocrine system? For example, it has long been known that in many adult insects deprived of the corpus allatum the eggs do not develop and do not build up yolk. There is the same effect in protein starvation, where the same histological eff'ects appear. A. S. Johansson (1958. Nature (Loud.), 181, 198), working on the milkweed bug, showed that if an active corpus allatimi is implanted in the starved insect there is normal egg development. In other words, the immediate effect of protein feeding seems to be to activate the corpus allatum. The prime deficiency in a protein-starved insect is inactivity of the gland. In your experi- ments it would not necessarily be the corpus allatum; it might be the neurosecretory cells or some other component in the endocrine system. Hinton: In bees doesn't the queen substance inhibit the develop- ment of the corpora allata and keep their volume small? Maurizio: The degree of development of the ovaries of the honey- bee worker is directly connected with the endocrine system and the queen substance (see references to Butler, Pain and Voogd). It is 244 Discussion possible that the development of the pharyngeal glands and the fat body also come under endocrine influence but, as far as I know, this has not yet been examined closely in the honey-bee. Rockstein: Dr. Maurizio did imply in her paper that it was a matter of endocrine control. I personally think that, generally, the ultimate controlling influence will be found to lie in the endocrine system. I do not think that Dr. Maurizio said that vitamins or proteins were involved in this physiological senescence. She did indicate that there was a depressed development of the pharyngeal gland, which suggests hormone control. Comfort: Is the rate of loss of nerve cells different in summer and winter bees? Rockstein: At the two levels of the brain which were counted there were about 500 cells in the young bees and only about 325 in the old bees, regardless of whether these were indoor (summer) bees without a queen, or outdoor (summer) bees which had been living a normal life in the hive. Comfort: In other words the difference in lifespan does not depend on the difference in the rate of loss of cells? Rockstein: Yes, it does, because the old bees in both conditions showed precisely the same degree of loss. The outdoor bees I had were only seven weeks old, whereas, indoors, I had bees still living at ten weeks of age. For one thing, my bees were exposed to light 24 hours a day. According to Dr. Maurizio the activity may have nothing to do with age, but nevertheless I could not keep them alive any longer under these conditions even on a completely unlimited diet. I do think that the similarity in development of the overwinter- ing or the summer-broodless bees is very significant. It points to an entirely different phenomenon from that in the caged bees — namely that these summer, "queenright" animals normally are not able to attain their potential longevity owing to the demands of brood- rearing upon their nutrition. As soon as you eliminate this really exorbitant nutritional dernand, by removing the queen, the bees are able to attain a lifespan of something like six to eight months. As Prof. Wigglesworth suggested and Dr. Maurizio implied, the diet does this, but where it produces its primary effect is not known. Most probably it is through the endocrine system ; certainly, that is evident in the effect upon the development of the pharyngeal glands and of the fat body, which must be controlled at a biochemical level. The most plausible explanation at a biochemical level is that of altering hormone control. Nigrelli: What would happen if you did not feed the caged bees? Maurizio: They would be dead of starvation in two or three days. Discussion 245 Nigrelli: Have they no capacity for food storage? Maurizio: They have proteins but no sugar. Wigglesworth: As I understand it, the queenless worker develops ovaries but the winter bee does not. That would suggest that there is a more subtle difference in addition to the direct nutritional effect. Maurizio: The queen substance is there in the overwintering colony. Wigglesowrth : So there is a dual effect. Hinton : Do the workers develop ovaries whatever time of the year you remove the queen? Maurizio: Yes. Kershaw: There is evidence in blood-sucking flies that they take blood under hormonal influence, and that there is an antibiting hormone (Lavoipierre, M. M. J. (1958). Nature (Lond.), 182, 1567). This may be a similar phenomenon. Maurizio : We will probably ask Dr. Butler to do this experiment with the queen substance. Bourliere: Have you made any determination of oxygen consump- tion in winter bees as compared with summer bees? Maurizio: That was done by Corkins and Gilbert (Corkins, C. L. and Gilbert, C. S. (1932). Bull. Wyo. agric. Exp. Sta., No. 187, 1). They found that carbon dioxide output at a hive temperature of 4° was 52-62 per cent of the output at 16°. Sacher: Is the difference between summer bees and the over- wintering bees in the colony determined directly by nutrition and activity, or is there a change in the behaviour of the colony which is caused by the environment and which in turn leads to these differ- ences? Maurizio: The winter bees are inactive; there is no brood, they cannot fly and they stay in the colony in a cluster. The only function is to keep warm in the cluster. Sacher: Then the fact that it gets cold and they cluster in this fashion and cease their brood rearing is what leads to this change in survival. Maurizio: But the bees begin to brood in February in Switzerland. It depends on the climatic conditions. Hinton: There must be some indirect effects here because there are many long periods in the winter which are just as warm. Maurizio: In Northern Europe bees are unable to forage for four or five months in winter, and the brood-interval is much longer than in England and Central Europe. Rockstein: There may, however, very well be a light-dependent factor, such as a diurnal or photoperiod effect. Thirty minutes 246 Discussion before sundown, for example, most foraging bees will start to come back to the hive, even on a hot summer day, so it is not a matter of temperature. Holt: What is the mode of life of tropical bees? Maurizio: I think that in the tropics there are no winter bees, because they can fly all the year. They have pollen and nectar, they seem to have brood the whole year and there are no long-lived bees in the colony. But a proper study of bees in the tropics has yet to be made. Hinton: That would be primarily an adjustment to food supplies and not necessarily to temperature. Maurizio: It is complex. In the tropics there may be no over- wintering, but during the two or three months of the dry season there may be long-lived bees in the colony. Rocksiein: There is probably more than one factor, such as light or temperature, that would produce the end-result of conserving the colony. In Minnesota, bees begin brood-rearing in January when temperatures are sometimes at —30° f. These animals are being continually decimated during the winter months because some of them have emerged in August, some in September and some in October, and we are going to end up with a very small nucleus from the last-laid eggs by the end of winter. There appears then to be a social pressure upon the hive which stimulates brood-rearing late in the winter. Somehow, the queen is aware of this, even in the dead of winter. As far as light is concerned, the shortest day is in mid- January in Minnesota, but the queen begins to lay eggs again at that time. Hinton: What is the temperature in the cluster in winter in Minnesota? Rockstein: The centre of the cluster is always maintained at a temperature of about 33° c. This has been established by thermo- couple measurements. The bees on the outside of the cluster change places with bees from the inside; otherwise those on the outside would fall torpid from cold. These outside bees must feed, generate heat, enter and allow the inner bees to move out. If the cluster gets too far away from the food stores, the whole cluster will die, with the queen the last one to go. Hinton: Thus social insects can be said to be homoio thermic. Rockstein: I would rather say that the colony as a whole can be said to be homoiothermic. THE BIOLOGY OF AGEING IN INSECTS Morris Rockstein Department of Physiology, New York University College of Medicine t New York For well over a decade the present author's interest in ageing has been that of establishing firm, quantitative bio- chemical criteria for physiological ageing, particularly in structures manifesting senescence in the form of declining function as well as degenerative anatomical alterations with advancing age. As a direct consequence of these physiological studies, longevity data have been obtained for large numbers of individuals for each of two species of insects and for strains of similar genetic constitution in each case. This discussion will present details of the lifespan of the common housefly, Musca domestica L., and the honey-bee, Apis mellifera L., with special emphasis upon recent studies directly concerned with the factors affecting the lifespan of male and female houseflies. In an earlier series of experiments, involving a comparison of change in cell number with alteration in cholinesterase activity with age, in the adult worker honey-bee (Rockstein, 1950a), the time of final samplings of living animals was determined by the availability of sufficient numbers of bees for enzymological and histological evaluation. In all cases this occurred when one per cent or less of the original number of bees employed were still living, and represented values very close to the maximum lifespan for this strain of Italian golden honey-bees. Thus, for 3,750 queenless worker bees, maintained indoors in large cages supplied with honey, pollen and water in excess and exposed to artificial lighting day and night, the final sampling was made at 68 days. A similar value of ten 247 248 Morris Rockstein weeks was obtained for approximately 4,000 of the same strain of honey-bee maintained indoors, in a more recent study involving changes in alkaline and acid phosphatase in ageing bees (Rockstein, 1953). For a similar study of 2,700 "queen- right" bees, marked with coloured lacquer immediately after emergence and returned to the hive to engage in normal hive activities (during the summer months), the last sample of bees had to be taken at 51 days of age. These values compare well with observations of apiculturists and other students of the biology of the honey-bee (see Rockstein, 1950fc). Indeed, Dr. Maurizio's own studies (1954) include data for two strains of Italian bees of maximum lifespans of 54 and 62 days, respectively, maintained in the hive during summer months. In the earlier studies (Rockstein, 1950a), the number of neurones at two distinct levels of the honey-bee brain was em- ployed as an anatomical criterion for biological old age; the absolute number of cells remaining in the brains of old bees (as well as the percentage loss from emergence to old age) was remarkably similar (325 ; 350) for both kinds of old bees, whether they were living the normal lives of the hive bee or were maintained indoors in a small queenless colony under the conditions described. This loss of about 35 per cent of the original number of brain cells in the adult worker bee is singularly similar to that of a 35 to 40 per cent loss in mam- malian brain cell number reported for humans by Hodge (1894), Ellis (1919, 1920), Andrew (1938) and Gardner (1940) and for the white rat by Hatai (1902) and by Inukai (1928). From a recent study of the decline with age in the activity of enzymes concerned with the energizing of flight activity in the common housefly, from emergence to senility, longevity data have been obtained for thousands of male and female flies of the NAIDM standard laboratory strain of houseflies, which had been intensively inbred for more than one hundred generations. In our laboratory these animals are reared and maintained on a standardized laboratory diet in an air- conditioned room kept at 80° f and 45 per cent relative The Biology of Ageing in Insects 249 humidity. Except for specific experiments designed to test the role of parental age in determining the longevity of the offspring, all flies were reared from eggs laid by parents as soon as they were capable of oviposition, i.e. between the fourth and fifth days of adult age. A well-regimented strain under the conditions of rearing and maintenance described, these flies emerge as adults exactly two weeks following the emergence of adults of the previous generation and young females begin laying eggs on exactly the fourth day after they have reached the imaginal state. Longevity and diet During the course of collecting adult male and female house- flies for biochemical study (Rockstein, 1956), it was observed that there were relatively fewer and fewer males available for enzyme determinations, especially by the end of the second week. Thus, from a sex ratio of one to one, the male to female population composition fell to a one to two ratio by the end of two weeks and to a less than one to three ratio by the end of the third week. A pilot follow-up study was made of six cages of about 125 flies each in which mortality (rather than survival) counts were made; the results (Rockstein, 1957) clearly confirmed the fact that male longevity was consider- ably smaller than that of the female housefly. However, in both of these cases, although the larvae were reared on a standard laboratory medium of powdered whole bovine milk (KLIM, Borden's) dried brewer's yeast and agar, the adults had been maintained on sucrose and water alone (in order to eliminate extraneous factors of diet and egg-production in the females particularly, in the age-dependent enzyme study). Under such conditions of restricted diet, it was thought that the observed sex-related diff'erential in longevity, favouring the female of this species, might have resulted from the pos- sible availability of nutrient reserves (such as oocytes or fully developed ova in the ovarioles of the young female at emerg- ence) to individuals of that sex exclusively. Figs, la and lb 250 Morris Rockstein graphically illustrate the data obtained for two sets of ex- periments involving a total of 600 males and 600 females maintained as adults on sucrose and water, as before, and 600 males and 600 females maintained on a diet of sucrose, water and powdered whole milk. (The latter is the normal diet for stock breeding adult houseflies.) As is clearly seen in Fig. la, 00 — 90 - ° • 80 • 70 — 60 c • 50 . <« 40 30 - • KUM-CASEI ^ •% • KUM- CASe a 20 ~ ,m oSUGAR-CASei o 'SUGAR' CASE n 10 • •o 1 1 1 _L o o q: 0 10 20 30 40 50 AGE IN DAYS Fig. la. The effect of diet on male houseflies. Reprinted from Rockstein and Liebennan (1959), by courtesy the Editor, Gerontologia. no appreciable advantages accrued to the longevity of the male houseflies from inclusion of this (high protein, lactose, butter-fat and mineral-containing) adult dietary component. On the other hand, female flies, with a greater longevity than males even under conditions of restricted diet (sucrose and water alone), showed a considerable prolongation of lifespan as a result of including KLIM in the adult food, beginning with The Biology of Ageing in Insects 251 the second week of life (Fig. lb). As can be seen from Table I (reprinted from Rockstein, 1957), the mean longevity for a total of 600 female flies maintained on the enriched diet was about 31 days, in contrast to a mean value of about 19-5 days for females maintained on sugar and water alone. Moreover, the maximum longevity for females on the KLIM-enriched CO Uj —I < Uj U. >■ o Uj U q: 00 Q o** .•••• 90 o O ° • •• 80 ~ o , • o ° • 70 o • • 60 _ • •• *• 50 - o * • O * 40 — • *• 30 • • — • • • • •KLIM-CASE I ,.' ' KLIM-CASEU 20 "^ y ° SUGAR-CASE I ° v» • SUGAR-CASE IC '•°. 10 0 10 20 30 40 50 AGE IN DAYS 60 70 Fig. lb. The effect of diet on female house- flies. Reprinted from Rockstein and Liebennan (1959), by courtesy the Editor, Gerontologia. diet was about 30 per cent higher than for females on the restricted diet. These results are in direct contrast to the well-known findings of McCay and his co-workers (1935, 1939, 1941) that restricting the dietary intake immediately or soon after weaning retarded the growth of male rats and concomitantly extended the average lifespan from 483 to 894 days (and maximum 252 Morris Rockstein longevities from 927 to 1,306 days). Female longevities, on the contrary, were relatively unaffected by such alterations in diet. It would appear from their results that retardation of growth in the male rat by such a dietary restriction eliminates the sex differential in longevity, normally favouring the female rat. However, McCay emphasized the low calorie aspect of his restricted diet, which was in reality a high protein, salt and vitamin diet. It is therefore likely that the Table I Effect of diet on longevity of male and female house flies WUh KLIM Sugar and water only With KLIM Sugar and water only Males Fe- males Males Fe- males Males Fe- males Males Fe- males Average longevity (days) 17-5 32-7 15-6 19-5 14-4 30-8 14-6 19-7 Maximum longevity (days) 44 64 40 50 40 56 29 35 critical factor involved in such a restricted diet might well have been the effective protein content of the diet. Indeed, the ingestion of such a "low calorie" diet might mean the utiliza- tion of the protein in this diet to meet the basal energy requirements (in the absence of carbohydrates and fats customarily utilized by the body in energizing processes). There would result an unavailability of protein required for normal cell growth, replacement and addition and, therefore, retardation of normal body growth and maturation. This (and evidence to follow) suggests that there may exist for each species a specific protein-calorie optimum for normal growth The Biology of Ageing in Insects 253 and maturation. Indeed, an early paper by McCay and Crowell (1934) reported that reduction in the protein content of the diet fed to trout doubled their longevity. In insects too, there appears to be an optimal dietary level of protein for growth and development and indirectly for lifespan. Thus in three species of cockroaches (all of which are not fully grown or mature at the onset of the final adult stage), Haydak (1953) reported specific optimum dietary protein requirements both for development and survival of Table II Effect of diet on the development and longevity of three species of cockroaches (after haydak, 1953) Effect of diet upon : Optimum protein content of diet P. americana B. orientalis B. germanica Minimal nymphal mortality : Shortest development period 49%-79% 22%-24% 22%-24% Greatest average adult longevity 22% 11% 11% the nymphal instars as well as for longevity of the adult. Table II, prepared from Haydak's data, shows that for the American cockroach the optimum protein for shortest nymphal development and associated lowest nymphal mortality was a broad range of 49-79 per cent; for either the Oriental or German cockroach, this was a much lower 22-24 per cent. For maximum adult longevity, on the other hand, the optimum protein content of the diet was 22-24 per cent for the adult diet of the American cockroach and 11 per cent for both the Oriental and German cockroaches. Haydak con- cluded from his data that for adults the heaviest protein 254 Morris Rockstein eaters had the shortest lifespans and that the total lifespan was inversely proportional to the protein intake from eclosion to death. Maurizio's work has also demonstrated the importance of protein (and possibly vitamins) in the longevity of adult worker honey-bees (1954). She has shown that captive worker honey-bees which are fed pollen early in their adult- hood resemble overwintering bees in two major respects; they are longer-lived than similar bees rearing brood and, secondly, their "physiological state", i.e. well-developed fat body and pharyngeal glands maintained for a longer time, is typical of overwintering bees (which may live as long as six to eight months in contrast to six to eight weeks for maximum summertime longevities). Thus, retardation of ageing in the worker bee is associated with adequate pollen (protein and vitamin) intake at an early adult age, in the face of reduced demands upon such food reserves, such as occurs in over- wintering or non-brooding "summer bees". This in turn results in slow build-up and extended maintenance of those structures upon which the prolongation of adult life is significantly dependent. In all the cases mentioned above it appears that, where growth or maturation has not yet been completed, an optimum level of food (perhaps protein, specifically) intake is necessary to complete that development. Restricting the diet, as in the case of the young male white rat, trout and at least three species of cockroaches, delays the attainment of adult form and con- comitantly defers the cessation of growth and therefore senescence. In the case of holometabolous insects, like the housefly or the honey-bee, the situation is quite different. These animals are essentially fully matured within a few hours to several days, respectively. Heavy demands upon the food (protein) reserves of the female such as oviposition in the case of the housefly (Rockstein, 1958) and brood-rearing in the case of the worker bee shortens the lifespan in these animals. One can perhaps speak of a "negative protein The Biology of Ageing in Insects 255 balance" operating against the attainment of the longevity potential of a particular sex for a particular species, when the diet is inadequate for the physiological needs of such animals. Rockstein (1958) suggested that food reserve in the ovaries might indeed be a factor important in higher female longevity even during starvation. In this connexion, a paper by Grosch (1950) has shown that starving female wasps, Habrohracon juglandis (Ashmead), draw upon the reserves of the ovarioles, as evidenced by the gradual resorption of ova from the ovari- oles which is especially marked during the latter days of their lives. Woke, Ally and Rosenberger (1956) further support this idea of the ovaries as a source of nutrition in starvation for the female mosquito, Aedes aegypti L., in the observations that delaying the first blood meal or decreasing the size thereof lowers total egg production markedly. In the Levant housefly, Musca vicina Macq., Ascher and Levinson (1956) have also found protein essential to the adult diet for ovi- position and cited similar evidence for other species of muscoid adults. However, the common housefly does lay eggs, albeit much more spottily, even on a protein-free diet, but no information is available as to the number and viability of such eggs. V Longevity differences and sex In order to obtain further insight into the differences between male and female longevities in the housefly, a study on a much larger scale was undertaken. Fig. 2 (taken from Rockstein and Lieberman, 1958) shows the survival curves for about 8,500 flies of both sexes of the same NAIDM strain extending over nine generations, reared and maintained on a complete diet under the controlled laboratory conditions described earlier. Both survival curves show a rectangular character typical of animal populations manifesting sene- scence, namely, a very low mortality rate during the early days of the cohort's existence, and a middle period of rapid dying off. However, during the final five to ten per cent of the 256 Morris Rockstein population's existence the survival curves for both sexes show a logarithmic-like character (in the extremely retarded rate of dying off of this small proportion of long-lived members of the population). In actual fact, for the males about 90 per cent die off during the period of ten to thirty days after 100 90 . 80 70 I 60 a. ^ 50 I- a 40 a. UJ ^ 30 20 r\. • MALES o FEMALES AGE IN DAYS Fig. 2. Survival curves of male and female houseflies. Reprinted from Rockstein and Lieberman (1958), by courtesy of the Editor, Nature, "senescence" emergence as adults; for females this period of occurs during the longer period of ten to fifty days of adult Ufe. One striking feature of the particular colony studied is the remarkable homogeneity of the population, as regards longevity at least. Thus, both for the total 600 flies per sex, The Biology of Ageing in Insects 257 from the earlier study on diet (Rockstein, 1957), and the more recent total of over 4,000 flies per sex, fifty per cent mortality occurred at 16 days for the males and at 30 days for the females, respectively. The mean longevity values for males of about 17 days and for females of about 29 days, for this large 0.01 I — 0.05 0.1 02 0.5 I 2 95 98 99 99.5 99.8 99.9 • MALES o FEMALES / o» J I I I I 1 1 II J L 2 3 4 5 6 78910 20 30 40 50 6070 80 LOG TIME Fig. 3. Log-probit plot for male and female housefly survival. Reprinted from Rockstein and Lieberman (1959), by courtesy of the Editor, Nature. number of flies, also compare favourably with those reported in 1957. The close approximation of mean and medium longevity values is also a strong indication of a good natural distribution in so far as the populations of the two sexes in this strain are concerned. A log-probit plot of these data (probability of survival AGEING — V — 9 258 Morris Rockstein plotted against time), shown in Fig. 3 (Rockstein and Lieber- man, 1959), portrays in more dramatic fashion the actual course of the survival (or, conversely, the mortality) trends of Fig. 2, above. From the similar flat slopes of the log-probit curves of both males and females during the first ten days, a similar low mortality rate is apparent. After the tenth day, however, a marked shift upward of the male curve, which con- tinues as a straight line through the fortieth day, represents a high and uniform mortality rate during 93 per cent of the male population's existence. Thus a single (or at least relatively simple) mortality factor appears to be operating in the great majority of the male houseflies of this strain. In the females, however, one sees that the same log-probit curve follows a more irregular course, with a slight upward trend at the tenth to 21st days, a steep shift upward at the 21st day, another more gentle upward rise in slope at the 30th, and a final shift upward at the 40th day, which persists unchanged as a steep, straight line from the fortieth to final day of the female population's existence. One might infer from this kind of plot that, for the female, mortality (and therefore survival) is a much more complex phenomenon, with more factors for mortality becom- ing effective w4th increasing age. These inferences have been further substantiated in the life-tables which have been recently completed by Rockstein and Lieberman (1959) from mortality data for this population of male and female flies. Despite the wide distribution of the ubiquitous housefly, only a few other studies have been made on its longevity. The study of Wilkes and co-workers (1948) employed the Peet-Grady method (in which some of the dietary components are not standard) and obtained average longevities of 12 days for males and 20 days for females of a laboratory strain and about 12 days and 24 days, respectively, for male and female flies of a wild strain, kept at 80° f and 50 per cent relative humidity. Aside from dietary variation, the authors also describe major difficulties of overcrowding, especially in the larval stage and also for adults. In a recent letter, Rollins (1959) has supplied The Biology of Ageing in Insects 259 some interesting data on differences in longevity between the sexes for over 2,000 males and females of a 15-year inbred strain of normal houseflies, originating from a wild strain collected in the Sacramento, California, area, and reared and maintained on a Peet-Grady medium at 80° f. For male houseflies, he obtained a 50 per cent mortality at 11 days (as opposed to our values of 16 days) and for females at 29 • 5 days (very much like our own data for females). Thirty-day mortalities of 98 per cent for males and 52 per cent for females corresponded very closely to those obtained for our own NAIDM strain. For the Levant house-fly, Musca vicina Macq., however, Feldman-Muhsam (1944) and Ascher and Levinson (1956) found no significant difference in longevity between the two sexes. Parental age In an attempt to establish the possible effect of parental age of houseflies at oviposition upon the longevity of the offspring, a preliminary investigation was made according to the procedure of Lansing (1947, 1948, 1954). For standard breeding and in all previous studies, eggs were collected from parents at the youngest possible age (at the fourth to fifth day after emergence). In this series of experiments, however, 225 eggs each were collected at the fourth, sixth, ninth, 15th, 23rd and 27th days and the offspring otherwise reared and maintained as before on complete diets. Table III shows that the female offspring longevity is progressively diminished with advancing age of the parents at the time of oviposition ; e.g. survival, as expressed as 30-day mortality for female offspring, falls from 50 per cent mortality for eggs from young parents, to 92 per cent mortality for eggs laid at 27 days of parental life. The 30-day mortality data for male offspring, on the contrary, were interpreted as meaning that male offspring longevity was unaffected by increasing parental age. Values also included in Table III show corresponding declines in average longevities for females from about 32 days for 260 Morris Rockstein young parents to 22 days for females from very old parent flies. In the case of Lansing's parthenogenetically reproducing rotifers, his "cumulative reversible aging factor" could be directly attributable to cessation of growth in the old female parent. In the present study, however, several interpreta- tions were possible for the sexually reproducing housefly; Table III Effect of parental age of houseflies on longevity of offspring Parental age (daifs) Males Females % Mortality % Mortality Average longevity 4 95 50 32 days 6 93 52 29 days 9 97 70 28 days 15 92 82 25 days 23 97 80 24 days 27 90 92 22 days either or both of the old parents might be contributing to the reduced female offspring longevity which is associated with the advancing age of the parents at oviposition. A second possibility is that the observed data represent the adverse effect of long-term storage of spermatozoa in the female spermatheca. With Dr. Lieberman, an expanded study of the possible r61e of parental age in the longevity of the housefly The Biology of Ageing in Insects 261 was undertaken last year, with four types of matings as follows: young males by old (about 29 days) virgin females, old (about 23 days) males by young virgin females, old males by old virgin females, and "modified old-old crosses" (in which, as in the original study of parental age, males and females were allowed to mate freely from emergence, but where eggs were collected only from parents at an advanced age). A fifth set of cages involved the usual young-by-young matings employed in routine stock breeding and in other studies in which parental age was not a variable ; this was the control series. Offspring longevity data were obtained for over 1,000 flies of each sex for each type of cross, for three to four generations in each case, in order to test the possible presence of a Lansing-like, cumulative ageing factor related to parental age, in the housefly. The results obtained indicate that the slightly lowered 30-day mortality for males from oldest parents, seen in our earlier study (Table III), might indeed have been significant. Statistical analysis of data from our current study indicates that, for two generations, the average male longevity (20-6 days) for offspring from " modified old-old parents" is higher, by three days, than the mean longevity for young-by-young crosses (P is less than 0-01). That this result may be due to the effect of the female parent upon male offspring is supported by the observation that a similar increase in male offspring longevity was obtained for crosses between young males and old virgin feynales and for two cases of crosses between old males and old virgin females. Thus, the mechanism involved may be one of selection of a long-lived strain of male offspring from the longer-lived female parents surviving to almost 30 days of age and still sexually functional at that advanced age. As for the effects of parental age on female offspring longevity, data obtained in this more recent and extensive study suggest that the parental ageing effect on this sex is more complex: it is considered essential at this point to continue these studies with single- pair crosses in order to follow the longevity of the adults as 262 Morris Rockstein well as of their offspring on an individual, rather than on the group population basis employed in all studies to date. In Drosophila suhohscura Comfort (1953) found no apprecia- able change in longevity of the population resulting from selection of eggs from old parents, over eight generations of breeding. Tracey (1958), on the other hand, found that larvae of eggs from week-old adult beetles of the mealworm, Tenebrio molitor, had longer developmental periods than those from eggs laid by five-week-old (middle-aged) beetles and consider- ably longer than those from eggs laid by very old (nine weeks) parents. The adult lifespan was likewise shortened for beetles reared from eggs laid by oldest parents. Thus, as the parental age lengthened, the entire lifespan of offspring was shortened in this species. Although the above held true for animals reared and maintained at 25°c, at 30° c some larvae from oldest parents had much longer larval development periods than those from younger parents ( ! ). However, increasing the temperature to 30° c did not alter the shorter adult lifespan observed for offspring from oldest parents. What role the age of parents plays in the lifespan of children in humans is not statistically clear, although Murphy (1954) reported that known abnormalities like congenital cardiac im- pairment, Mongolian idiocy, spina bifida and stillbirth, and possibly hairlip and cleft palate, increase in frequency as the age of the mother at the time of birth increases, after 30 years of age. However, Sonneborn (1957) has pointed out that the age of the mother is frequently positively correlated with the paternal age, for foetal deaths. Indeed, for data covering foetal deaths and a 10 per cent randomized sample of 330,000 live births from 1954-55 New York City Department of Statistics, he found there was a consistently higher foetal death rate for older fathers, when single rather than five-year age classes were employed for females over 30 years of age. The Biology of Ageing in Insects 263 Summary Queenless, Italian golden worker honey-bees, Apis mellifera, maintained indoors on an excess of honey, pollen and water, lived to a maximum of about 9 • 5 weeks ; in a second study of over 4,000 bees a similar maximum value of ten weeks was obtained. A maximum longevity of 7 • 5 weeks was found for over 2,700 bees, properly marked and returned to the hive to perform normal hive activities. As a criterion of old age, the number of brain cells at two distinct levels of the adult honey- bee brain was found to be virtually identical for both indoor and outdoor (hive) senescent bees. In a study of over 8,500 houseflies, Musca domestica, of the NAIDM strain, females were found to have a mean longevity of 29 days and males a mean longevity of 17 days. Curves for probit-log time plots indicated a relatively simple mortality factor for the male population but a complex of several mortality factors for the female cohort. Female longevity was enhanced by the inclusion of powdered whole milk in the adult diet of sugar and water. No such beneficial effect was obtained for male houseflies. A study of the role of parental age at the time of oviposition indicated a possible adverse effect on the longevity of female offspring from oldest parents ; for males reared from eggs from oldest parents there appears to be an enhancement of the mean longevity, probably by a selection of long-livedness through the old surviving female parents. Acknowledgements Figs, la and lb were drawn by Miss Mary Lorenc. Fig. 3 was drawn by Mr. Rudolph Cavalcante. The author is deeply appreciative of the assistance of Dr. Albert S. Perry of the USPHS Technical Development Laboratories, Savannah, Georgia, for generously supplying the pupae of the NAIDM strain from which this colony has been developed. The kind assistance of Miss Gertrude Uhr, Secretary of the Depart- ment of Physiology, and Mrs. Elaine S. Rockstein, in the preparation of the manuscript, is acknowledged. 264 Morris Rockstein The research described in this report was supported in part by the United States Army Surgeon General's Office and the United States PubHc Health Service Division of Research Grants. REFERENCES Andrew, W. (1938). Z. Zellforsch., 28, 294. AsciiER, K. R. S., and Levinson, Z. H. (1956). Riv. Parassit, 17, 217. Comfort, A. (1953). Nature (Lond.), 172, 83. Ellis, R. S. (1919). J. comp. Neurol, 30, 329. Ellis, R. S. (1920). J. comp. Neurol, 32, 1. Feldman-Muhsam, B. (1944). Bull ent. Res., 35, 53. Gardner, E. (1940). Anal Rec., 77, 529. Grosch, D. S. (1950). Biol Bull, Wood's Hole, 99, 65. Hatai, S. (1902). J. comp. Neurol, 12, 107. Haydak, M. H. (1953). Anti. ent. iSoc. Amer., 46, 547. Hodge, C. F. (1894). J. Physiol, 17, 129. Inukai, T. (1928). J. comp. Neurol, 45, 1. Lansing, A. I. (1947). J. Geronl, 2, 228. Lansing, A. I. (1948). Proc. nat. Acad. Sci. {Wash.), 34, 304. Lansing, A. I. (1954). Ann. N.Y. Acad. Sci., 57, 455. Maurizio, a. (1954). Landw. Jh. Schweiz., 68, 115. McCay, C. M., and Crowell, M. F. (1934). Sci. Mon., N.Y., 39, 405. McCay, C. M., Crowell, M. F., and Maynard, L. A. (1935). J. Nutr., 10, 63. McCay, C. M., Maynard, L. A., Sperling, G., and Asgood, H. S. (1941). J. Nutr., 21, 45. McCay, C. M., Maynard, L. A., Sperling, G., and Barnes, L. L. (1939). J. Nutr., 18, 1. Murphy, D. P. (1954). Ann. N.Y. Acad. Sci., 57, 451. Rockstein, M. (1950a). J. cell. comp. Physiol, 35, 11. Rockstein, M. (19506). Ann. ent. Soc. Amer., 43, 152. Rockstein, M. (1953). Biol Bull, Wood's Hole, 105, 154. Rockstein, M. (1956). J. Geronl, 11, 282. Rockstein, M. (1957). J. Geronl, 12, 253. Rockstein, M. (1958). J. Geronl, 13, 7. Rockstein, M., and Lieberman, H. M. (1958). Nature {Lond.), 181, 787. Rockstein, M., and Lieberman, H. M. (1959). Gerontologia {Basel), 3, 23. Rollins, R. Z. (1959). Personal Communication. SoNNEBORN, T. M. (1957). Proc. Ageing Conference, Gatlinburg. Wash- ington, D.C. : A.I.B.S., in press. Tracey, K. M. (1958). Ann. enl Soc. Amer., 51, 429. Wilkes, A., Buciier, G. E., Cameron, J. W. McB., and West, A. S., Jr. (1948). Canad. J. Res., D., 26, 8. Woke, P. A., Ally, M. S., and Rosenberger, C. R., Jr. (1956). Ann. ent. Soc. Amer., 49, 435. Discussion 265 DISCUSSION Gerking: Did McCay keep these trout you mentioned until they died? You referred to an increased Hfespan for these fish. Rockstein: My recollection is that he was able to double the life- span of the trout. Comfort : The total duration of the experiment was only a matter of months, while they were in the hatchery. The lifespan of trout is at least 10 years. Berg: The advantage of using rats for lifespan studies is that we can determine causes of death in this species, whereas the pathology of other species has not been studied as thoroughly. Rockstein : We are not as fortunate as you in having a pathologist. There are insect pathologists, but they are very rare and very costly. Berg: McCay employed drastic underfeeding so that his animals were severely retarded in growth and did not attain sexual maturity. The greater longevity of McCay's rats has been attributed in part to the retardation of sexual development. In our experiments, sexual maturity of the females was delayed only three to four weeks, and skeletal measurements were only 5 to 7 per cent less than in ad libitum-fed rats. These findings indicate that longevity of the rat can be increased by dietary restriction without seriously affecting skeletal growth or sexual maturity. Rockstein: We are all eagerly awaiting the results of your experi- ments on longevity. Comfort: 1 am a little disturbed, Prof. Rockstein, about drawing analogies between the nutritional requirements of rats and those of insect imagos. For example, I take it there is very little extragonadal mitosis in your flies. Any change you get is not due to altering the number of cell generations or the stage of development. Rockstein: No, I did not mean to draw that analogy. I should have mentioned that adult flies are essentially fully-grown animals, much like the honey-bee, but even more so because the honey-bee takes about ten days to attain full maturity. The housefly on the other hand is completely mature within a few hours except for the ability to lay fertile eggs, under standard laboratory conditions. But I did mean very definitely to compare the human with the rat, and to emphasize that starvation would hardly be of any use in the human if one wanted to prolong life. I do not know that it has ever been shown to do so; if anything, it would shorten the lifespan in man. When McCay and his workers came out with the pronounce- ment that a low protein, low calorie diet was what we needed for a 266 Discussion long life, I thought that was rather a broad inference from their particular study on white rats. Comfort: I do not think any human population has ever been sub- jected to the sort of controlled and selective restriction of diet which Dr. Berg and Prof. Simms have been using. Starved populations are deficient in all foods, and do not receive adequate vitamin supple- ments, as McCay's rats did. There is a difference there. I agree with you entirely about the general principle. Sacher: Restrictions in diet during and after the past war in several countries may have had a relation to the observed decrease in mortality from heart disease. Tanner: In some of the degenerative diseases, for example diabetes, the incidence and the death rates went down. But this is different from the notion that at the same time children are being starved and therefore they might live longer later on. The starvation during the war lasted a sufficiently short time, so that those children who were starved probably picked up on to their natural growth curves a little later on. It was acute or sub-acute starvation, which is probably compensated for pretty rapidly. We know that the human, like the rat, gets back to the normal growth curve fairly rapidly, even after severe disease or severe starvation. It is for this reason that I do not think this data is particularly relevant. Sacher: Nevertheless, such children constitute a cohort which can be followed in successive decades. Even though normal growth is resumed, there 'may still be permanent after-effects detectable in later susceptibility to disease. Jalavisto: Did you measure the death rate at earlier dates in this parental age series. Prof. Rockstein? Rockstein: Yes, I have curves, but this was a very limited study, involving about 150 flies in each case, and so the data are not really adequate for preparing such curves. Maynard Smith: My colleague Miss Clarke has been doing experi- ments on the effect of larval nutrition on the longevity of Drosophila, The animals are kept as adults in the same environment on the same food, but are fed as larvae on diets varying from 0 • 03 per cent up to about 16 per cent of dead yeast. I do not think she would want to commit herself very much on the results, because she has not finished doing the sums. However, it is quite clear that the effect, if any, of larval nutrition on adult longevity is very small. It has an effect on the time it takes the animals to develop, and on how big they are when they emerge from the pupae, but it has only a very tiny effect on their adult survival in either sex. I confess I find that surprising. Discussion 267 Rockstein: Do they lay eggs? Maynard Smith : Yes. The more protein you give them as larvae, the higher the rate of egg-laying when they are adults. We therefore suspected that the ones which had a lot of protein might not live as long as the others, but there is no overall effect of any great magni- tude. If there is an effect it is of the order of 10 or 20 per cent — not more than that. Kershaw: There is an analogous situation in parasitism. The ability of the tsetses to act as vectors of sleeping sickness is largely determined by the temperatures at which the pupae are maintained before the adults come out. What effect that has on the longevity of the adult is not known, but it will be a sort of parallel viability, or parallel parameter that we put against longevity. It is the middle-aged insect which is the important survivor for parasites, because the ones which die young do not hve long enough to transfer the parasite. Secondly, a very complex pattern of mortality is evident in the development of a parasite in different selected organs of an insect. A third point is the ability of the insect to support the parasite, so that those at the tag end of their life cannot act as vectors. That has an obvious application in the field but what this means biologically one does not know. Unfortunately there are no means at the moment of quantitatively assessing the capacity of insects to support some parasites. Wigglesworth : Could you describe the sexual difference in the effect of protein feeding by saying, rather as Prof. Kershaw is im- plying, that the adult male is not protein- starved, but that the adult female is starved of protein by its reproductive activities — therefore in the absence of the extra protein feeding it succumbs early? Rockstein: Yes, that inference would be very appropriate. Sacher: I found a survivorship curve for male Drosophila almost identical with what you found for the male housefly. Prof. Rockstein. Unfortunately I did not get data for female Drosophila. In regard to your remarks about the complexity of the survivorship curve for the female (and I think we should say for the male too), perhaps we should recall the controversy between Crozier and Pearl (Pearl, R., White, P., and Miner, J. R. (1929). Proc. nat. Acad. Sci. (Wash.) 15, 425). Pearl had studied the resistance of Drosophila to alcohol as a function of age and he got a curve which he graduated smoothly. Crozier (Crozier, W. J., Pincus, G., and Zahl, P. A. (1936). J. gen. Physiol., 19, 523) objected that such smoothing was not proper and he did a far more extensive experiment. He established that the resistance of Drosophila to alcohol as a function of age went through many stages and was an exceedingly complex curve. This comes 268 Discussion back to the fact that we must eventually dissect the life-table into several components, as Dr. Benjamin has pointed out, and as H. S. Simms (1940. Science, 91, 7) has shown previously. Brauer (personal communication) at the Naval Radiological Defence Laboratory in San Francisco is using McCay's technique now. He re-feeds rats at various ages and finds that the rate of growth when full feeding is re-established carefully is preserved up to quite late ages, and then diminishes. This is somewhat like Comfort's regeneration and re-feeding experiments. Brauer is also examining the effect of previous X-irradiation on the ability to resume growth. Rockstein: McCay could accelerate senescence in an old animal which appeared to be young because of starvation. By re-establish- ing the full diet, the animal's appearance rapidly shifted to that of an old animal and mortality was accelerated. THE RATE OF AGEING IN DROSOPHILA SUBOBSCURA J. IVIaynard Smith Department of Zoology, University College London To a geneticist, the oddest feature of gerontology is the absence of a coherent and generally accepted theory of age- ing, comparable to the chromosome theory of heredity. In case this remark should cause any misgivings, it should be said that no attempt will be made to remedy this defect. Instead, two kinds of theory which seem to be possible will be indicated, since this will help to interpret some experiments to be described later. We have to accept that a theory of ageing may be valid only for a single species or group of related species. It may be that we shall find a theory which proves to have the same universality in the study of ageing as does the chromosome theory in genetics, but this does not at present seem very likely. What kinds of theory, then, can be put forward to explain ageing in a single species, say in men or in mice or in fruitflies? A distinction should be made between two types of theory, which may be called "single" and "multiple" theories of ageing. A multiple theory would postulate that there are a number of partially independent processes occurring in every in- dividual, any one of which may ultimately cause death. It is not intended to imply that two processes in a single in- dividual can ever be wholly independent; by "partially independent" is meant only that each process would continue, perhaps at an altered rate, in the absence of the others. Now some ageing processes are fairly certainly independent in this sense. For example, the mechanical wearing away of the teeth of herbivorous mammals would occur even if other 269 270 J. Maynard Smith ageing processes were arrested, and, unless there is contin- uous tooth growth, would ultimately lead to death. Similarly, in so far as cancer is the result of cumulative environmental insult, it is partially independent of other ageing processes, though it would be rash to assume that it is wholly so. But it is always possible that apparently unrelated symptoms of ageing may be due to a single cause, just as apparently un- related abnormalities of development may be the pleiotropic effects of a single gene. A "single" theory would postulate that all or most of the symptoms of ageing are the consequence of a single process (or of a single series of processes), either at a cellular or organism level. There is one observation which at first sight appears to support such a single theory. In a given species, the deter- ioration of different organ systems proceeds at roughly the same rate; if this were not so, individuals dying of "old age" would always die of the same immediate cause. This synchrony might suggest a high degree of physiological inter- dependence, with some one particular process acting as a timekeeper. But the synchrony can be explained in another way. Suppose that ageing in mammals is in fact multiple in character. Then if in any species one ageing process, say the deterioration of the central nervous system, proceeded at a much higher rate than did other ageing processes, there would be strong natural selection tending to slow down the rate of ageing in this system, if necessary at the expense of accelerating other ageing processes. In other words, natural selection will tend to synchronize different ageing processes, even if these are physiologically independent of one another. The example of tooth wear already mentioned demonstrates that selection can in fact act in this way. The volume of tooth worn away in unit time is proportional to the volume of food eaten, which in turn is roughly proportional to the surface area of the animal. Consequently, the expectation of life of the teeth of small mammals is less than that of large mam- mals. However, many small herbivores (rodents) have Rate of ageing in Drosophila suhohscura 271 evolved molar teeth which grow throughout life, whereas large herbivores have not. Similarly, the age of onset of cancers in species with different life expectancies (say in mice and men) is roughly proportional to those expectancies, and this proportionality seems more likely to be a consequence of synchronizing selection than of a direct physiological con- nexion between ageing generally and cancer. It follows that a decision between a single and a multiple theory in any species is impossible without experimental Table I Mean survival times in days of adult flies at various temperatures Temperature °c. Males Females No. of flies Survival time in days No. of flies Survival time in days 20 „(, /raised at 15^ ^\ raised at 25 30-5 33 50 25 25 50 10 67-4 ± 2-46 29-5±107\^ 24-6 ± 110/" 7-58 ± 0-28 0-79 ± 008 50 25 25 50 10 55-9 ± 2-58 40-5 ± l-68\„. .. 30-6 ± l-65/'^^"^'' 17-60 ± 0-65 0-82 ± 005 interference with the process. The grafting of organs from young individuals into old and vice versa, or between individ- uals with different genetically determined rates of ageing, is perhaps the most promising experimental approach (Jones and Krohn, 1959; Medawar, 1957). Some experiments w411 now be described on ageing in Drosophila suhohscura, using a different approach, but leading to the conclusion that the ageing pro- cess is a multiple one (Maynard Smith, 1957, 1958). It has been known for a long time that in poikilothermous animals the expectation of life decreases with increasing temperature. It was the purpose of the investigation now to 272 J. Maynard Smith be described to discover how far the processes responsible for death in D. subobscura are the same at different temperatures, differing only in the rate at which they proceed, and how far different processes are concerned at different temperatures. 100-1 c 4 4) O ^ 10- 1- J I I L J I I I I L -1000 \- 100 ^ c 20 35 10 25 30 Temperature (°C.) Fig. 1. Survival time of flies at different temperatures. A, in food vials; B, in saturated air; C, in dry air; A» ^j females; 0> •> males; □, ||, sexes combined. Figs. 1-3 and Table II reproduced by courtesy of the Editor, Journal of Experimental Biology. The mean ages at death (measured from adult emergence) of adults kept continuously at various temperatures are shown in Table I. In Fig. 1 these values are plotted on a logarithmic scale, together with the survival times of flies exposed to higher temperatures without food or water in dry and in saturated air. The rather sudden change in the slope of the curve in Fig. 1 suggests that the causes of death at high Rate of Ageing in Drosophila suhohscura 273 temperatures may be different from those acting below about 31°. This suggestion is confirmed by the finding that the changes which occur at high temperatures are, wholly or in part, reversible, whereas the changes which occur at 30-5° are irreversible. The reversibility of a change is judged by exposing in- dividuals to high temperatures intermittently, with interven- ing periods at a lower temperature. Thus if flies are exposed to a high temperature (33-5° in dry air or 34-3° in saturated air) for 50 minutes (i.e. for about half their expectation of life at that temperature) and are then kept for three hours at 20°, their survival times when they are again exposed to the high temperature are as great or greater than the survival times of flies not previously exposed. Thus death in these conditions is due to changes which can be reversed at a lower temperature; therefore the changes responsible for death at high temperatures are not regarded as processes of senescence. Experiments in which flies were kept in food vials alternately for eight hours at 33° and for 16 hours at 20° showed that the changes responsible for death at 33° in food vials are also in part reversible. In contrast, as is shown in Table II, the changes responsible for death at 30-5° are not to any appreciable extent reversed at lower temperatures. There is evidence for a small degree of recovery in males, since the first eight-day interruption at 20° did slightly increase the further expectation of life at 30-5°, although the second interruption did not. Females which were exposed intermittently had total survival times which were if anything slightly shorter than those of flies exposed continuously. Since the changes responsible for death at 30-5° are, at least in the females, irreversible, and since they take an appreciable time to reach completion (mean of 17-6 days for females), it seems reasonable to regard them as processes of ageing. The question then arises, are they the same processes as are responsible for ageing at 20°? If the processes of ageing 274 J. Maynard Smith at the two temperatures were in fact identical, it would be possible to predict the total lifespan of flies kept for varying periods at the two temperatures. For example, a female exposed for eight days to 30-5° soon after emergence would Table II Expectation of life at 30-5° c. No. of flies Further expectation of life at 30 • 5° (days) Females (1) Exposed continuously to 30-5° 50 17-60 ± 0-65 (2) After 5 days at 30-5° (a) Exposed continuously 50 12-60 ± 0-65 (b) 8-day interruption at 20° after 5 25 1102 ± 0-28 days at 30 • 5° (3) After 13 days at 30-5° (a) Exposed continuously 44 5-82 ± 0-49 (b) 8-day interruption at 20° after 5 25 302 ± 0-28 days at 30-5° (c) Two 8 -day interruptions at 20° 22 5-23 ± 0-32 after 5 and 13 days at 30-5° Males (1) Exposed continuously to 30-5° 50 7-58 ± 0-28 (2) After 5 days at 30-5° (a) Exposed continuously 49 2-64 ± 0-27 (b) 8-day interruption at 20° after 5 44 5-23 ± 0-38 days at 30 • 5° (3) After 8 days at 30 • 5° (a) Exposed continuously 21 1-40 ± 0-35 (b) 8-day interruption at 20° after 5 36 2-97 ± 0-36 days at 30-5° (c) Two 8-day interruptions at 20° 25 2-42 ± 0-23 after 4 and 8 days at 30 • 5° be expected to have completed about half its expected life- span, and therefore to have a further expectation of life at 20° of about 28 days. Experiments do not confirm this simple additive hypothesis. Fig. 2 and Table III show the results of exposing young adult females to 30-5° for varying periods, and then keeping Rate of Ageing in Drosophila subobscura 275 them at 20° until they died. The exposure, so far from decreasing their expectation of Hfe, in fact increased it, by as much as 50 per cent in females exposed for eight days. In a similar experiment, a group of males were exposed to 30-5° for five days, or two-thirds of their expectation of life at that temperature. The further expectation of life of these males at 20° did not differ from that of a group of unexposed con- trols. 56 64 72 80 Age in days 96 104 112 120 128 136 Fig. 2. Survival time at 20° of females previously exposed to 30-5°. A, unexposed; B, exposed for 5 days; C, exposed for 8 days; D, ex- posed for 12 days. It follows that, both for males and females, different pro- cesses are responsible for death at the two temperatures; we are therefore obliged to accept a multiple theory of ageing for this species. The situation is further complicated by the different response of males and females to exposure to 30 • 5°, w^hich prolonged the life of females but left that of males unaltered. The clue to this difference was found when it was observed that exposure to a high temperature caused a partial regression of the ovaries of females, which sub- sequently laid eggs at only about half the rate of unexposed females. This suggested that the process of egg-laying might accelerate ageing in females, and that the exposure to a high 276 J. Maynard Smith Table III Expectation of life in days of females kept at 20° c. No. of flies Further expectation of life in days at age 17 days Kept continuously at 20° Exposed to 30 • 5° for 5 days (6th to 10th day after emergence) Exposed to 30 • 5° for 8 days (6th to 13th day after emergence) Exposed to 30-5° for 12 days (6th to 17th day after emergence) 50 47 18 15 38-9 ± 2-6 57-2 ± 30 67-8 ± 4-9 500 ± 6-6 temperature prolongs life because it slows down the rate of egg-laying. This suggestion has been confirmed by experiments using virgin females (which lay eggs at a reduced rate), and females lacking ovaries. The latter were obtained by using females homozygous for the mutant " grandchildless " (Spurway, 1948), whose offspring appear to be normal in all respects Table IV Expectation of life of femai.es kept at 20° c. No. oj flies Further expectation of life in days at age 10 days Mated females kept continuously at 20° exposed to 31° for 5 days Virgin females kept continuously at 20° OVARILESS FEMALES kept continuously at 20° exposed to 31° for 3 days 48 23 89 28 22 331 ± 1-6 61-2 ± 5-7 58-7 ± 2-7 67-6 ± 4-7 64-2 ± 5-6 Rate of Ageing in Drosophila suhohscura 277 except for the absence of gonads. The results of these experi- ments are shown in Fig. 3 and Table lY. As before, the exposure of normal mated females to a high temperature increased their expectation of life at 20°. As would be ex- pected from the hypothesis being tested, both virgin and ovariless females lived for longer than did the controls, and closely resembled the females exposed to a high temperature. The final confirmation of the hypothesis comes from the fact 16 24 32 40 48 56 64 72 80 88 96 104 112 120 Age in days Fig. 3. Survival time of females at 20°. A, normal mated females; B, ovariless females; C, normal virgin females; D, normal mated females exposed for 5 or 6 days to 31°. that the expectation of life of ovariless females, as of males, is not increased by exposure to a high temperature ; since such females will not lay eggs in any case, exposure cannot further prolong their life. The greater longevity of virgin as com- pared to mated females has previously been demonstrated by Bilewicz (1953) in Drosophila melanogaster, and by Griffiths and Tauber (1942) in Periplaneta americana. It is natural to suppose that the causes of ageing of ovariless or virgin females at 20° are the same as the causes of ageing of males at the same temperature. But these experiments leave one question unanswered. Does egg-laying shorten the life 278 J. Maynard Smith of females because it accelerates processes which occur in any case in ovariless females, or is it a process which would by itself ultimately result in death, even if other ageing processes could be arrested? We are again faced by a choice between a single and a multiple theory. As yet we have not been able to find a way of deciding between them, but we hope we may be able to do so by studying ageing in genetically different strains on varying diets, since in this way we have other means of altering both the rate of ageing and of egg-laying. Table V The longevities of inbred and of outbred flies IN days at 20° c. Mean lifespan Coefficient of variation Females Males Females Males Nine inbred /range lines \mean Four outbred r range populations'^ mean 17 -2-53 -8 36-4 55 - 9-64 1 GOO 17-1-69-2 40-0 44-7-67-4 56-8 0-35-0-69 0-51 0-29-0-35 0-32 0-35-0-66 0-55 0-23-0-50 0-33 We may now turn to the genetics of ageing in D. subobscura. Our interest in ageing originated with the discovery (Clarke and Maynard Smith, 1955) that the hybrids between inbred lines live for longer, and are less variable in lifespan, than their inbred parents. These findings have been confirmed by later work (Table V), although we were perhaps fortunate that the particular pair of inbred lines originally available for study showed the effect in a particularly striking manner. But later work has shown that, in addition to genetic variance due to "heterosis" or "overdominance", much of the genetic variance of longevity is due to genes with sex-limited effects, i.e. to genes with different effects on the longevity of males and Rate of Ageing in Drosophila suhohscura 279 of females (Maynard Smith, 1959). This can be shown in two ways. Table VI shows the lifespans of males and of females of nine inbred lines, two kinds of F^ hybrids between inbred lines, and the offspring of two groups of wild-caught females, one from Kent and one from Galilee. In eight of these 13 populations there was a significant difference between the Tables VI Relative longevities of males and females, in days, at 20° c. Lifespan Females Males Ratio P Inbred lines K 17-2 31-2 0-55 + + . M 35-3 51-8 0-68 + F 53-8 69-2 0-77 + O 48-7 52 o 0-93 NFS 40-7 42-4 0-98 D 50-2 47-5 106 B 33-3 25-8 1-29 + G 30 0 22-6 1-33 + E 36-2 171 212 + + Fj Hybrids K/NFS 55-9 67-4 0-83 + + B/K 61-5* 61-6 100 Offspring of WILD flies Kent 58 -C 53-4 110 Galilee 64 1 44-7 1-43 + + + +, significant at 0-001 level; +, significant at 0 10 level. longevities of the two sexes, but in four cases it was the males and in four cases the females which lived for longer. This can only be explained by the presence of genes which affect the longevity of the two sexes differently. The same conclusion emerges from a study of the correlations between the longevi- ties of relatives in a population derived from females caught in Galilee (Table VII). All the correlations are rather low; this means only that many differences between members of the 280 J. Maynard Smith population were due to uncontrolled variations in environ- mental conditions. But all the correlations between relatives of like sex were significant and positive, whereas only one of the four correlations between relatives of unlike sex was significantly different from zero. The presence of sex-limited genetic variance of longevity is understandable in view of the physiological findings described earlier. Since the causes of ageing in males and ki females are Table VII CORREL/V-TIONS BETWEEN RELATIVES AMONG THE DESCENDANTS OF FEMALES CAUGHT IN GaLILEE Like sex Unlike sex Brother-brother < p,^ Sister-sister < ^^ Father- son Mother-daughter 013 0-19 012 0-20 0-29 015 Brother-sister < p,^ Father-daughter Mother-son 0 04 004 0 19 -004 at least in part different, it is to be expected that gene dif- ferences should exist with different effects on the longevity of the two sexes. The moral seems to be that the genetics of a character can often be better understood if something is known of its physiology. To sum up, the most important conclusion which has emerged from this work seems to be that ageing in Drosophila subobscura is "multiple" in character. The processes re- sponsible for death at 30-5° are reasonably regarded as pro- cesses of "ageing" or "senescence", since they are not reversed or repaired at 20°, and since they take an appreciable time to reach completion. Yet they are not the same as the processes responsible for ageing and death at 20°. Further, the process of egg-laying either accelerates the normal Rate of Ageing in Drosophila subobscura 281 ageing processes of females at 20°, or is itself an age processing capable independently of causing the death of females. REFERENCES BiLEWicz, S. (1953). Folia biol. {Krakow), 1, 177. Clarke, J. M., and Maynard Smith, J. (1955). J. Genet., 53, 172. Griffiths, J. T., and Tauber, O. E. (1942). Physiol. ZooL, 15, 196. Jones, E. C, and Krohx, P. L. (1959). Nature {Lond.), 183, 1155. IVIaynard Smith, J. (1957). J. exp. Biol., 34, 85. Maynard Smith, J. (1958). J. exp. Biol., 35, 832. Maynard Smith, J. (1959). J. Genet., in press. Medawar, p. B. (1957). The Uniqueness of the Individual. London: Methuen. Spurway, H. (1948). J. Genet., 49, 126. DISCUSSION Danielli: There may be an alternative explanation for your experiments to the one you suggest, namely that the causes of ageing are multiple. In the study of cell division it is now common practice to synchronize cells by giving them a cycle of temperature changes. The logic behind this is that the synchronization is due to the breaking of a cycle or to interference with some phase of the cycle of metabolic activity, so that when the constraint due to temperature change is removed, all the cells take up a new cycle at the same point. In your animals the variance decreased in some of the experiments, which would suggest that some measure of syn- chronization was occurring. You may fail to get the additive effect one expects in ageing, not because the cause of death is different at different temperatures, but because you break the initial ageing cycle by moving from one temperature to another and then later the animals begin somewhat closer to the origin of a cycle than they would have done had you kept them constantly at one temperature. This would mean that there is possibly only one cause of death, although they are behaving as if there were two causes. Maynard Smith: I would accept that as a very possible explanation of the reduction in variance of the population exposed to 30 • 5° inter- mittently, compared to that exposed continuously. We shall have to repeat the experiment which showed this striking reduction in variance to see whether it was just one of those things that happen once, or whether it will happen every time. But I would not accept your suggestion as an alternative to the existence of the two processes of ageing. After eight days at 30 • 5° one knows that, although all the animals are alive, they are all actually "half dead". They are all 282 Discussion halfway through a programme towards death, and if left at that high temperature most of them would be dead in another eight days. If it was the same programme which is responsible for death at 20°, then one would expect that flies exposed for eight days to 30 • 5° and then kept at 20° until they died would behave as if they were halfway through the programme, and not, as actually happens, back at the beginning. Danielli: I was actually suggesting that the abrupt change of temperature swung the animals back to the beginning of the cycle in each instance. This would mean that you could take the animals halfway through their expectation of life at 30 • 5° and then the actual change in temperature swings them back to the beginning of their cycle again, or somewhere closer to it. Maynard Smith: If that were true, one could presumably make them almost immortal. You are suggesting that flies kept at a constant temperature die of physiological boredom. Sacher: I do not know whether your evidence clearly establishes the hypothesis of multiple as against a single cause of ageing. We had a diff'erent experimental situation which leads to results similar to yours. Fruit-flies were given daily doses of X-rays throughout their lives from emergence onwards. Under these circumstances flies that received about 1 • 5 to 3 kilorontgens per day throughout life lived more than 30 per cent longer than their controls and at the same time manifested a markedly decreased variance. Subsequently I discovered that W. P. Davey (1917, 1919. J. exp. Zool, 22, 573; 28, 447) had also done this with flour-beetles. My interpretation is that X-irradiation is a stress, and that a moderate degree of stress invokes adaptive responses that are not invoked in the animal's natural environment. This leads me to ask whether you could do an experi- ment in which temperature shocks are given daily or at frequent intervals? Maynard Smith: I shall have to do such an experiment. Whether you do or do not accept my conclusion that we have a multiple process here hinges largely on whether you accept my argument that what ultimately kills them at the high temperature is properly re- garded as a process of ageing. If you just starve a population of Drosophila they will all die in about three days, and you will get a survival curve with an increasing force of mortality, looking just like a life-table. But I do not think any of us regard this as a proper process of senescence because it is fully reversible ; if you give the flies food after two days' starvation they recover completely. I am arguing that the process at 30-5°, which takes as long as 18 days to reach completion in a fly whose normal life-expectation is only Discussion 283 about 50 days, and which does not seem to be reversible, can properly be regarded as a process of ageing in its own right. Therefore I am rather unhappy about thinking of exposure to 30 • 5° simply as the application of a stress. Rotblat: Our own results could be explained by either single or multiple processes, but like Sacher I think your results do not necessarily contradict the single theory. You assume that high temperature produces only ageing processes. But it may cause something else ; it may cause trauma or some other process which is not normally present, and consequently it may not be just an ageing process. Griineberg: The question of whether we are dealing with single or multiple processes of ageing could probably be tackled by investigat- ing the effects of individual genes on the ageing process. In the ex- periments you described, considerable differences were found between different inbred strains. These, of course, differ in a multi- tude of genes and in practice it is impossible to sort out the effects of individual genes following a cross between two inbred strains, as I have repeatedly found in my mouse crosses. It would probably be a better plan to start with single-gene differences. We are about to do that in mice, to see whether genes without obvious pathological effects in some way affect the longevity of the animal. In the mouse this will take about three years, whereas if you are so inclined, you could probably produce significant results in Drosophila by Christ- mas. Maynard Smith: I certainly could produce results by Christmas. R. Pearl (1928. The Rate of Living. University of London Press) found a long time ago that the gene vestigial in Drosophila melano- gaster halves the expectation of life. I have no doubt that a number of other genes would alter the lifespan. But I am not a good enough insect pathologist, and to learn anything from such an experiment one should analyse the causes of death. Griineberg : The gene vestigial is not a good gene to use because it is itself obviously pathological: because of their reduced wings the vestigial flies tend to get stuck in the food. One should use genes that have no obvious pathological effect. Maynard Smith: Suppose that you find a particular gene, which is not obviously pathological, but which reduces or extends the life. It would not tell you much unless you could then show that animals with the gene did not die of a specific cause that the others are dying of, or vice versa. Griineberg: That is exactly what I mean. Once you have shown that a gene affects the lifespan, whether it has an effect one way or 284 Discussion the other, it may then be possible to identify the physiological channels through which the gene affects the lifespan. But this is difficult if not impossible if you are dealing with the joint results of a multitude of genes which differentiate two different inbred strains or populations. Maynard Smith: It will be easier in the mouse, although it will take longer, because you have more idea of what mice die of than I have of what flies die of. Rockstein: Your results do not really agree with Dr. Maurizio's or mine. The honey-bee lives longer when it has a functioning ovary, certainly in the queenless colony. The housefly lives longer when it is well fed and is laying lots of eggs. On the other hand, you implied that the effect of temperature in prolonging life was through the destruction of the ovaries. Of course, these occur together and may not necessarily involve cause and effect. Instead, this may be the net effect of temperature through a more important higher level of control, which affects ovary development (as well as other processes) so as to result ultimately in the rapid dying off of the population. Sacher: There is also the opposite interpretation, that since the ovary is regenerated it is not the destruction, but rather the regener- ation, that extends life. Maynard Smith: You can get the same results with animals which never had and never will have ovaries — these animals will live much longer than their double first cousins who have got ovaries. It is reasonable to take the simple hypothesis that it is the egg-laying that matters. Since, as Prof. Wigglesworth commented earlier, females are liable to suffer from protein-shortage, my results are quite con- sistent with yours. On sugar and water your females may have died young because they were suffering from serious protein shortage. My experiment suggests that if you deprive a female of her ovary or cause it to regress, then she does not suffer from protein shortage as much as she would if she were laying eggs. Gerki7ig: Can you give an estimate of the size of the ovary in relation to the body? Is it as much as 50 per cent? Maynard Smith: I do not know exactly, but it is certainly not as much as that. Gerking: In the fish I talked about the ovary may weigh as much as 20 or 30 per cent of body weight at maturity. Once the eggs are shed then you can hardly find the o^"ary. Its restoration to this 20 or 30 per cent level requires a great amount of energy. I wanted to point out that in both the fish and Drosophila a very large proportion of metabolism is devoted to egg production. You have concluded Discussion 285 that egg production is at least one factor responsible for ageing in Drosophila, but I have been unable to find any relation between fecundity and age in egg-laying fishes. Hinton: In most flies when the ovaries ripen the abdomen swells greatly, and the ripened ovaries occupy a considerable percentage of the total volume of the abdomen. The other organ systems are frequently much displaced by the ripened ovaries,which may account for a high percentage of the total weight of the female. Maynard Smith: They are not so big in Drosophila. Eggs are pumped through at a rate of about 30 or 40 per day, but the ovary itself is never very large. Berg: Is the ovary a self-regulatory mechanism in the fly? A selective effect on the ovary without aff'ecting hormonal regulatory mechanisms would be unusual. Maynard Smith: What do 3^ou think the regulatory mechanism might be here? The ovary is not itself a hormone producer, is it? Wigglesworth : One of the main detectable abnormal effects on the insect of raising the temperature is the effect on hormone action. You can get an insect which is apparently metabolizing normally, but you knock out the action of the growth-promoting hormone. The ovary certainly appears to have a hormonal influence, a sort of feedback influence, upon the endocrine system. So that even in your ovariless insects produced genetically you might be impairing their endocrine system through lack of this feedback mechanism. Maynard Smith: The ovariless flies are the off'spring of females homozygous for the mutant grandchildless, and the suspicion, which is not adequately demonstrated, is that females carrying this gene produce eggs without pole plasm. This would explain the fact that the female off'spring have no ovaries and the male off'spring no testes. If the ovariless females had lived for a shorter time than the controls, I would have said something else was wrong too, but they lived 50 per cent longer. I was very reluctant to accept this simple mechani- cal explanation, that they live longer because they do not lay eggs, but everything seemed to fit in so well that until something does not fit, I have to accept it. Kershaw : The overwhelming effect of ovaries and this relation to nutrition has been shown in some experiments that we have done. We have exactly the opposite results from those in your Drosophila. In Aedes aegypti, which depends largely on blood meals, the virgin females live for a much shorter time than the normal egg-laying females (Lavoipierre, M. M. J. (1958). Nature (Lond.), 181, 1781). This disparity would fit in with the complete dislocation of nutrition invoked by the disturbance of normal ovarian function. GROUP DISCUSSION Comfort: I would like to emphasize the variety of the material to which we have been obliged to apply the term lifespan. I am inclined to say that if a single parameter must be used to designate a curve — it is better it should not be, but if it must be — for most of the purposes we have been talking about I would favour the modal age of adult death which Dr. Benjamin showed us. Unfortunately there are many curves for which you cannot use the modal age of death. In zoo animals there is effectively no mode, since the survival curve is almost an arithmetic straight line (Com- fort, A. (1957). Proc. zool. Soc. Lond., 128, 349; Ciba Found. Coll. Ageing, 3, 14.) Another possible parameter that has been mentioned is the median. It has the advantage for experimental purposes that you need not wait till the animals studied are all dead — you can rush into print when half of them are dead. But I think its standard error is a little difficult to handle. It also has the drawback that it is very sensitive to the effects of environment on the survival curve. The last decile is far more stable in this respect. You could also use the limit. The limit has the advantage that even in small populations of animals one or two commonly survive much longer than their fellows — their performance is a better index of "physiological" performance than the crude mean or median. Its drawback is the existence of a large number of doubtful records of very old age in man and animals. In Bourliere's curves for birds, and also many of Beverton's curves for fish, standing mortality at low ages is so high that it is effectively independent of age; the most obvious parameter is the half-life, but since these populations contain some long-lived individuals the limit is also possible. But they cannot be made to indicate what happens to the mortality at ages which are so rarely reached. To compare lifespans we might fit a set of curves and compare their time scales. Without using any equations I superimposed those for the K/B Drosophila, for a human population (1941 United States males), for Murie's wild sheep, and for my thoroughbred mares, on different time scales, by fitting the last three quartiles of each unsmoothed curve. That is another way of defining and com- paring lifespans ; but if you do that you must allow for the fact that man has a uniquely long developmental period, whereas sheep have 286 Group Discussion 287 not, while in Drosophila I used only the imaginal lifespan and ignored the whole of its previous larval career. I have not included among these examples the agricultural type of lifespan, which Dr. Hartwig mentioned for his cattle and horses ; that is yet another question. And I should add that in fish kept in the laboratory we obtain a series of curves under different degrees of environmental comfort which are very like those for mammals, including man. In all cases the force of mortality rises with in- creasing age. Muhlbock: The mouse has the advantage that there are inbred strains available, as you know, and we use them in our work on cancer research. These inbred strains come from brother-sister matings, made for at least 20 generations, and all the strains I shall refer to here have been through more than 100 inbred generations. Survival curves for the females of the DBA and 020 strains show that hybrids from the two strains live longer than the pure inbred lines. An analysis of the DBA strain shows that the males die first and the virgin females live longest; in between come the breeding females. By breeders we mean females which are allowed to rear their young. For special purposes in cancer research there is another group of females which is described as force-bred. That means that the young are discarded after birth, so that no lactation occurs. The lifespans of the force-bred females are plotted from the 12th month, after the fertility period has ended, so their shorter lifespan cannot be due to accidents in pregnancy and there must be some influences which affect lifespan in the second half of life when the fertility period has ended. The CBA strain is one of the longest-lived strains we have; some of the virgin females live to an age of 35 months. Here again the virgin females have the longest survival. In this strain the males are not so different, but the force-bred females, with rapid pregnan- cies and without lactation, have a shorter survival. But that is not the case in all the different strains we have. In the O20 strain the differences between the various states are not so great as in the other ones. There are therefore differences in these different strains but I do not know what is the reason for them. Rockstein: In the life-tables of the male houseflies I was interested particularly in the d^ values because from about the 10th to the 24th day of life this represents a fairly large part of the population. We are really at the peak of the mortality during this period of cohort existence. In the females there is a grouping in the d^ values, so that they reach a peak, then fall a little, and then reach another peak, and so on. This illustrates the idea which the probit curve seems to sug- gest, namely that the female is involved in a more complicated type 288 Group Discussion of mortality for the populations, so that at a different age an addi- tional factor seems to be interjected into the mortality picture. Perhaps Mr. Perks would comment on these tables [not printed]. Perks: First, I find it rather strange that you have such a large number of "ages" in your life-tables. We actuaries, of course, string out the rates of mortality for each year of age, but that is for the practical purposes of life assurance. For understanding the mortality that underlies the life-table we would certainly compress it, and we would not have 67 values of the independent variable. We would probably group these in fives and show the values of qx for five intervals at a time. That is a general question of presentation, and of trying not to confuse the reader with too many figures. The next point is that the distribution of dx for male houseflies gave me the impression of a curve very much like the Karl Pearson type 3 frequency curve, that is the gamma distribution, that comes up to a peak fairly quickly and has a long tail away to the right. The mortality curves for electric light bulbs that E. G. Pearson published 25 to 30 years ago had very much of that characteristic ; they were fitted fairly well by the type 3 distribution. If we are actually to understand anything about the underlying mathematical processes of mortality curves, we should start with the simpler organisms, or simple physical objects, and electric light bulbs are particularly suitable for this purpose. You can get them fairly homogeneous, and put them on a uniform circuit, so cutting down much of the extraneous variation. It certainly is interesting to see a death curve with a long tail to the right. The human death curve tends to have the tail to the left — coming up slowly to the peak, and then coming down very sharply. Beard has fitted incomplete gamma functions to a number of human life-tables, but he had to do a bit of manipulation with the data first, to remove the accident and infectious diseases mortality, otherwise the tail on the left-hand side would not asymptote to zero. There is a mathematical model that provides some analogy with the death process. Imagine that a population of objects are put on a wall, and shot at at regular intervals so that each is equally likely to be hit. Then suppose you define death as when a particular object has been hit n times; then the death distribution is in fact the type 3 distribution. Thinking along those lines may help us to get mathematical representations of the death curves of more complicated objects. I am particularly interested to see that the housefly appears to give a relatively simple distribution. My impression of the female table is that it is rather similar to the male, except that the peak is much flatter. So far as the so-called Group Discussion 289 subsidiary peaks are concerned, I would be astonished if they were not just the result of random error, q^ is irregular, and so obviously no graduation process has been applied. I assume that the life-table was obtained by following the history of a cohort, and that the Ix figures are in effect the numbers actually surviving to each age, the numbers being reduced to a radix of a thousand. Rockstein: That is right. There were about 4,000 animals of each sex there, 3,875 females and slightly more males. Perks: That is a technique that with humans we have not found very helpful, because it takes 100 years to follow a cohort through. But even though you get your rates of mortaUty that way, before I would draw any conclusions whatever from bumps on the dx curve I should want to put a light graduation through the q^ values, and then recompute the l^ and dx columns. You would have to apply a goodness of fit test. But a quite elementary graphical graduation would probably be sufficient to get rid of the accidental bumps in the dx column. Gerking: What do you mean by a light graduation? Perks: I mean putting a smooth curve through the points repre- sented by qxy so that you get rid of, or greatly reduce, the random errors — on the hypothesis that if you had a much larger number in your sample the departures from the smooth curve would largely disappear. I agree that the assumption of the smooth curve for the qx is only a hypothesis, but there is a great deal of observational evidence for that assumption, provided you keep your condition reasonably constant. S acker: When you say smooth, you do not necessarily mean simple? Perks: No. There is no satisfactory mathematical definition of smoothness. There has been some controversy in the Institute of Actuaries on what we mean by smoothness, and some people suggest it should be absence of roughness ! Rotblat: How sensitive is the gamma distribution to the value of n? Perks: You can get a wide range of different curves with different values for n. Comfort: I am impressed with that remark, because this is almost what Failla or Szilard have suggested is in fact happening, isn't it? (Failla, A. (1958). Proc. Ageing Conf., Gatlinburg. Washington, D.C.: A.I.B.S., in press. Szilard, L. (1959). Proc. nat. Acad. Set. (Wash.), 45, 30). The objects are actually being shot at by radiation, and this may be one of the causes of chromosome deterioration. Perks: If you are interested in that, R. E. Beard developed the subject further some years ago (see Appendix, p. 302). If you increase AGEING — ^V — 10 290 Group Discussion the probability of being hit according to the number of times the object has been hit before, and if you proceed further and increase the speed of the shooting, the mathematics develop in the direction of the Makeham curve and certain modifications of it. Maynard Smith: I do not believe for one moment that the shape of these curves has anything to do with the organism that has been studied. I think it has something to do with the environment in which it was studied. Perks: I have thought about this problem for a long time and I believe the shape has something to do with time. Tanner: You would expect a different-shaped curve in a less long- lived organism? Perks: No, I think it is the cumulative factor in life; injuries and so on are additive, or cumulative in almost a geometrical sense, and you must expect to have exponentials coming into the form of the mortality rates. Tanner: Does this make the change, then, from the ganmia func- tion into the Makeham type of curve? You think that the reception state changes according to the number of shots impinging? Perks: Yes. Holt: In the models for fitting the depreciation of motor cars death is assumed to result from either single big accidents, or an accumulation of small ones, and a complex death curve follows; it seems to have some possible application to animal mortality. Rotblat: I am concerned to know what lifespan is because we have to express some of our findings in terms of lifespan. For example, people often speak about the effect of radiation in causing a shorten- ing of the span of life and they put down figures of the percentage of fife- shortening per rontgen. But everyone I have asked what he meant by the term lifespan gave a different answer. I was hoping that perhaps in this meeting we might come to some agreement on that. I was impressed by your remarks on the first day. Dr. Benja- min, when you brought in what you called the senescence peak. This peak seems to me to be the quantity to put down as the *' normal lifespan". I was particularly interested to see data presented by various people which showed that although the shapes of the mortality curves may differ enormously, nevertheless for a given species they all reach nearly the same end-point. If we could draw the ideal curve, i.e. if we could eliminate all deaths due to accidents, infections, etc., then the "normal lifespan" would be obtained im- mediately from the senescence peak. But in practice this will be very difficult, because most mortality curves do not approximate to this ideal. We are, therefore, still left with the question of what we Group Discussion 291 mean by lifespan. If one could establish, once and for all, the normal lifespan for a given species, then the observed mortality curve would tell us the deviation due to environmental conditions, such as being kept in a zoo, or hunted. Then we might introduce another index to take these conditions into account. We could take the 50 per cent survival time from the actual mortality curve and the ratio of this to the normal lifespan might serve as an index of, say, longevity or senescence. These two quantities, the normal lifespan and the 50 per cent survival time, might be the proper parameters to use. Comfort: The limit, as I say, is not always advantageous, because of the very great divergence in the reports of maximum age in animals. My own feeling is that once you can get a family of curves like this, or a rough indication of the family, you could perhaps take a point 10 per cent back from the limit. Roiblat: If you take the mortality curve, then depending on the condition in which the animals are kept we would obtain different times for the 50 per cent survival but the same limit. Comfort : I am thinking of cases where you have not got curves as good as that. If you do have such a family of curves, there is no difficulty. The trouble is to know what to do with the sort of thing that Sacher was describing, where he wants to compare a whole range of mammals ; merely to get a rough correlation it is necessary to give some sort of figure which one can compare. Perks: The difficulty with the modal value is that it is influenced to a certain extent by the infantile mortality, and by mortality from accidents and infections. If you reduced those mortalities, then, as a mere piece of arithmetic, the mode is advanced and the curve of death becomes much steeper, by the mere fact that more lives survive to the ages at which the rates of mortality are high. I have been thinking about this question of what would be a useful measure of lifespan. Some sort of technique such as actuaries use, which we call multiple decrement technique, might be used, at least in theory. It is rather laborious, but if you could separate from the death curve all those deaths which have nothing to do with lifespan — accidental, predatory, infantile, and anticipatory deaths, — then there is a technique for getting a residual life-table, a hypothetical life-table, that is concerned only with those causes of death which affect biological lifespan. This would have all the disturbances taken out of it, and of course you would get a much later and taller death curve. Comfort: The trouble with that is to know which causes are age- dependent. Rockstein: I think we are being unrealistic about this. I would suggest that mean longevity seems to be a thing that you could 10* 292 Group Discussion obtain all the time, regardless of the cause of death. If you are con- cerned with a standard value or reference against which you can com- pare, why not use an animal, as you do in the laboratory, from which you can get such data? I have been amazed that for over ten generations in the housefly, for example, one can continue to get the same mean value. I say this value is a good basis for comparison in an experimental study. With the white rat, under specific laboratory conditions, for example, the only thing that will vary in irradiation tests will be the extent or dosage of radiation. With humans we run into another problem because we always deal with life expectation based on a population that is not always an identical cohort, but for experimental studies we can get cohorts of an identical nature for inbred lines of a number of species of lower animals. I do not see why the mean longevity is not as good a criterion as anything else ; it considers the accidental deaths, it considers the possible disturb- ance that the mode would have from having had early deaths or accidental deaths, and even a tail at the end resulting from the extended longevity of the few long-lived cohort members. Perks: I am sure you are right for your problem, but diff'erent problems require different solutions. Rotblat: If we say, for example, that radiation causes a contraction of the scale of life, and suppose we are dealing with the ideal case in which all animals die of old age, then for the irradiated animals we would obtain the same curve but bodily shifted to the left. This would be very nice, but in practice it may not be so. It may be that radiation has caused earlier deaths without changing the end-point. The curve would then change completely. Which of these will actually happen depends upon the effect that radiations have on the lifespan. Sacher: The average is certainly the first quantity to use in the experimental situation, but you cannot characterize all the effects of radiations, or of any other environmental influence, in terms of a single parameter. Empirically you can then proceed to the succeed- ing central moments. The question is to find out what parameters of the survival curve are being influenced by the particular environ- mental factors under investigation. I have pointed out how, in studies of radiation effects on mice, you could characterize the eff'ects of radiations in terms of the A and the a parameter of the Gompertz equation [q^ = Ae°^]. A single dose of radiation — to restate what I said yesterday — causes a change in A, without a change in a. Con- tinuous exposure causes a change in the a coefficient. It is perfectly true that this Gompertz equation is not an entirely adequate des- cription of the life-tables of natural populations, but it should be Group Discussion 293 borne in mind that this is an oversimpHfied form, suitable for dis- cussing general principles. In application to data, more complicated expressions are used. Each major disease category needs a separate Gompertz term, as has been shown by Simms (1940. Science, 91, 7). In addition the experiment need not be simply proportional to age, a?, but may be a function of x. Thus the general expression for the description of mortality in terms of a summation of Gompertz terms is gx = 2 A^efi^^) i = l Perks : If you are fitting mathematical expressions to your data of statistical distribution, then clearly you estimate the parameters, and your estimated values for the parameters sum up the statistics. I thought the problem we were really talking about was how to characterize statistics for which you have not got a mathematical expression. For a single measure to be sufficient the distribution would have to be a very simple one, such as l^ = an exponential, in which case you have got a single parameter and the measure might be the constant rate of mortality or the half-life. In any other case, you cannot sum up the distribution by a single measure, and you cannot even say that any particular measure is the best one. All you can say is that for some purposes one measure may be better than another. You have to accept that the expectation of life or the half- life or mode or whatever you may regard as the lifespan, gives you only part of the information contained in the statistics. When the life-tables of different animals are compared, it may be that the lifespan is good enough, and I think that view was expressed yesterday. When the mortalities of the same species in different environments are compared, I would agree that expectation of life is probably as good as any. In general terms, if you are going to have more than one figure as a measure of a death curve, probably the 1st, 2nd, and 3rd moments of the death curve would be as good as any. It is not until you get a mathematical expression for the death curve that you can really say that any parameters are better estimators than any others. Benjamin: All this discussion of lifespan seems to be only a means to an end ; we really want to get away from lifespans to considering the ageing effects, for example, of changes in environment. For that we really want two things. First, we need some kind of function which is as discriminating as possible of the effects of ageing, so that it is very sensitive. That suggests that what is wanted is the middle part of the survival curve where a small change in the survival risk 294 Group Discussion may make a difference between the curve bulging up one way, or bulging down the other. The largest possible dispersion of effects is found here, which brings us to the idea of the 50 per cent survival age point. To get a time scale which enables comparisons to be made between different kinds of animals, on the other hand, you need something which is not sensitive to that kind of change. This suggests that you should use the peak of the curve of deaths, because although the height of the peak is ^'e^y much correlated with antici- pated deaths and so on, the actual movement of the peak is not particularly sensitive. So while it is unprofitable to talk about an ideal lifespan, it is quite practical and profitable to look at large families of curves of death for the same species, and see what kind of shape they tend to in general, so that you can get for the different species a typical modal attained length of life, which you could use as the time scale. Comfort: This is Bodenheimer's "physiological longevity" (Bodenheimer, F. S. (1938). Problems of animal ecology. Oxford University Press). Perks : I would like to plead that you should all take an interest in the international actuarial notation. I think it is a very good thing that all scientists should use the same notation if there is one which is generally accepted. Dr. Benjamin has included some of it in his paper and I think it probably could be extended to cover all your needs. We have heard the phrases lifespan and ageing over and over again but I do not think you will ever succeed in rigorously defining them. I think they are best left as rather vague concepts, as we know in general what we are talking about. I have been a little puzzled by the extent to which logs have been taken of various observed figures. To me it only confuses the issues, particularly when you have a graph on a logarithmic scale, although I do understand that sometimes it is necessary to do that to com- press the graph to reasonable dimensions. But there is no excuse for taking logs in arithmetic merely to get rid of some of the variations, and apparently to produce a correlation or regression which possibly is not there if you do not take logs. What does the logarithm of a residual mean? When you take a logarithm of the cephalization index and then associate it with a logarithm of body weight, and finish up with a logarithm of a lifespan, what does it all mean? There may be a very simple technique that might be useful for those of you who study lifespans or mortality of animals in their natural state. I understand that it is very difficult to get their ages, but often I imagine you will find stationary populations in the natural Group Discussion 295 state. So if you take the crude death rate over an interval of time, then a fair measure of the expectation of hfe, in terms of that inter- val, is the reciprocal of the crude death rate, provided that popula- tion is stationary, or nearly stationary. If it is not nearly stationary, you can probably make an approximate adjustment. The only other thing I want to say is that the actuary's use of life- tables is very different from yours. The life-table is not an end in itself for actuaries ; it is merely a step on the way from a set of mortality rates to the calculation of premium rates, reserves, bonuses and surrender values. Sacher: Logarithms are not introduced to mystify. They are actually a great convenience for computation. The classical law of allometry is that one dimension of an organism is related to another as [F = AX'^'] so that one of them varies as a power of the other; these allometric relations are almost always presented graphically on a double logarithmic scale. When you take the log of Y and the log of X there is then a linear relationship between these values. There are great advantages in using logarithms to fit a power func- tion by least squares. All of these considerations apply to brain weight, body weight, and lifespan as I have analysed them here. The index of cephalization is a pure number. It is the logarithm of the ratio of the actual brain weight of a species to the brain weight that is predicted by the overall regression of log brain weight on log body weight. Logarithms are also convenient in the present application because they introduce the property that all of the observations have approximately the same statistical weight in terms of the logarithmic transform. The lifespan, brain weight and body weight measurement all have about the same percentage error from mice to elephants and therefore the error in logarithmic units is roughly constant, even though the original absolute values have a million-fold range of variation — from 5 grams or so to 5,000 kilograms. Perks: I agree that if you have reason for a relationship of that form, then logarithms may ease the arithmetical processes. Sacher: There is no reason, in the sense of a general theory of growth and of the relationship between parts of an organism, that is capable of explaining why the allometric relationships should be of this form. It is, however, a fact of observation that the power function does describe these relations, and no other function does it as well. Comfort: It is also true that in drawing the survival curves of birds and small mammals, where over a large part of their lifespan their mortality is so high that it is almost age-independent, most people 296 Group Discussion use an arith./log scale, arith. for time and log for survival, so that constant mortality gives a straight line. Tanner: I have been sitting here for the last half-hour with a very strong feeling of dejd vu. The people who are interested in growth have been fitting growth curves with decreasing enthusiasm for about 35 or 40 years. It seems to me that you are pursuing a vertiginous and descending pathway ! I do not think there is anything in the general aspects of growth which leads one to suppose that the allometric relationship is very useful in general. There may be instances where it is necessary, not from any theoretical considerations but because using logs produces a straight line; I do not think that there can be any other justifica- tion. Sacher: A transformation, such as the logarithmic, cannot increase the amount of information contained in a set of data, so if the cor- relation of the transformed variables is 0 • 99 + this expresses a fact about the data, i.e. that only a fraction of a percentage of the total variance is error variance when the proper functional relation be- tween the variables is found. No a priori justification is needed for the use of the power function. In my own and Brody's data there is no question about its appropriateness. The fact that some other data are adequately rectified by a linear plot is interesting but it cannot contravene the allometric relations as they have been estab- lished in many other cases. Chitty: My particular problem is to find out why animal popu- lations in nature do not go on increasing indefinitely, and what it is that they die of. ^lost people up to the present have considered that deaths in nature could be almost entirely accounted for through predation or epidemic disease, heavy infestation with parasites, or food shortage, but it is now clear that this is a wholly inadequate explanation, particularly for the huge mortalities which occur in the young stages. The problem arises — what exactly is it that they die of? The suggestion was first made by P. H. Leslie and R. M. Ranson in 1940 (J. Anim. EcoL, 9, 27), for the field mouse, that the life-table type of explanation might be applied to field populations. In the laboratory you recognize that, with age, there is an increasing prob- ability of death from a variety of causes which are peculiar to the particular environments — that is to say a group of mice in one laboratory would not have the same final causes of death as they would in another — but in each case there would be the common fact that as they grew older they became increasingly liable to die of whatever it was that was peculiar to those environments. The field evidence strongly suggests that this may be a profitable way of Group Discussion 297 looking at natural populations, too. The problem now becomes to try and find some general law which is applicable in spite of the fantastic variety of conditions in the field. I think one would never hope to find any common causes of death associated with the actual time of death in nature. Every environment differs in its hazards frona every other one. The point is, can we find anything in the properties of the animals which does obey some sort of general law under all these varied circumstances? In other words, what is it that makes an animal increasingly likely to die regardless of what actually kills it in the end? We might divide the problem into both multiple and single processes. The final causes of death would be the multiple processes, and there we have very little hope of introducing much unity; but by concentrating on susceptibility there may be some hope of finding a common process which can equally well be studied in those animals which live to a great age in the laboratory, and those which die at a much younger age in nature. Exactly how one goes about this I do not know, and that is the point at which my work is hung up. I have to account for very violent changes in the probability of survival at different times of the population cycle, and exactly where does one go to look in the organism for something which may be an index of this change in properties? That seems to me to be very similar to the problems with which one is faced in trying to account for the increasing probability of death with age. Animals in the field very seldom live to an age at which you can say they are senescent. Nevertheless, it is a fact that even at a much younger age than they die at in captivity, some species periodically show this very great increase in probability of dying (see Green, R. G. and Evans, C. A. (1940). J. Wildlife Mgmt, 4, 220, 267, 347). The question arises of whether or not we should regard these as problems of senescence and ageing, or as much more analogous to the high probability of human beings dying young, or whether age is irrelevant and some index of physiological condition is the only thing worth trying to find in any species. Tanner: There does not seem to me to be any connexion between the situation in non-domesticated animals and the situation in the human. All the other mammals and birds would have died early in terms of human growth. The human was the only animal which seemed to be surviving long enough to experience senescent processes. Cellular ageing might paradoxically haA^e been closer to the situation in man, than is the situation in the passerine, for example. After what Dr. Chitty has said I feel like withdrawing this comment, because if in other species the probability of their dying in the field increases with age, this is the fundamental thing. It comes back to 298 Group Discussion what I was trying to say before, that just as in growth we talk about developmental age, so we can talk about the probability of death in any given situation as developmental age further down the scale. If that is really true, if these analogies can be made the same in animals and man, then we can use the results of both field workers and the experimentalists much more usefully to shed light on the human situation. Comfort : The figures we get for small mammals have often misled people into thinking that all mammals behave in this way. We have only very few figures for the larger ones but there is some evidence that many larger mammals in the wild achieve high ages quite often. I would be very surprised if some of the larger wild ungulates didn't live long enough to senesce. Wild horses, if one had the chance of observing them, may perhaps live to a fair proportion of the age reached by tame horses. Sacher: Porcupines can survive until they are so arthritic they can hardly climb trees — nobody comes up against them except mountain lions. Verzdr: Dr. Chitty, did you imply that the time of survival for a certain species depends on the number of individuals living at the time? If you put a pair of fish in a pond, they will have certain survival characteristics for their age. When the pond becomes full with fish the population will become constant. Then the individual survival curve will probably be different, otherwise the number of fish would continue to increase, which is impossible because there is not enough food. Chitty: The number alive has a very great influence on the survival rate but the survival time can certainly not be predicted from the numbers of the animals alone. One must also take into account the behaviour of the population ; experimentally you can have a family group of 50 animals which is perfectly amicable and has a good survival rate, yet with two animals in a similar space who are strangers, the survival time of one of them will be about half a day. The presence of other hostile animals of the same sort is, I think, one of the strongest environmental factors which does affect survi^^al, regardless of the amount of food available. Danielli: I shall not attempt in any way to summarize this meet- ing. I have been considerably refreshed by it in many ways. We have learned, for example, that man, skunks and porcupines are amongst the few animals who know how to survive into the period of Group Discussion 299 senescence. The nature of this grouping probably has some moral significance. A point that impressed me very much was the difficulty of evaluat- ing the significance of data obtained from animals in captivity, which are living in conditions of constant diet, no exercise, constant illumin- ation, very little in the way of seasonal change, and so on, to which even animals which have been selected for laboratory purposes are not really 100 per cent adapted. After all, the period of adaptation to laboratory life is comparatively short compared with the period of evolution. I do not really know what can be done about that. What is outstandingly important is that, wherever possible, a pathologist should look at the animals when dead, and this should be done for insects as much as for any other form of animal life. The concept of a biological time scale continues to interest us, though it is obviously even vaguer than some of the other concepts, which, as ]\Ir. Perks remarked, are better kept vague. However, this particular one is of no use to us unless we can measure it. It is useless to continue to use it without exact definition. Whatever the source, if we exclude accidents to which the individ- ual concerned does not contribute, in most instances susceptibility to mortality in a species or a strain does develop in a typical manner. It is perfectly clear from what has been said that if one cause of mortality is removed, for most of the animals for which we have had data analysed in sufficient detail here, some other cause of mortality would rapidly cause life to come to an end in the individuals which have survived. The total gain of lifespan which would result from eliminating one cause of death is not great. In other words, opera- tionally we appear to be dealing with a unitary process. Whether ageing is in fact unitary, however, cannot be determined from the data we have been presented with so far. I am not at all clear to what extent Maynard Smith's theory of synchronization of inde- pendent lethal processes is a valid one, but it is, I am sure, a very important matter indeed to have had raised, and one which necessi- tates a good deal of further thought and investigation. The concept of a limit to the life of a tissue which is set in terms of the energy conversion per unit mass per life cycle is a very attractive one, but possibly dangerously attractive. If we accept it at its face value, it presumably means that living matter commits accidents at a rate which is proportional to the rate of energy conversion in it ; this is quite a tenable view physically, and one which must surely be open to investigation. We need more data, for example from the study of hibernating animals, and from the use of metabolic poisons. We ought to use metabolic poisons and radiation in attempts to 300 Group Discussion desynchronize some of the processes which may be synchronized. It may be in this connexion that we could make some progress by studying systems in which one is normally dealing with symbiosis. Here the breaking-up of the symbiotic relationship might reveal phenomena which would be difficult to reveal by any other method. A cellular approach to ageing would under certain circumstances facilitate examination of the actual process of ageing, if this is in fact to some degree unitary, but I am not convinced that this is so with higher animals. The ageing process in higher animals may be fundamentally a function of the complex of cells and operate at a higher level of organization than is present in cells. Another point which we have not discussed yet, but which we are probably all agreed upon, is that it is quite possible that lifespan in man, as we see it now, is an entirely accidental by-product of selection for breeding efficiently at a much earlier age, and does not in any sense correspond to social needs of the moment. An accumu- lation of experience has become much more important, or at any rate equally as important as physical vigour; therefore socially speaking there is a very good case for anything which will enable us to modify the expectation of life in a radical manner. Looking at this from the long-term point of view, I think the possibilities for so doing would in fact be good by chemical means, provided ageing does in fact occur by a unitary process. If on the other hand it is due to a very large number of non-unitary processes which do not have any common mechanism, then I think the chemical approach to extension of lifespan is fraught with so many difficulties that it is hardly worth considering. To illustrate the order of magnitude which I would expect if there is a unitary process involved, I would like to refer briefly to the rate of mutation, or rather the rate at which damage is caused by radiation. It is of course a tenable hypothesis that somatic mutation is, one way or another, directly concerned with the ageing process. There are cells which are killed by a dosage of the order of 100 r. ; there are other cells which require a dosage of 1,000,000 r. to kill them. This difference of four orders of magnitude is not at present, as far as I know, accountable for in any cytological or physiological terms whatever. Therefore, if ageing were associ- ated primarily with somatic mutation, or some other generalized mechanism of this type, we might expect that under suitable con- ditions lifespans could be varied by that order of magnitude. This sounds a little like science fiction, but all of us have seen so much of what we regarded as science fiction 20 years ago turning up as reality, that I think we must envisage the possibility that we may be able to make radical changes in lifespan. Group Discussion 301 My final comment is that I regard it as singularly unfortunate that so much of the research on cancer is conducted essentially without reference to the phenomena of ageing, of which I think it is a part. I believe that not only would a study of ageing benefit from more knowledge of what is going on in older animals, but the whole field of investigation of cancer might benefit equally, and perhaps far more. This for the simple reason that to a very considerable degree, at the moment, cancer experiments are conducted by putting trans- plantable tumours into young, often rapidly growing animals, and then seeing what could be done in the way of chemotherapy in these conditions — a thing which does not approximate very closely to what happens in the average human patient. Prof. Miihlbock's laboratory is a shining example of an institute where this isolation of cancer research from ageing research does not occur, but I hope it will not long continue to be one of the very few shining examples. V APPENDIX NOTE ON SOME MATHEMATICAL MORTALITY MODELS R. E. Beard Pearl Assurance Co. Ltd., London 1. A satisfying basis for a law of mortality would be a formula that, starting from some fundamental concepts about the biological ageing process, led to a distribution of deaths by age which was comparable with observational data. Such comparison would not be simple and straightforward because environmental and secular factors would introduce distortions as compared with the theoretical underlying distribution. 2. In the course of numerical work, extending over a number of years, on the expression of human mortality functions by mathe- matical formulae, various attempts have been made by the writer to develop an approach on this basis. The results obtained have not led to any satisfying formulae, but they are suggestive of different lines of approach and have been summarized below in the hope they may be of value to others interested in the subject. The note follows the sequence in which the ideas have developed in the mind of the writer and leads from considerations based on the force of mortality, /x^, to those based on the curve of deaths, ^xjl^. 3. The first mathematical expression which provided a reasonable representation of the observed force of mortality in human data was that first proposed by Gompertz (1825) and later modified by Make- ham (1867). Basically the "law" was derived by postulating a relationship between the rate of change of the force of mortality at any age and its value at that age. The next significant modification to the Makeham law was the system of curves devised by Perks in 1932 and of which the important formula was the logistic. Many human life-tables have been graduated by this basic cur^'e, modified in some instances to allow for special features of the data, particu- larly at the younger and early middle ages, and the clear fact emerges that adult human mortality can be very well represented by a logistic curve of the form ix^- A = Be^^'Kl+De^') (1) 302 Appendix 303 which will be referred to as a Perks curve since this is the name by which it is generally known by actuaries (Perks, 1932; Beard, 1936, 1939a, 1951a, 1952a; Registrar General, 1951; Mortality of Assured Lives, 1956). 4. Now [Xg is the ratio of the ordinate at age x of the curve of deaths to the area under the curve above age x. We may look upon the curve of deaths as a frequency distribution of deaths by age at death and for many types of frequency curves it will be found that this ratio has a sigmoid form. It is not apparent whether the satis- factory representation of /x^ by a Perks curve is because the formula has a theoretical significance or because the formula does provide a good approximation to the particular function of a family of fre- quency curves which can be used to represent the distribution of deaths by age (Perks, 1953). 5. What evidence is available tends to support the idea that the force of mortality does not continue to increase indefinitely with age. The concept of a limiting age by which all individuals must be dead (i.e. a maximum lifespan) does not seem to be in accordance with the facts — the use of a limiting age as a mathematical device to cut off" a long slender tail has nothing to do with the present dis- cussion. Formula (1) leads to an upper limit of BID for fx^ and it is not without interest to note that the numerical values of B/D obtained from the graduation of human mortality data are of the same order as the force of mortality which can be deduced from select mortality tables as being appropriate to "damaged lives", i.e. about 0-57 (Beard, 19516). 6. If the rapidly decreasing mortality associated with the infantile and growth period be ignored the pattern of human mortality then exhibits a basic sigmoid form on which are superimposed waves and other disturbances. The waves appear to be due largely to secular effects (e.g. selective effect of war deaths); the main disturbances are those arising from accidental deaths and the (rapidly disappear- ing) hump at the early adult ages from deaths from tuberculosis. 7. For a broad mathematical approach we will be concerned with (a) accidental deaths (assumed to be at a constant rate at all ages), (b) an upper limit to the rate of mortality, and (c) a progression in time. Gompertz' law arises by using condition (c) only, i.e. djjijdx = A/x^ whence fi^ = B e^' (2) Makeham's law arises by using conditions (a) and (c), i.e. dfxjdx = X{ix^—A) whence ii^ = A+Be^ (3) 304 Appendix Perks' law arises by using conditions (a), (b) and (c), i.e. dfijdx = X(yi.-A) {E-tJi,)l{E-A) whence /x^, = ^ + ^ . -n .x. (^) The Perks (logistic) relation can be expressed as stating that the rate of change of yi^ is proportional to the product of its value and the amount by which it falls short of its upper limiting value. 8. If the requirement of a constant upper limit for the rate of mortality is relaxed other formulae can be developed on similar lines to those of the preceding paragraph. For example, d^ _ X{fi^-A) _ dx - l+B(fji^-A) ^'^^' "^'^ ~^^ where Wj, = B{jx^—A) (5) and ^ X{fM,-A)(l+^fi,-A dx ~ I 2D (l+-g/x,-^j gives /Lt, = ^ + ^(-l + \/l+4Z)e^^) (6) Formula (6) is equivalent to a continued fraction form for /x^, i.e. ^■^^^ D?^ 1 + ... and the relationship between formulae (2) to (6) is clearly seen by expanding the expressions for /^^ in terms of powers of e^'^ i.e. formula (2) gives B e^' „ (3) „ A+Be^' (4) „ A-\-B e^^'-BD e'^'^ + BD'' e^^ _. . . (5) „ A+Be^'-BDe''^ + lBD''e^^'^ —^' • (6) „ A+Be^-BDe''^ + 2BD^e^^' _. . . Appendix 305 9. The differences between formulae (4), (5) and (6) will become apparent only at the old or very old ages and unless the data were extensive the differences would be unlikely to be significant for many numerical processes. From a scientific point of view the models are, of course, quite different. 10. An alternative approach to the question, but still based upon rates of mortality, is to determine the conditions necessary for /x^ to be a Perks (logistic) curve, given that the population can be stratified according to a longevity factor and that the basic mortality law is Makeham in form (Beard, 19526). Thus letfx'f, be the force of mortality at time ( = age) k for the group having longevity factor s and let (f)(s) ds be the proportion of the initial population having factor s. Then the survivors of (j){s) ds at time k are {s) ds . exp ( - r fit dt\ (7) and the total survivors at time k h = j {s) exp (- J* fJLtdt^ds (8) where the integral is taken over the whole range of s. The force of mortality at time k {= —d log Ikjdk) is then (j>{s) jLtfc exp ( — jLt* dt) ds <^(5) exp ( — yi,ldt\ ds 11. From formula (9) it will be noted that fx^ is a weighted mean oi ix\ (= /x^ say). Since the number of lives with heavier mortality will diminish more rapidly than those with lighter mortality, s will decrease with increasing k. If the basic mortality is Makeham in form, then dfij^/dk will show a slackening off at the higher ages, i.e. the sigmoid feature shown by a logistic curve. In order to meet practical conditions some limitations are necessary on the form of (5) is a gamma function such that (f){s) ds = ks^ ery' ds (0 ^ s (s) is fil = (x + ps e^^, we have f ks^ €ry\ = -p+'-^ (16) 18. Attempts to use the formula of paragraph 16 on human mortality data have been unsuccessful, the shape of c^(log /x^y/^^ not 308 Appendix fitting well to observed values which show a negative second dif- ferential coefficient over the adult ages. 19. As an extension of formula (12) a model can be set up in which the "hits" in an interval can be single, double, etc., in known proportions. The basic relation then takes the form J" = -plT+p^^f{r)ir (17) This can be integrated to If = e-^' J pe^' 2 f{r) Vf^ dt (18) and by noting that ll = e-'" /„ values of If can be obtained by succes- sive integration. No experiments have been made using this form, mainly because the form of c/(log fJLtQIdt seems to be unsuitable for human data. The form of /(r) is also speculative. 20. A different forward model can be devised in which the proba- bility of a "hit" is dependent on the number of "hits" recorded already. We then have the following /77a ^ = -(^4-pa) Zf + (^+^.a-l) ir (19) This can be integrated to give with 'te^) = -(^ + 0^^) +^^^^^- (21) Here again the form of equation (21) does not accord with observa- tions from human data. 21. In the attempts to fit these forward type formulae to human data it was found (Beard, 1950, 1952c) that satisfactory numerical results could be obtained by expressing /x, /^ in gamma function form subject to a terminal age cu, i.e. the infinite tail of the curve is the opposite way round to what would be considered natural. This formula, after elimination of a constant element representing accidental mortality, can be derived from the difference-differential equation ^^=plf-pf-' (22) Appendix 309 the solution of which leads to If = l^ ^pe^-*) {p{oj - 1) Yjoi ! (23) from which dt ^ co-t if the deaths occur at the ath hit. In this formula p'~0-3,a^ll and oj --^ 110 for human mortality. 22. No obvious physical model applies to equation (22), but the relationship can be written in the backward form f = - ^ Z,« + «-±l r^ (24) at to — I co — t in which the rate at which a unit is lost is proportional to the number of units remaining divided by the years of life remaining to the final age CO. From a biological point of view the concept of a final age by which the organism must be dead is unsatisfactory, but the fact that satisfactory numerical results arise only from a backward formula suggests that a closer study of this type of model might be more profitable. 23. The simplest backward model arises from the relationship ^= -plf+pir (25) where the organism is assumed to lose a unit at rate p. This has a solution If = l^ e-^* {pt)^°'l(n - a) ! (26) where n is a maximum number of units. If death is assumed to occur when the number of units faUs below ?•, we have dCioQ 11.1) n — r ,„^. This is of similar form to equation (16) and is not suitable for human data. 24. By assuming that the rate of loss of a imit is proportional to the number of units remaining the relation ^ = -piP + c.) lf+piP + o^ + 1) r^ (28) 310 Appendix may be set up. This has the solution If = k e^'l(l +De^'y+°'+'- (29) If death occurs when the units fall below a, we have We also have = I D{\ +Z))^+«/(l -{-De^y^'' (30) dt ^ \+De^' ^ ' and I"' - ^^Td^ (^^) We have now found a difference equation model which leads to a Perks (logistic) formula for /x^. In formula (31) the upper limit of /x, is _p(/S + a); p ^ 0-1 and the limit ^ 0-7 so that (jS + a) ~ 7. 25. The distribution of a in the population at age 0 implied by equation (29) is a decreasing geometrical progression, i.e. D D D 1+D' {l+Df"'(l-{-D)'^ For human mortality D is small (of the order of 10~^) so that the dis- tribution is very slowly decreasing with increasing a. 26. The significant result which emerges from the experiments made along these lines is that to provide results which have some reasonable semblance to observed human mortality the backward type of model has to be adopted. This is consistent with death being regarded as the culmination of a degenerative process such that death occurs when the organism reaches a certain level of degeneration. The mathematical models are based on numerical results for adult ages and interpolation back to birth is possibly a questionable process, a more suitable approach being to regard the life and death process as a period during which the organism is building up to a complex situation with a subsequent degeneration. This would lead to models in which the whole of life process would be looked upon as the resultant effect of two opposing forces. Appendix 311 27. Calculation of the moments of the distribution of deaths by age for a population of mice (Greenwood, 1928) shows that a Pearson type III (gamma function) would give a fair representation, but, as with the himian data, the curve is the "opposite way round", i.e. subject to a terminal age. By inference the Perks (logistic) curve would give a fair representation of this data. No calculations have been made on animal data or on physical objects such as electric light bulbs and motor cars (e.g. Cramer, 1958) but it would seem worth while trying to find out if observed data of this latter type would distinguish between the two types of processes. REFERENCES Beard, R. E. (1936). J. Inst. Acta., 67, 53. Beard, R. E. (1939a). J. Inst. Actu., 70, 53. Beard, R. E. (19396). J. Inst. Actu., 70, 373. Beard, R. E. (1950). Proc. Centen. Assembl. Inst. Acta., 2, 89. Beard, R. E. (1951«). J. Inst. Actu., 77, 382. Beard, R. E. (1951b). J. Inst. Actu., 77, 394. Beard, R. E. (1952«). J. Inst. Actu., 78, 82. Beard, R. E. (19526). J. Inst. Actu., 78, 201. Beard, R. E. (1952c). J. Inst. Actu., 78, 341. Cramer, J. S. (1958). J. R. statist. Soc, 121, 18. Gompertz, B. (1825). Phil. Trans., 115, 513. Greenwood, M. (1928). J. Hyg. (Lond.), 28, 282. Makeham, W. M. (1867). J. Inst. Actu., 13, 325. Mortality of Assured Lives (1956). J. Inst. Actu., 82, 3. Papers of Royal Commission on Population (1950). 2, 154. Perks, W. (1932). J. Inst. Actu., 63, 12. Perks, W. (1953). J. Inst. Actu., 79, 199. Registrar General's Decennial Supplement, England and Wales 1951, Review in J. Inst. Actu., 83, 168 (1957). AUTHOR INDEX Plain numbers indicate a contribution either in the form of an article or as a contribution to the discussions. Italic numbers indicate a reference to an author's work. Aasen, O. . Akyuz, E. Ally, M. S. Aim, G. . d'Ancoma, U. Andersen, J. Anderson, J. Andrew, W. Ansell, S. . Appelget, J. Armbruster, L. Arora, H. L. Ascher, K. R. S. Asgood, H. S, Austin, O. L. Austin, O. L. Jr. 149, 156, 174 . 149, 174 . 255, 264 144, 155, 159, 168, 174 . 167, 174 92, 94, 95, 102 232, 234, 241 . 248, 264 21, 31 . 155, 174 232, 241, 242 149, 174 255, 259, 264 . 251, 264 98, 102 98, 102 Back, E. Bagenal, T. B. Baird, D. Banfield, A. W. F. Barcroft, J. Barnes, L, L. Barnet, H. A. R. Barton, R. A, Bauer, K. . Beard, R. E. Beeton, M. Belding, D. L. Bell, F. H. Bendell, J. F. Benedict Beneke, R. Benjamin, B. Berdegue, J. Berg, B. N. Berg, L. S. Bertalanffy, L. von Bertholf, L. M. . Bertin, L. . Best, A. T. Beutler, R. 235, 238, 241 192, 193, 200, 201, 202, 205, 207 . 33 92, 102 . 133 251, 264 . 5, 15 53,54 59, 69 4, 14, 288, 289, 302, 303, 305, 307, 308, 311 21, 31, 33, 53, 54 197, 207 151, 177 101, 102 . 124 110, 112 2, 15, 16, 17, 18, 19, 32, 34, 70, 293 . 755, 175 16, 32, 33, 72, 73, 74, 75, 76, 79, 79, 80, 82, 83, 84, 85, 86, 87, 88, 114, 137, 265, 266, 285 . 212, 225 157, 169, 175 . 235, 241 . 168, 175 . 104 235, 236, 241 Beverton, R. J. H. . 142, 149, 151, 158, 175, 177, 178, 179, 226, 227, 229 Bezem, J. J. 97, 103 Bidder, G. P. 173, 175, 182, 189, 190, 191, 207 Bilewicz, S. . 277, 281 Blackburn, M. . 146, 149, 151, 175 Blair, W. F. 96, 102 Blazka, P. . 170, 175 Bodenheimer, F. S. . 294 Bottcher, T. . 60, 61, 69 Bonin, G. von 119, 123, 128, 132 Boroughs, H. . 170, 175 Bourliere, F. 70, 84, 91, 96, 102, 103, 104, 105, 112, 113, 119, 132, 137, 217, 225, 227, 228, 245 Boyd, H. . . 101, 102 Breder, C. M. . 212, 217, 218, 225 Bretherick, O. . . 237, 243 Brody, S. . 124, 132 Brown, C. J. D. . . 168, 175 Brown, M. E. 147, 148, 149, 175, 217, 225 Brues, A. M. . 138 Buchanan, A. D. 59, 69 Bucher, G. E. . . 258, 264 Burd, A. C. . 149, 175 Burns, C. M. 21, 31 Butler, C. G. 237, 241, 243 Buxton, P. A. . 240, 241 Cameron, J. W. McB. . 258, 264 Cannon, C. 59, 69 Carlson, L. D. . . 89 Chalmers, T. A. . . 56 Chang, H-W. . 145, 146, 153, 175 Chipman, W. A. . 170, 175 Chitty, D. H. . 105, 296, 298 Chugaeva, M. 53, 54 Ciba Foundation . 128, 132 Clark, F. N. 149, 155, 175 Clark, H. W. . . 212, 226 Clarke . 266 Clarke, J. M. . . 278, 281 Clarke, R. D. . 4, 5, 6, 7, 8, 10, 11, 14, 15 Cohen, D. M. . . 151, 175 Cohrs, P. . 110, 112 313 314 Author Index Comfort, A. 17, 17, 32, 34, 35, 1 Fitch, H. S. 96, 97, 102 36, 37, 43, 44, 47, 52, 53, Fitch, J. E. . 155, 175 54, 55, 56, 70, 71, 84, 85, Flade, J. E. 66, 69 87, 88, 103, 104, 113, 114, Flower, S. . 119, 128, 132, 168, 175, 116, 117, 119, 132, 134, | 217, 225 m \, 137, 169, 173, 175, Foerster, R. E. . 151, 156, 175 \n 5, 179, 208, 208, 209, Fox, H. . 108, 109, 112 210, 217, 225, 111, 11^, 1 Franca, P. da . 153, 175 23C ), 244, 262, 264, 265, Franke, F. 62, 69 26e i, 286, 286, 289, 291, Frankland, H. M . T. . . 53, 54 294, 295, 298 Franz, V. . . 202, 207 Cooper, E. L. 144, 153, 176 Eraser, E. A. . 189, 207 Corkins, C. L. . 245 Freudenberg, F. . 62, 69 Count, E. W. 119, 123, 132 Freudenstein, H. 235, 236, 237, 241 Crowell, M. F. 169, 176, 251, 253, Freudenstein, K. . 232, 241 264 Frost, W. E. . 153, 175 Crozier, W. J. . . 267 Fry, D. H. 168, 170, 175 Curtis . 87 Fry, F. E. J. . 192, 207 Daly, C. . 21, 31 Gardiner, E. M. 21, 31 Danielli, J. F. 1, 15, 17, 19, 32, 33, Gardner, E. . 248, 264 34, 70, 103, 104, 114, 133, Geiser, S. W, 167, 175, 177 13( ), 137, 177, 178, 209, Gerking, S. D. 18, 79, 171, 175, 229, 230, 281, 282, 298 17S ), 180, 181, 183, 207, Dasmann, R. F, 92, 94, 103 209, 210, 211, 228, 229, Davey, W. P. . 282 230, 265, 284, 289 Davies, D. F. 72 Gerner, K. 63, 69 Davis, D. E. 95, 102 Gibson, J. R. 29, 31 Davis, W. S. . 192, 207 Gilbert, C, S. . 245 Deason, H. J. . 151, 176 Gompertz, B. ?, 15, 72, 79, 302, 311 Decider, C. L. . 159, 175 Gontarski, H, . 232, 241 Deevey, E. S. 91, 102 Gordon, J. E. 21, 31 Detkens 67, 69 Gottlieb, E. . 149, 175 Devoid, F. . 149, 175 Grainger, E. H. 147, 153, 175 Dickie, L. M. . 151, 175 Green, R. G. . 297 Dietrich, H. 60, 69 Greenwood, M. . 311 Dilley, W. E. . 169, 176 Griffiths, J. T. . 277, 281 Dinkhauser, F. 63, 69 Groot, A. P. de 235, 236, 237, 238, Doljanski, F. . 208 241 Dubois, E. . 123, 132 Grosch, D. S. . 255, 264 Duetz, G. H. . 85 Griineberg, H. 19, 20, 79, 80, 283 Dymond, J. R. . 175 Gumbel, E. J. Gunter, G. 117, 132, 134 . 167, 175 Edmonds, S. J. . 169, 175 Eidrigevits, E. V. 53, 54 Hafter, E. 196, 208, 218, 225, El-Deeb, A. L. A 232, 234, 241 226 Elkin, R. . . 168, 176 Hagel, L. 59, 69 Ellis, R. S. . 248, 264 Hald, A. . . 120, 133 Engeler, W. 59, 69 Haldane, J. B. S. 53, 54 Erickson, D. W. 196, 197, 208 Hall, D. A. . 113 Eschmeyer, P. H . 197, 207 Hansen, E. 59, 69 Evans, C. A, . 297 Hanyu, I. . . 151, 175 Evans, H. . 92, 102 Harmison, C. R. 73, 79 Evenius, C. . 232, 241 Harrison, J. L. . 95, 96, 102 Evermann, B. W . 212, 226 Hart, J. L.. Hart, J. S. . 144, 151, 155, 175, 176 . 170, 175 Fahr, H. O. . 110, 112 Hartmann, W. 59, 69 Failla, A. . . . 289 Hartwig, W. 56, 57, 65, 66, 67, Earner, D. S. . 97, 101, 102, 103 69, 70, 71 Farr, W. . . 9 Hatai, S. . . 248, 264 Farran, G. P. . 194, 207 Haydak, M. H. 232, 236, 237, 238, Farrar, C. L. 232, 234, 241 239, 241, 242, 253, Feldman-Muhsai Tl, B. . 259, 264 264 Finnell, J. . . 168, 176 Heady, J. A. 21,31 Author Index 315 Hejtmanek, J. Hems, J. Herald, E. S. Herrington, W. C Hervey, G. F. Hess, G. Hickey, J. J. Hickling, C. F Hile, R. . Hinton, H. E. Hinton, S. . Hodge, C. F. Hoffman, H. Hogreve, F. Holmes, S. J. Holt, S. J. . 86, 1 158, 179, Hoogendoorn, D Hubbs, C. L. Hunter, W. R. Huxley, J. Hynes, H. B. N. Hytten, F. E. Inukai, T. Irwin, J. O. Isupov, A. P. Ivlev, V. S. Jackson, B. H. Jalavisto, Eeva, Jenkins, R. Johansson, A. S. John, M. . Jones, E. C. Jones, J. W. Jordan, D. S. June, F. C. Kadic, M. . Kalabouchov, N. Karzinkin, G. S. Katz, M. . Kennedy, W. A. Kershaw, W. E, King, G. . King, J. O. L. Kluijver, H. N. Knox, G. . Koch, A. . Kocher, V. Koehler, A. Konig, K. Konopinski Kratky, E. Krause, C. Krohn, P. L. Krumholz, L. A. 235, 237, 242 . 227 155, 176 151, 177 . 227 235, 242 97, 101, 102 195, 196, 205, 207 151, 176 34, 112, 243, 245, 246, 284 191, 217, 219, 225 . 248, 264 . 237, 243 . 64, 65, 69 21, 31 34, 142, 149, 151, 155, 167, 175, 176, 178, 209, 210, 226, 227, 228, 229, 246, 290 21, 31 146, 155, 176 . 179 . 104 153, 156, 176 . 33 . 248, 264 37, 54 53, 54 186, 207, 208 . 89 16, 17, 21, 21, 31, 31, 32, 33, 81, 113, 266 . 168, 176 . 243 . 237, 242 . 261, 281 153, 156, 176 . 212, 226 . 193, 208 43, 54 96, 102 . 186, 208 196, 197, 208 145, 151, 176 56, 56, 70, 105, 114, 245, 267, 285 10 . 56 99, 102 21, 31 236, 238, 242 . 237, 242 . 232, 242 . 61, 62, 69 67, 69 232, 236, 242 . 110, 112 . 271, 281 171, 176, 187, 188, 208 Kuhn, W. , , 240, 242 Kuiken, A. . 237, 243 Kuznetsova, G. . 53, 54 Lack, D. . . 97, 101, 102, 104 Lansing, A. L . 259, 264 Larkin, P. A. . 182, 208 Lavoipierre, M. M. J. ". . 245, 285 Lea, E. 145, 149, 176 Lesher, S. . . 80 Leslie, P. H. . 296 Levin, M, D. 232, 236, 242 Levinson, Z. H. 255, 259, 264 Lieberman, H. M. 250, 251, 255, 256, 257, 258, 260, 264 Lindauer, M. . 233, 242 Lindop, Patricia J. 113, 136, 138, 141 Lorenz, E. . 89 Lotmar, M. 232, 233, 234, 242 Louveaux, J. . 236, 242 Luecke, R. W. . . 237, 243 McCay, C. M. . 169, 176, 251, 252, 253, 264, 265, 268 McCracken, F. D. . 151, 175 McFadyean, J. . 50,54 McGregor, E. A. 192, 208 McHugh, T. . 105 McKeown, T. 29, 31 MacMahon, L. . 21, 31 Maar, A. . 191, 208 Mackintosh, J. . 21, 31 Makeham, W. M. 2,4,15,302,311 Mannes, A. 59, 60, 69 Margetts, A. R. . 151, 176 Marlow, H. 63, 69 Marr, J. C. 153, 176 Martin, W. 59, 64, 69 Matsumoto, K. . 43,54 Mauermayer, G. 232, 234, 242 Maurizio, Anna . 231, 232, 234, 235, 236, 238, 242, 243, 244, 245, 246, 248, 254, 264 Maynard, L. A. 183, 208, 251, 264 Medawar, P. B. . 271, 281 Mellen, I. . . 217, 226 Menon, M. D. . . 149, 176 Menzel, D. W. . 183, 184, 185, 208, 210 Milinsky, G. L . . 207, 208 Miller, R. B. 168, 176 Miner, J. R. . 267 Miner, R. W. 53,54 Moltoni, E. . 103 Moore, H. L. . 155, 176 Morgulis, S. . 186, 208 Morris, J. N. 21, 31 Moskovljevic, V. . 234, 242 Miihlbock, O. . . 33, 71, 80, 287 Murie, A. . 54, 92, 102 Murphy, D. P. . ^. 262, 264 Mussbichler, A. . 2 35, 238, 242 316 Author Index Nail, G. H. Nickel, H. K. Nieberle, K. Nieto, D. . Nigrelli, R. F Ogborn, M. Olsson, V. . Opfinger, E. Orton, J. H. J53, 165, 176 . 232, 242 . no, 112 . 107, 112 113, 138, 147, 176, 178, 210, 211, 212, 213, 216, 217, 226, 216, 111, 228, 229, 244, 245 . 5, 15 99, 103 235, 236, 241 190, 191, 193, 198, 200, 201, 208 Paccagnella, B Pain, J. Pallaske, G. Palmer, L. S. Parker, R. R. Pearl, R. . Pearson, E. G. Pearson, K. Pearson, P. B. Perks, W. Phillips, E. F. Phillips, E. W. Phillips, J. B. Piel, H. Pincus, G. . Polyakov, E. V. Ponomareva, L. I. Pospelov, S. P. . Probst, R. T. Pycha, R. L. . 153, 176 235, 237, 238, 242, 243 . 110, 112 237, 242, 243 . 182, 208 . 267, 283 . 288 3, 15, 21, 31, 33, 53, 54, 288 . 237, 242 3,4,5,14, 75,81,288,289, 290, 291, 292, 293, 294, 302, 303, 311 232, 234, 242 4, 5, 15 . 149, 176 59, 63, 64, 69 . 267 53, 54 53, 54 53, 54 144, 153, 176 . 153, 176 Qasim, S. Z, Quirling, D. P. Radhakrishnan, Raitt, D. S. Rakowicz, M. Ranson, R. M. Rasquin, P. Reibisch, J. Renton, R. M. Ribbands, C. R. Rice, T. R. Rich, C. O. Richdale, L. E. Ricker, W. E. Ripke, G. . Robertson, J. A. Robertson, T. B. Robinson, O. J. 153, 156, 176 . 119, 133 N. . . 755, 776 149, 176, 198, 199, 200, 201, 202, 205, 208 . 155, 176 . 296 196, 208, 218, 226 . 202, 208 . 189, 207 232, 233, 242 . 170, 175 . 4, 15 100, 101, 103 145, 149, 151, 155, 176 59, 69 . 149, 176 . 137 59, 69 Rockstein, M. . 15, 31, 55, 83, 84, 112, 114, 179, 210, 226, 228, 232, 234, 243, 144, 245, 246, 247, 247, 248, 249, 250, 251, 254, 255, 256, 257, 258, 264, 265, 267, 268, 284, 287, 289, 291 Rosch, G. A 243 Rottgermann, W. RoUefsen, G. Rollins, R. Z. . Rosenberger, C. R. Rotblat, J. 59, 62, 69 149, 176 258, 264 255, 264 15, 19, 20, 86, 87, Rounsefell, G. Rubner, M. Rumbauer Runnstrom, S. Russell, E. S. Sacher, G. A. 103, 135, 177, 209, 210, 283, 289, 290, 291, 292 . 797, 208 124, 125, 126, 127, 128, 130, 133, 134 . 59, 63, 69 . 149, 176 190, 199, 208 87, { 133, 136, 140, 266, Sakagami, S. F. Samokhvalova, G. Sarkar, B. C. R. Sato, R. . Saxl, H. . Scheidegger, S. Scherer, H. J. Scheyer, W. J. Schieren, J. Schneider, F. Scholz, W. Schotterer, A. Schroder, G. Schwarz, J. Selye, H. . Shell, H. M. Shiraishi, Y Silliman, R. P. Simms, H, S. Simpson, A. C. Sluiter, J. W. Smith Smith, L. L. Smith, J. M. 87, 266, 279, 284, Smyly, W. J. P. . Snyder, D. P. . 75, 18, 31, 52,54, 80, 18, 104, 115, 775, 727, 133, 134, 135, 755, 137, 138, 138, 139, 141, 147, 176, 245, 282, 284, 289, 292, 295, 296, 298 233, 243 . 209 237, 243 151, 176 . 113 103, 106, 112, 113, 226 108, 112 89 59, 69 240, 243 107, 112 43,54 65, 69 238, 242 . 88 237, 243 151, 176 . 149, 176 11, 73, 74, 75, 76, 79, 266, 268, 293 203, 204, 205, 208 97, 103 59, 69 . 155, 174 16, 32, 33, 55, 86, 104, 134, 137, 141, 267, 268, 277, 278, 281, 281, 282 283, 285, 290 7-^5, 153, 169, 175, 176 95, 103 Author Index 317 Sonneborn, T. M. Soudek, S. Southern, H. N. Sperling, G. Spitskaya, T. D. Spohde, H. Spurway, H. Stevens, C. F. Stockklausner, F Stolbova, A. Strawn, K. Suau, P. . Summers-Smith, D. Sundnes, G. Sutter, J. . Svardson, G. Svoboda, J. de Sylva, D. Szilard, L. . Tabah, L. . Taber, R. D, Tainter, M. L. Tanner, J. M. Tanquary, M. C. Tauber, O. E. Taylor, C. C. Templeman, W. Terpenning, J. G Tester, A. L. Thiele, T. N. Thompson, W. F Thomson, A. M. Tibbo, S. N. Tiews, K. . Todd, F. E. Townsend, C. H Tracey, K. M. Tucker, D. W. Vaas, K. F. ' . Vallois, H. V. . Van Cleave, H. J. Van Heerdt, P. F. 32, 35, 53, 54, 262, 264 235, 243 101, 103 251, 264 53, 54 64, 69 276, 281 21,31 59,69 53,54 155, 176 149, 177 99, 101, 103 169, 176 21,31 171, 176, 182, 190, 208 . 235, 243 155, 175, 228 . 19, 129, 133, 289 21, 31 92, 94, 103 . 136 84, 85, 86, 88, 266, 290, 296, 297 . 232, 242 . 277, 281 149, 167, 177 146, 151, 177 182, 208 149, 177 . 3, 15 151, 177 . 33 149, 177 155, 177 237, 243 217, 226 262, 264 174, 177 170, 177 . 134 . 179 97, 103 Verzar, F. Vitt, V. O. Vives, F. . Vivino, E. VNIRO . Voogd, S. . Wahl, O. . Walford, L. A. Walker, E. P. Wallace, B. Wallace, W. Weaver, N. Webb, C. S. Weber, R. . Weidenreich, F Weitnauer, E. West, A. S. Weyer, F. . Weygand, F. White, P. Wigglesworth, V. 82,85,55, 112, 133, 137, 138, 141,298 35, 43, 46, 49, 54 . 149, 177 237, 238, 242, 243 . 149, 177 . 237, 243 159, B. 243, Wilkes, A. Wilson, I. . Winnigstedt, R Wittwer, S. H. Wohlschlag, D. Woke, P. A. Wolkoff, K. Wolstenholme, G. E. W Wright, N. C Wurzel, W. Wussow, W. Wynne-Edwards, V. C Yerkes, A. Yerkes, R. Yerushalmy, J. Yockey, H. P. Yokota, T. Zahl, P. A. Zamyatin, N. Zander, E. Ziegenhagen, G. 243, . 235, 243 177, 214, 226 . 119, 133 . 20 . 189, 208 . 237, 243 . 104 43, 54 . 134 . 104 . 258, 264 . 234, 243 . 237, 243 . 267 87, 133, 240, 245, 267, 285 . 258, 264 21, 31 59, 69 . 237, 243 145, 151, 177 . 255, 264 . 110, 112 . 71 59, 69 59, 69 . 66, 67, 69 . 193, 194, 195, 208 . 119, 133 . 119, 133 21, 31, 53, 54 . 129, 133 155, 163, 177 . 267 53, 54 . 232, 243 59, 60, 64, 69 SUBJECT INDEX Accipitres, arteriosclerosis in, 109 Acipenseriformes, lifespan and size of, 152, 220, 222 Acipenser fiihescens, survival curves of, 143, 144 Acipenser ruthenus, oldest age of, 191 Actuarial aspects of human lifespan, 2-20 Aedes aegypti, effect of diet on, 255 Age, changes due to in fish, 218 effect on fecundity in fish, 198-200, 201 effects of in insects, 247-268 effect on reproduction in fish, 181-182, 186-200 parental, and lifespan, 21-34 Ageing, genetics of in Drosophila, 278- 280, 283 in Drosophila, 269-285 molecular changes in, 121-131 physiological changes in fish due to, 181-211 theory of, 129-131, 169-111, 297, 281- 282 Albatross, lifespan of, 100, 103 Albumin, effect on bees, 235 Allometry of lifespan, 125 Alosa sopidissima, fecundity of, 191-192 Amiiformes, lifespan of, 222 Amino acids, effect on bees, 238 in pollen, 237 Amyloid, effect of diet, 79 Amyloidosis, in mice, 80-81 Angelfish, lifespan of, 221 protein metabolism in, 183-185 Anglia cattle, lifespan of, 60 Anguilliformes, lifespan of, 223 Anguilloidei, lifespan and size of, 152 Animal populations, control of, 296-297 Anseres, arteriosclerosis in, 109 Anseriformes, mortality rate of, 101 Apis mellifera, 247 Apodiformes, mortality rate of, 101 Arctic char, lifespan of, 147 Ardea cinerea, lifespan of, 99 Argentine, lifespan and size of, 150 Arteriosclerosis, aetiology of, 107-108 brain in, 107 cholesterol in, 110-111, 113 in birds, 106-114 in fish, 225 in mammals, 109 in monkeys, 108 site of, 1 10 Bass, lifespan of, 220, 221, 223 Batrachoidjfonnes, lifespan of, 221 Bats, lifespan of, 97, 103 Bees, ageing in, 248 brain cells in, 248 effect of diet, 254 effect of protein on, 254 in tropics, 246 lifespan of, caged, 235-239 factors influencing, 23 1-246 in free-flying colony, 231-235 physiological condition of, free-flying, 231-235 caged, 235-239 yearly life cycle in, 239-240 Beryciformes, lifespan of, 220 Birds, arteriosclerosis in, 106-114 lifespan of in captivity, 103 lifespan of in nature, 90-105 Bison, lifespan of, 105 Blackfish, lifespan of, 221 Black-pied cattle, lifespan of, 60 Biennioidei, lifespan and size of, 1 52 Blennius pholis, growth of, 156, 157, 158 Blenny, growth of, 156, 157, 158 lifespan and size of, 152 Blue-head, lifespan of, 221 Blue striped grunt, lifespan of, 221 Body weight, relationship to lifespan, 115- 139 Bowfin, lifespan of, 222 Brain, in arteriosclerosis, 107 internal environment of, 133 weight of, relationship with lifespan, 115-139 Brain cells, ageing in, 248 Breast cancer, in cattle, 71 Bullhead, effect of diet on, 169 lifespan and size of, 152, 223 survival curves of, 143, 145 Bulls, causes of death in, 65 lifespan of, 64-65 Burbot, lifespan of, 223 Biiteo buteo, lifespan of, 99 Butterfly fish, lifespan of, 121 Buzzards, Kfespan of, 99 Callionymoidei, lifespan and size of, 152 Callionymus lyra, mortality rate of, 146 survival curves of, 143, 145 Cancer, in fish, 216, 226 Cancer eye, in cattle, 71 Capelin, lifespan and size of, 1 50 mortality rate of, 146 Capreolus capreolus, lifespan of, 92-94 Carbohydrate, in pollen, 237, 238 318 Subject Index 319 Caribou, lifespan of, 92-94 Cat, lifespan of, 134 lifetime energy expenditure of, 127 Catfish, lifespan and size of, 154, 220, 223 Cattle, breeding of, 57-58 lifespan of, 57-65, 70-71 Cell, effect of temperature on, 281 Centenarians, accuracy of age of, 1 3 Central nervous svstem, control of vital functions by, 128-129 Centrarchidae, protein metabolism in, 182-183 Cervus elaphus, lifespan of, 92-94 Char, fecundity of, 191-192 lifespan and size of, 152 Characin, lifespan of, 224 Charadriiformes, mortality rate of, 101 Chickens, arteriosclerosis in, 110, 113 Chiropodomys gliroides, lifespan of, 96 Cholesterol, in aetiology of arteriosclerosis 110-111, 113 Cholinesterases, age activity of, 247 Chrysops, lifespan of, 105 Chub, lifespan and size of, 150, 221 Ciconiformes, mortality rate of, 101 Cisco, lifespan and size of, 150 Citellus pygmaeus, lifespan of, 97 Clupea harengus, growth of, 156, 157, 158 Clupea pallasii, fecundity of, 196-197 Clupeiformes, lifespan of, 220, 222 Clupeoidei, lifespan and growth in, 160- 166 lifespan and size of, 148 Clupeoids, survival curves of, 143, 145 Coalfish, lifespan and size of, 148 Coate's knifefish, 224 Cockroaches, effect of diet on, 253-254 lifespan of, 253-254 Cod, hfespan and growth of, 160-166 lifespan and size of, 148 Columbiformes, mortality rate of, 101 Cormorants, arteriosclerosis in, 109 Corregonus clupeaformis, survival curves of, 143, 144 Cottoidei, lifespan and size of, 152 Cottus gobio, survival curves of, 143, 145 Cow(s), average age of different breeds, 59 cause of death, 63-64 lifespan of, 58-64, 70 lifetime energy expenditure of, 127 Cowfish, lifespan of, 221 Creatinuria, in rats, 83 Cristivomer namaycush, mortality rates of, 145 Croaking gourami, 224 Cunner, lifespan of, 221 Cyprinids, metabohsm of, 169 Cypriniformes, lifespan of, 220, 222, 224 Cyprinodontiformes, lifespan and size of, 152, 224 Cyprinoidei, lifespan and size of, 1 52 Dab, fecundity in, 192, 200-202 Dace, lifespan of, 223 Dasyatis akajei, lifespan and growth of, 163 Death, accuracy of age at, 12-13 Death curves, 6-9, 15, 17, 286-296 of horses, 56 Death rate, anticipated, 4, 5, 7, 12, 15 laws governing. 2-4 senescent, 4, 5, 6, 7, 10, 11, 12 Deaths, accidental, 17 Deer, lifespan of, 91-94 Diet, effect on amyloid, 79 effect on bees, 254 effect on cockroach, 253, 254 effect on fertility, 34 effect on fish, 178,215-216 effect on flies, 249-255 effect on growth, 84-85, 177, 178, 254- 255 effect on lifespan, 169, 177-178, 249- 255, 265-267, 268, 282 effect on hfespan of rats, 78, 83-85, 87, 88,251-252,254 effect on mice, 79-80 effect on mosquito, 255 effect on sexual maturity, 84-85 effect on trout, 253, 254 effect on wasps, 255 protein in, 252-254 Dipodomys heermani, lifespan of, 96 Disease, onset of and longevity, 72-89 Doctor fish, hfespan of, 221 Dog, hfespan of, 134 hfetime energy expenditure of, 127 Dog snapper, lifespan of, 221 Dragonet, lifespan and size of, 1 52 mortality rate of, 146 survival curves of, 143, 145 Drosophila subobscura, lifespan of, 262, 266 rate of ageing in, 269-285 Drum, lifespan of, 221 Ducks, arteriosclerosis in, 109 Dwarf top-minnow, reproduction and senescence in, 189 Eel, lifespan and size of, 152, 223 natural death in, 174 Electric catfish, hfespan of, 224 Electric eel, lifespan of, 224 Environment, effect on hfespan, 167-168, 229 effect on onset of disease, 86-87 Epinephalus guttatus, protein metabolism in, 183-184 Falconiformes, mortality rate of, 101 Fat body in bees, 233, 235, 237, 238, 239, 244 Fat(s), in pollen, 237 storage in bees, 237, 238, 239 Fecundity, variation of in fish, 191- 192 Fertility, effect of parental age on, 56 Finch, lifespan of, 99, 103 320 Subject Index Fish (see also under common names) changes in due to ageing, 218 effect of diet on, 167 effect of fat diet on, 215-216 effect of growth and size on lifespan, 147-159 effect of toxic substances on, 177 egg counting in, 192-193 fecundity in relation to age in, 186- 206 growth and senescence in, 217 growth of, utilization of protein in, 182-186 growth rate and lifespan, 227, 229 infectious disease in, 213-214, 228 lifespan of, characteristics of long, 217-218 in different species, 219-224 in captivity, 212-230 in Nature, 142-180 lipid metabolism in, 227 metabolic disease in, 216 metabolism of, 169-170 natural death and reproduction in, 170- 174 natural mortality of, 142-147 neoplasia in, 216 nutrition in, 215 ovaries of, 193 parasites in, 213-216 physiological changes due to ageing, 181-211 protein utilization in, 210 relationship of age, mortality and growth in, 160-166 survival curves of, 143-147 variation in fecundity of, 191-192 Flamingoes, effect of diet on, 112 Flounder, lifespan and size of, 1 50 Fly (see Housefly) Food, conversion of in fish, 186 Gadiformes, lifespan of, 223 lifespan and growth of, 160-166 lifespan and size of, 148 metabolism of, 169 Galli, arteriosclerosis in, 109 Galliformes, mortality rate of, 101 Gambusia a. affinis, reproduction and senescence in, 187-189 Gannets, arteriosclerosis in, 109 Gar, lifespan of, 222 Gasterosteiformes, lifespan of, 152, 223 Gasterosteus aculeatiis, effect of environ- ment of, 168 Geese, arteriosclerosis in, 109 Genetics, and lifespan, 137 of ageing, in Drosophila, 278-280, 283 Glomerulonephritis, in man, 77 in rats, 74-76 Glycogen, storage in bees, 237, 238, 239 Golden Shiner, 223 Goldfish, lifespan of, 223 metabolism of, 170 Gompertz' Law, 2, 72, 302 Gompertz-Makeham equations, 1 1 7 Grayling, effect of environment of, 168 Grouper, lifespan of, 220 Growth, 296, 298 effect of diet, 84-85, 254-255 effect on lifespan, 147-159, 160-166 Grunion, lifespan and size of, 154 Guinea pig, lifetime energy expenditure of, 127 Guppies, effect of diet on, 169 growth of, 229, 230 regeneration in, 208-210 reproduction in, 171, 209 Habrobracon juglandis, 255 Haddock, effect of age on fecundity, 198- 200 Hfespan and size of, 148 relationship of fecundity and body weight, 200 sex organs in, 190 Hake, lifespan and size of, 148 Halibut, hfespan and size of, 148, 150 Heart disease, in man and rat, 74-77 Herons, lifespan of, 99 Herring, fecundity of, 196-197 growth of, 156, 157, 158 hfespan and growth of, 160-166 lifespan and size of, 148 relationship of gonad growth to body weight, 193 relation of size and maturity, 172 survival curves of, 143, 145 Hesperoleucus venustus, effect of environ- ment on, 168 Heterandria formosa, reproduction and senescence in, 189 Hibernation, effect on lifespan, 103-104 Highland cattle, lifespan of, 59, 61, 62 Hippocampus hudsonius, lifespan of, 147 Hippoglossoides platessoides, fecundity in, 192 Hippoglossus spp,, lifespan of, 147 Holocanthus bermudensis, protein meta- bolism in, 183-185 Horses, breeding of, 55 causes of death in, 68 coat colour, and longevity, 49-50 death curves of, 56 lifespan of, under various climatic con- ditions, 65-69 lifespan of English thoroughbred, 35- 56 lifetime energy expenditure of, 127 survival curves of, 35-42 Housefly, ageing in, 247-262 effect of diet on, 249-255 lifespan of, 249-262, 287-288 effect of paternal age, 259-262 sex differences in, 255-259 sex ratio in, 249 Humming-birds, lifespan of, 105 Subject Index 321 Ide, lifespan of, 223 Index of cephalization, 123, 135 Infectious disease, in fish, 213-214 Insects (see also under names of species) ageing in, 247-268 lifespan of, 105 overwintering in, 232-234, 240 Irradiation, effect on lifespan, 19-20, 138- 141,282,290,292 effect on onset of disease, 87, 88 Jack, lifespan of, 220 Kidney disease, in man and rat, 74-77 Killifish, lifespan of, 220 Labidesthes, growth of, 156, 157, 158 Labidesthes sicculus, mortality rate of, 146 Lagonostica senegala, lifespan of, 99 Lamniformes, lifespan of, 220 Laws of mortality, 2-5, 302-311 Lebistes reticulatus, effect of diet on, 169 fecundity in, 209 regeneration in, 208-210 Lepidosteiformes, lifespan of, 223, 224 Leucichthys kiyi, survival curves of, 143, 145 Leucichthys sardinella, metabolism of, 169 mortality rates of, 145 Leuresthes tenuis, survival curves of, 143 Lifespan, allometry of, 125 and parental age, 21-34 effect of brain and body weight on, 1 1 5- 139 effect of diet on, 169, 249-255, 282 effect of disease on, 72-89 effect of egg laying on, in Drosophila, 275-278 effect of environment, 167-168 effect of growth and size on, 147-159 effect of hibernation, 103-104 effect of metabolism, 103-104, 124-126, 129, 136-137, 169-170 effect of mitotic inhibitors, 136 effect of parental age, 43-47, 53, 259- 262 effect of radiation on, 19-20, 138-141, 282, 290, 292 effect of reproduction, 179, 275-279, 284-285 effect of temperature on, in Drosophila, 271-279, 283, 284 effect of thyroid gland on, 137 genetic aspects of, 33, 55, 137 in nineteenth century, 23, 32 mathematical basis of, 286-296 measurement of, 133-134, 138-141 methods of study, 21-26 of albatross, 100, 103 of Arabian horses, 42 of Ardea cinerea, 99 of bats, 103 of bees, factors influencing, 231-246 Lifespan of bison, 105 of birds in captivity, 103 of birds in Nature, 90-105 of bulls, 64-65 of Buteo buteo, 99 of buzzards, 99 of Capreolus capreolus, 92-94 of caribou, 92-94 of cat, 134 ofcattle, 57-65, 70-71 of Cervus elaphus, 92-94 of Chiropodomys gliroides, 96 of Citellus pygmaeus, 97 of cockroaches, 253-254 of deer, 91-94 of Dipodomys heermani, 96 of dog, 134 of Drosophila subobscura, 161, 266 of English thoroughbred horses, 35-56 of finch, 99, 103 offish, characteristics of long, 217-218 offish in captivity, 212-230 offish in Nature, 142-180 offish, of different species, 219-224 of flies, 105, 249-262, 287-288 of Hafling mares, 43 of herons, 99 of Hokkaido ponies, 43 of horses, 35-56, 65-69 of houseflies, 249-262, 287-288 sex differences in, 255-259 of human beings, actuarial aspects of, 2-20 of humming-birds, 104 of insects, 105 of kangaroo rats, 96 of Lagonostica senegala, 99 of Lapitsa horses, 43 of mammals in Nature, 90-105 of man and woman compared, 10-11, 16 of Megadyptes antipodes, 100 of mosquitoes, 105 ofmouse, 95-96, 287 of Myotis mystacinus, 97 of Odocoileus hemionus, 92-94 of Ovisdalli, 91-94 of owls, 99 of parent and progeny correlated, 47- 49 of Par us major, 100 of Passer domesticus, 99 of Passerines, 99-100 of Peromyscus leucopus, 95-96 of rabbits, 97 of Rangifer articus, 92-94 ofrats, 95-96, 251-252 effect of diet, 78 effect of disease, 72-89 of sheep, 91-94 of shrews, 97 of sousUk, 97 of Sorex araneus, 97 322 Subject Index Lifespan of sparrows, 99 of Sterna hirundo, 97-99 of Strix aluco, 99 of swifts, 100, 103, 104 of terns, 97-99, 104 of tits, 100 of trout, 265 of ungulates, 91-95 relationship to index of cephalization, 123-125 relationship with growth, 160-166 Life tables, 9-11 limitations of, 11-12, 18 Liver, in fish, 215-216, 227 Longevity {see also Lifespan) onset of disease and, 72-89 Look-down, lifespan of, 221 Lovettia seali, mortahty rate of, 146 Lung disease, in rats, 82 Lungfish, lifespan of, 224 Mackerel, lifespan and size of, 1 54 Makeham's Law, 2, 302 Malayan flying barb, lifespan of, 224 Mallotus villosiis, mortality rate of, 146 Mammals, arteriosclerosis in, 109 lifespan of in Nature, 90-105 life tables of, 117-118 relationship of brain and body weight to lifespan, 115-139 Man, disease in, effect on lifespan, 72- 89 lifetime energy expenditure of, 127 Mandibular glands, in bees, 233 Maternal age, effect on lifespan, 23-27, 31,32 Meal worm, effect of paternal age on life- span, 262 Megadyptes antipodes, lifespan of, 100 Melanogrammus aeglefinis, fecundity of, 198-200 sex organs in, 190 Metabolic disease, in fish, 216 Metabolism, effect on hfespan, 103-104, 124-126, 129, 136-137, 169-170 Mice, amyloidosis in, 80-81 effect of diet on, 79-80 lifespan of, 287 white-footed, lifespan of, 95-96 Miller's thumb, lifespan of, 223 Minerals, in pollen, 237 Minnow, lifespan and size of, 1 52 Mitotic inhibitors, 136 Molecular chanjges in ageing, 129-131 Monkeys, arteriosclerosis in, 108 Moonfish, lifespan of, 221 Mortality, relation with age and growth, 160-166 theory of, 127-129 Mortality rates, 286-296 laws governing, 2-5 mathematical models for, 302-311 sex difi"erences in, 81 Mosquito, effect of diet on, 255 lifespan of, 105 Mosquito fish, mortality rates of, 167 reproduction and senescence in, 187- 189 Mothers, age of, effect on lifespan, 23-27, 31,32 Mugiloidei, lifespan and size of, 154 Musca domestica {see Housefly) Muscular degeneration, 77 in rats, 74-76 Muskallunge, lifespan of, 222 Myocardial degeneration, 77 in rats, 74-76 Myotis mystacinus, lifespan of, 97 Neoplasia, in fish, 216, 226 Neothunnas macropterus, fecundity of, 193 Nephrosis, in man, 77 in rats, 74-76, 83 Nerve cells, in bees, 234, 244 Nitrogen, in pollen, 236 Odocoileus hemionus, lifespan of, 92-94 Oncorhynchus nerka, fecundity in, 197- 198 growth of, 156, 157, 158 Oncorhynchus spp., mortality rates of, 145 Ophiocephaliformes, lifespan of, 224 Orange chromide, Hfespan of, 224 Ovaries, effect of age on in fish, 201 growth of in herring, 195 in bees, 233, 235, 237, 238, 239, 243 of Drosophila, 275-277, 285 in fish, 193 changes due to age, 182 Overwintering, 232-234, 240 Ovis dalli, lifespan of, 92-94 Owls, lifespan of, 99 Palometa, lifespan of, 220 Parasites, in fish, 213-216, 228 Parental age, effect on fertility of off- spring, 56 effect on lifespan, 21-34, 43-47, 53, 259-262 Parental death, in fish, 189 Parrots, arteriosclerosis in, 109, 110 Par us major, lifespan of, 100 Passer domesticus, lifespan of, 99 Passeres, arteriosclerosis in, 109 Passerines, lifespan of, 99-100 Pelecaniformes, mortality rate of, 101 Pelicans, arteriosclerosis in, 109 Penguin, mortality rate of, 100 Perca fluviatilis, survival curves of, 143, 144 Perch, effect of diet, 178 lifespan and size of, 154, 220, 223 survival curves of, 143, 144, 145 Perciformes, lifespan of, 220, 223, 224 Percoidei, lifespan and size of, 1 54 Peromyscus leucopus, lifespan of, 95-96 Subject Index 323 Pharyngeal glands, in bees, 233, 235, 237, 238, 239, 244 Physiological function, stability of, 135 Pike, lifespan of, 222 Pike-killie, lifespan of, 224 Pilotfish, lifespan of, 221 Plaice, fecundity of, 202-206 growth of, 157, 158, 159 lifespan and size of, 1 50 mortality rates of, 167 reproduction and growth in, 189-190 Pleuronectes platessa, fecundity in, 202- 206 growth of, 157, 158 reproduction and growth in, 189-190 Pleuronectoidei, lifespan and growth of, 161-166 lifespan and size of, 148 Poeciliidae, reproduction and age in, 181, 186-189 Pollen, content of, 237 eflfect on bees, 236-240 Pompano, lifespan of, 220 Porcupine fish, lifespan of, 221 Porgy, hfespan of, 221 Poultry, arteriosclerosis in, llO, 113 Procellariiformes, mortality rate of, 101 Protein, effect on bees, 234, 254 effect on Drosophila, 266-267 in diet, 252-254 in pollen, 237. 238 storage in bees, 237, 238, 239 utilization of in fish, 210 Psittaci, arteriosclerosis in, 108-109, 110 Pudding wife, lifespan of, 221 Puffer, lifespan of, 221 Pygosteus pungUius, growth of, 156, 157, 158 Quillback, lifespan of, 222 Rabbits, cottontail, lifespan of, 97 Rajiformes, lifespan and size of, 154, 220 Rangifer articus, lifespan of, 92-94 Rasbora, lifespan of, 224 Rats, brown, lifespan of, 95-96 causes of death of, 82-83 disease in, effect on lifespan, 72-89 efiFect of diet on, 78, 251-252, 254, 265, 268 kangaroo, lifespan of, 96 lifespan of, 95-96, 251-252 lung disease in, 82 nephrosis in, 83 Rattus rattus, lifespan of, 95-96 Ray, lifespan and size of, 154, 163, 221 Red hind, lifespan of, 220 protein metabolism in, 183-185 Regeneration, in guppies, 208-210 Reproduction, effects of age on in fish, 181-182, 186-206 effect on Ufespan, 179, 275-279, 284- 285 in fish, 170-174 Roach, lifespan of, 222 Rockfish, lifespan of, 220 Rosy tetra, lifespan of, 224 Rubner's theory of ageing, 125-126, 128 Rudder fish, hfespan of, 221 Sailfish, lifespan and size of, 154 Salivary glands, in bees, 233, 235, 237, 238, 239 Salmo gairdneri, growth rates of, 182 Salmon, fecundity of, 192, 197-198 growth of, 156, 157, 158 lifespan of, 222 hfespan and growth of, 160-166 lifespan and size of, 150, 152 mortality rates of, 145, 166 natural death in, 174 relation of size and maturity, 172 reproduction in, 179 Salmonoidei, lifespan and growth of, 161- 166 hfespan and size of, 150 Salmo salar, fecundity in, 197-198 Salve! in us alpinus, hfespan of, 147 Sand dab, lifespan and size of, 148 Sardina pilchardis, eflfect of environment on, 168 Sardines, effect of environment on, 168 Sanger, hfespan and size of, 154 Scandinavians, lifespan of, 21-26 Scat, lifespan of, 221 Scombroidei, lifespan and size of, 1 54 Sea horse, lifespan of, 147, 154 "Senile "death, in fish, 189 Sex differences, in growth of fish, 1 59 in mortahty among fish, 145, 166-167 in reactions to environment in fish, 178 Sex organs, changes in due to age in fish, 218 in fish, 229 relationship to body weight in fish, 193-196 Sexual maturity, eflfect of diet, 84-85 in fish, 170-171 Shad, fecundity of, 191-192 Shark, lifespan of, 220 Sheep, hfespan of, 91-94 Sheepshead, lifespan of, 221 Shrews, lifespan of, 97 Siluroidei, lifespan and size of, 154 Size, eflfect on hfespan, 147-159 Skipjack, metabolism of, 170 Smelt, lifespan of, 147, 150 Snake-head, lifespan of, 224 Snapper, lifespan of, 221 Sole, hfespan and size of, 150 Sorex araneus, hfespan of, 97 Souslik, lifespan of, 97 Soya flour, effect on bees, 235 Spadefish, lifespan of, 221 Sparrows, lifespan of, 99 Sphenisciformes, mortahty rate of, 101 Sprat, lifespan and size of, 148 324 Subject Index Squirrel fish, lifespan of, 220 Stallions, causes of death in, 68 lifespan of, 65-67 Sterility, in cows, 63, 64 Sterna hirundo, lifespan of, 97-99 Stickleback, effect of temperature on, 168 growth of, 156, 157, 158 lifespan of, 152, 220, 223 Stress, 88 Strigiformes, mortality rate of, 101 Strix aluco, lifespan of, 99 Sturgeon, growth of, 156, 157, 158 lifespan of, 147, 220, 222, 226 lifespan and growth of, 163 lifespan and size of, 152 oldest age of, 191 reproduction in, 171 survival curves of, 143, 144 Sugar, effect on bees, 238 Sunfish, effect of environment of, 168 lifespan of, 223 protein metabolism in, 182-183 Swifts, lifespan of, 100, 103, 104 Syngnathiformes, lifespan and size of, 154 Tarpon, lifespan of, 220 Temperature, effect on bees, 245, 246 effect on cell, 281 effect on lifespan, 167-168 effect on lifespan in Drosophila, 271- 279, 283, 284 effect on protein metabolism, 185-186 Tench, lifespan of, 223 Tenebrio molitor, effect of paternal age on lifespan, 262 Terns, lifespan of, 97-99, 104 Tetrodontlformes, lifespan of, 221 Thunniformes, lifespan and growth of, 163 lifespan and size of, 154 Thymallus signifer, effect of environment on, 168 Tits, lifespan of, 100 Toadfish, lifespan of, 221 Totoaba, lifespan and size of, 154 Trigger fish, lifespan of, 221 Trout, effect of diet on, 169, 253, 254 effect of reproduction on, 179 fecundity in, 192, 197 food conversion in, 186 growth rates of, 1 82 lifespan of, 222, 265 lifespan and size of, 1 50, 1 52 mortahty rates of, 145 Tuberculosis, in cows, 63, 64 Tuna, fecundity in, 193 lifespan and size of, 1 54 Ungulates, lifespan of, 91-95 Vitamins, effect on bees, 238 in pollen, 237, 238, 239 Vultures, arteriosclerosis in, 110 Walleye, fecundity/length relationship in, 197 Wasps, 232, 255 Wax glands in bees, 233, 239 Weakfish, lifespan of, 221 Whitebait, lifespan and size of, 150 mortality rate of, 146 Whitefish, growth of, 156, 157, 158 lifespan of, 222 lifespan and size of, 1 50 survival curves of, 143, 144 Whiting, lifespan and size of, 154 Yellow tail, lifespan of, 221 Printed by Spottiswoode, Ballantyne db Co. Ltd.^ London and Colchester CUMULATIVE INDEXES TO VOLUMES 1-5 AUTHOR INDEX Numbers in bold type indicate volume number. Plain numbers indicate a contribution either in the form of an article or as a contribution to the discussions. Italic numbers indicate a reference to an author's work. Aas, K Aasen, O. . Abbot, W. E. . Abderhalden, E. Abercrombie, M. Abess, A. T. Abraham, K. Ackermann, P. G Adair, F. . Adair, F. L. Adams, C. E. Adamsons, K. Addis, T. Adlersberg, G. D Adolph, E. F. . Aebi, H. . Ahlman K. L. Akyuz, E. Albertini, A. von Albrectsen, S. R. Albright, F. Alex, M. . Alexander, F. Alexander, J. D. Alexander, L. C. Alexander, M. O. Alexejev-Berkmann, I. Allee, W. C. Allen, E. . Allison, A. C. Allison, J. B. Allott, E. N. Ally, M. S. Aim, G. . Alpatov, W. W. Altman, K. I. Alving, A. S. Amatruda, T. T. . 4, 275, 294 5, 149, 156, 174 . 4, 109, 113 4, 201, 202, 205 . 3, 70 . 4, 727, 134 . 1, 33, 48 1, 110, 121, 122; 4, 241, 246 2, 120, 121, 125, 127 . 1, 151, 158, 160 . 1, 143, 159 4, 92, 93 4, 254, 257, 259, 261 . 3, 136, 142 4, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 34, 34, 60, 75, 94, 95, 96, 97, 152, 162, 195, 196, 221, 228, 225,311,575 . 1, 77 . 4, 74, 76 . 5, 149, 174 1, 53, 55, 108 4, 67, 68, 69, 72 1, 109, 122 1, 91, 99, 103; 3, 97, 100 . 3, 757, 142 . 4, 275, 295 . 4, 289, 296 4, 40, 41, 57 A. . 4, 287, 295 1, 10, 14 2, 38, 50, 53 . 2, 235, 238 . 4, 757, 134 4, 38, 39, 41, 43, 56 . 5, 255, 264 5, 144. 155, 159, 168, 174 . 3, 24, 29 2, 277, 257, 234, 238, 239 4, 243, 244, 246, 248 4, 75 Amoroso, E. C. 1, 7^2, 159; 2, 1, 16, 27, 66, 67, 68, 85, 705, 106, 113, 115, 117, 125, 127, 128, 145, 755, 160, 126, 755, 241, American Academy of Pediatrics . 4, 160, 161 Ames, R. G. . . .4,6,11,97 d'Ancona, U Andersen, J. Anderson, A. R. Anderson, J. Anderson, J. L. Anderson, W. E. Andersson, B. Andjus, R. K. Ando, S. . Andrew Andrew, W. Andrews, M. C. Anitschkow, N. Anner, G. Ansell, S. Anslow, W. P. Anthonisen, P. Anthony, H. E. Appelboom, J. W Appelget, J. Arai, H. . Araki, Y. . Arataki, M. Aristophanes Armbruster, L. Ariiey, S. F. Aron, C. . Aron, M. . Arons, W, L. Arora, H. L. Artunkal . Arvidsson, V. B. Ascher, B. M. Ascher, K. R. S. Asdell Asgood, H. S. Ashton, N. Assheton, R. Astbury, W. T. Astrup, P. . Aub, J. C. . 175, 200, 240. 243, 246, 247, 253 . 5, 7<57, 174 5, 92, 94, 95 102 . 4, 301, 308 5, 232, 234, 241 . 3, 178, 181 4, 720, 727, 128, 131, 135 4, 37, 42, 56 . 1, 765, 169 4, 68, 69, 72 4, 225 . 5, 248, 264 2, 129, 132, 143 1, 89; 3, 707 4, 757, 795 5, 27, 57 4, 277, 298 4, 41, 56 2, 184, 187 4, 24, 31 5, 755, 77^ 2, 32, 34, 53 4, 68, 69, 72 . 4, 257, 259 . 1, 34 5, 232, 241, 242 . 2,206,209 2, 36, 37, 38, 39, 41, 42, 46, 47, 53, 60, 63, 66 2, 60, 63, 66 . 4, 109, 112 . 5, 149, 174 . 2, 84 3, 759 . 3, 752, 143 5, 255, 259, 264 . 1, 30 . 5, 257, 264 . 2, 30 . 2, 755, 755 . 3, 96, 100 . 4, 41, 56 1, 52, 54, 77, 104, 105, 124, 125, 205:2,775, 779, 750, 755 11 Cumulative Author Index Aumonier, F. J. Austin, C. R. Austin, O. L. Austin, O. L. Jr. Axelrad, B. J. Bailey, M. . Baillie, W. H. ' Bainey, J. D. Baird, D. Baker, B. L. Bale, W. F. Balfour, W. M Balinsky, B. Ball, M. R. Ball, Z. B, Ballon, J. Balo, J. . 3, 56 . 2, 86, 96 . 5, 98, 102 . 5, 98, 102 4, 292, 294, 297 Babcock, H. . . .3, 179, 181 Babkin, B. P. . .4, 62, 72 Back, E. . . .5, 235, 238, 241 Baclesse, M. . . .2, 168, 172 Bacon, F 3, 9, 18 Bacsich, P. .1, 142, 159; 2, 41, 53 Bagenal, T. B. . 5, 192, 193, 200, 201, 202, 205, 207 4, 146, 147, 149 3, 24, 29 4, 109, 112 . 5,33 4,260 2, 263, 238 2, 236, 238 1, 37, 38, 48; 3, 178, 181, 186 . 4, 114, 291, 297 . 1, 180, 185 1, 219, 223, 224, 225, 227, 232, 233, 236 1, 98, 99, 103, 104, 107; 3, 32, 44, 70, 97, 98, 100, 101, 101, 103 Banfield, A. W. F. . . 5, 92, 102 Banga, I. Banta, A. M. Barber, J. K. Barbour, A. Barcroft, Sir J. Barker, J. P. Barker, S. B. Barlow, J. S. Barnes, B. A. Barnes, L. L. Barnes, R. H. Barnet, H. A. R, Barr Barrett, G. R. Barrett, M. K. Barron, D. H. Barrows, C. H. Bassett, D. L. Bates, D. V. Bauer, K. . Bauer, L. S. Bartholomew, R Bartlett, Sir F. C Barton, R. A. Bartter, F. C. Beams, H. W. 1, 98, 99, 103, 104, 107; 3, 65, 66, 67, 97, 98, 100, 101, 103 3, 24, 29 4, 89, 92, 292, 293, 294, 297 . 4, 265, 268 2, 126, 151, 158; 5, 133 4, 3, 4, 11 . 2, 130, 143 . 4, 106, 112 . 4, 49, 56 1, 188, 193, 197, 200; 3, 8, 18; 5, 251, 264 . 1, 180, 185 5, 5, 15 . 1,89 . 1, 149, 160 . 3, 123, 124 . 2, 118, 125 . 4, 245, 246 . 2, 60, 66 1, 58, 62, 64, 65 . 5, 59, 69 . 4, 133, 134 . 2, 112, 113 1, 41, 202, 209, 214,215,216,217,218,2^5 . 5, 53, 54 4, 46, 57, 89, 92, 196, 280, 292, 293, 294, 297 . 2, 86, 96 Bean, H. W. Bean, W. B. Beard, R. E. Beck, J, C. Becker, W. H. 4, 120, 121, 131, 135 1, 78, 79, 80, 87, 105, 171 5, 4, 14, 288, 289, 302, 303, 305, 307, 308, 311 . 4, 293, 294 . 4, 275, 295 Beeton, M. 1, 238; 5,21, 31, 33, 53, 54 Belding, D. L Bell, F. H. Bell, R. G. Bendell, J. F. Benedict, F. G. Beneke, R. Benjamin, B. Bennett, H. S. Bensley Bercu, B. A. Berdegue. J. Berg, B. N. Berg, L. S. Bergerard, Y. Bergmann, K. Bergstrom, J. Bergstrom, W. H. Berkeley, A. Berman, C. Bernard, C. Bernhard, K. Bernstein, F. Bertalanffy, L. von Berthet, J. . Bertholf, L. M. . Bertin, L. . Bertucio, M. Best, A. T. Best, C. H Beutler, R. Bevan, W. . Bezem, J. J. Biale, J. B. Bichat, X. Bidder, G. P. Bienvenu, B. Bilewicz, S. Billeter, J-R. Billingham, R. E 5, 197, 207 . 5, 151, 177 . 1, 196, 200 . 5, 101, 102 3,9, 18;4,138;5,124 . 5, 110, 112 . 1, 198, 199; 5, 2, 15, 16, 17, 18, 19, 32, 34, 70, 293 . 2, 106, 112, 114, 129, 144 .2,103 .4, 280, 294 . 5, 755, 175 . 5, 16, 32, 33, 72, 73, 74, 75, 76, 79, 79, 80, 82, 83, 84, 85, 86, 87, 88, 114, 137, 265, 266, 285 . 5, 212, 225 . 2, 11, 13 . 4, 284, 286, 291, 296 4,298 . 4, 120, 134 1, 219, 223, 224, 225, 226, 227, 232, 233, 236 . 3, 106, 114 1, 205; 2,23,27, 108, 113, 118, 137 . 2, 177, 183 3, 195 5, 157, 169, 175 . 2, 136, 143 . 5, 235, 241 . 5, 168, 175 . 4, 140, 141, 146, 149 5, 104 3, 1, 36, 46, 49, 101, 103, 126, 127, 128, 146, 147, 182, 188, 192, 195 . 5,235,236,241 . 4, 127, 134 Beverton, R. J. H. 5, 142, 149, 151, 158, 175, 111, 178, 179, 226, 227, 229 . 5, 97, 103 . 2, 208, 209 1, 245 1,28; 3, 9, 18; 5, 173, 175, 182, 189, 190, 191, 207 . 1, 198, 200 . 5, 277, 281 . 4, 181, 195 1, 170; 2,229,231; 3, 116, 124 Cumulative Author Index Billis, L. de Binet, A. . Birch-Andersen, A. Birchard, W. H. Birren, J. E. Bischoff, F. Bishop, L. R. Biskind, G. S. Biskind, M. S. Bittner, J. J. Bjorkman, N. Bjorksten, J. Black, A. Black, D. A. K. . 4, 242, 246 . 3, 176, 182 . 4, 126, 136 . 4, 51, 56 . 1, 38, 48 4, 262 . 2, 206, 209 . 1, 153, 160 . 1, 153, 160 . 1, 155, 160 . 2, 155, 156, 158 . 3, 66, 67 . 4, 127, 135 4, 13, 38,40,41,50, Blackburn, M. Blackman, F. F. Blair, W. F. Blake, W. D. Blalock, A. Bland, J. H. Blandau, R. J. Blaxter, K. L. Blazka, P. Blegen, E. Bloch, E. . Bloom, W. L. Blum, H. F. Boas, E. P. Bodenheimer, F Bottcher, T. Bogdonoff, M. D 56,57,59,59,15, 113, 150, 150, 163, 208, 224, 226, 247, 248, 264, 267, 268, 269, 310, 317 5, 146, 149, 151, 175 2, 205 . 5, 96, 102 4, 275, 287, 294 . 4, 287, 296 4, 294 1, 149; 2, 81, 83, 85 4, 301, 307, 308, 309 5, 170, 175 4, 273, 294 4, 90, 93 4, 127, 134 2, 228, 231 3, 136, 142; 4, 136, 142 5,294 5, 60, 61, 69 1, 110, 120, 121, 122; 3, 75, 90; 4, 225 Bogomolets, A. A. . . 3, 133, 142 Boling, E. A. . 4, 102, 106, 108, 113 . .2, 191, 198 . 4, 109, 113 1, 42, 48; 3, 137, 143, 173, 182 . 4, 257, 259 M. . .4, 80, 92 5, 119, 123, 128, 132 . 2, 208, 209 4, 89, 92, 93 . 4, 132, 135 3, 117, 118, 120, 123, 124, 125 . 5, 170, 175 4, 60, 101, 114, 196, 248,260, 261, 269,298 . 4, 200, 205 . 3, 151, 168 . 2, 56 2, 183, 210, 211,247, 247,24S,249;3, 10, 17,18, 20, 28, 2P, 31, 129; 5, 70, 84, 91, 96, 102, 103, 104, 105, 112, 113, 119, 132, 137,217,225,221,228,245 3, 34, 35, 47, 50, 68, 72, 130, 144, 188, 189 Bommer, S Bondurant, J. Bondy, E. Bendy, P. K. Bongiovanni, A. Bonin, G. von Bonner, J. . Boog, J. M. Booher, L. E, Boot, L. M. Boroughs, H. Borst, J. G. G. Borum, E. R. Botwinick, J. Bounoure . Bourliere, F. Bourne, G. H. Boyd, E. M. Boyd, H. . Boyd, J. D. Boyle, A. J. Boyne, A. W. Bozic, B. . Brachet, J. Bradley, G. P. Bradley, S. E. Brambell, F. W. R. Brand, F. C. . 2, 122, 125 . 5, 101, 102 . 2, 185, 186 . 4, 289, 296 . 1, 196, 199 2, 11, 13, 17 . 1, 192, 193 . 4, 274, 294 4, 274, 275, 294 2, 34, 50, 53, 57, 60, 66, 81, 83 . 4, 23, 32 Brandfonbrener, M. 1, 237; 3, 77, 84, 88, 89, 90, 91 ; 4, 234, 236, 246 Brandt, F. A. Brannon, E. S. Breder, C. M. Breed, E. S. Bretherick, O. Breward, M. M Brewer, W. D. Bricker, N. S. Brien, P. . Briggs, A. P. Briscoe, W. A Brod, J. Brodsky, W. A. Brody, M. B. Brody, S. Bronks, D. Brooks, L. Bro-Rasmussen, F Brown, A. N. Brown, C. J. D. Brown, I. W. Brown, L. M. Brown, M. E. Brown, R. A. Brozek, J. Brubacher, Z. Bruce, R. A. Brues, A. M. Bruhin, H. Brull, L. . Bryden, J. G. Bryuzgin, V. L. Buchanan, A. D. Buchborn, E, Bucher, G. E. Bucht, H. . 3, 102 4, 279, 297 5, 272, 217, 218 225 4, 274, 297 5, 237, 243 2, 33, 53 1, 722, 725 4, 45, 56 1,244 4, 275, 277, 295 . 4, 264, 268 4, 248, 271, 272, 273, 279, 281, 284, 286, 288, 295, 296 Buchwald, N. A. Buehl, C. C. Buehler, C. Buettner, K. Bujard, E. . Bull, G. M. Bull, J. P. . 4, 24, 31, 40, 46,57 . 3, 775, 182 1, 188, 189, 190, 193, 197, 199; 5, 124, 132 3, 70, 96, 100 . 4, 109, 113 . 4, 63, 65, 68, 73 . 4, 131, 135 3, 26, 29; 5, 168, 175 . 2, 234, 238 . 4, 46, 57 5, 147, 148, 149, 175, 217, 225 3, 757 4, 757, 757, 752 . 2, 777, 183 . 4, 46, 56 5, 138 2, 777, 181, 183 1, 27, 68, 123, 124, 202, 207, 238, 241 . 3, 707, 114 . 3, 77, 18 5, 59, 69 . 4, 280, 292, 293, 295 . 5, 258, 264 4, 273, 279, 280, 283, 295, 298 3, 755, 7^2 1, 198, 201 . 1,49 4,100 2, 42, 53 4, 265, 268 4,98,95, 115, 150, 151, 163,207,228,249,269 Cumulative Author Index Bullough, w. s. Bulmer, M. G. Bunge, R. G. Bunting, H. Burch, H. . Burd, A. C. Burgen, A. S. V. Burger, M.. Burgess, E. W. Burgos, M. H. Burkl, W. . Burn, G. B. Burnet, F. M. Burns, C. M. Burrill, M. W. Burrows, H. Burton, D. Bush, I. E. Butler, A. M. Butler, C. G. Butt, W. R. Buxton, P. A. Byers, S. O. Bykov, K. M. Byrom, F. B. 2, 4, 2, 38, 53, 64, 66 4, 65, 69, 73, 74 1, 171 2, 191, 192, 198 1, 197, 199 5, 149, 175 4, 64, 73 3, 6, 18 1, 43, 48 2, 86, 87, 96 36, 42, 43, 44, 53 . 2, 235, 238 . 2, 228, 231 . 5, 21, 31 . 2, 12 2, 49, 51, 53 . 3, 97, 100 . 4, 181, 194 139, 141, 147, 149 5, 237, 241, 243 . 1, 130, 137 . 5, 240, 241 . 3, 137, 142 . 4, 287, 295 . 4, 258, 259 Caldwell, M. J. . Caligaris, L. C. S. Calloway, D. H. Calvin, M. Cameron, G. R. Cameron, J. W. McB Cameron, W. Campbell, H. A, Campbell, R. M. Canivenc Cannon, C. Cannon, P. R Capek, K. . Card, W. I. Carlo, J. . Carlsen, E. Carlson, A. J. Carlson, L. D. . Carter, S. B. Castle, W. E. Cathcart, E. S. . Caton, J. D. Caton, W. L. Cavan, R. S. Cavanaugh, M. W. Cavazzana, P. Challice, C. E. . Chalmers, T. A. . Chaloupka, M. . Chang, H-W. Chaplin, H. . 4, 200, 205 . 4, 89, 93 4, 131, 134, 135 . 2, 223, 232 1, 16, 26, 54, 55, 66; 3, 93, 100 5, 258, 264 . 1,57 1, 197, 200 2, 166, 172 2,175 5, 59, 69 4, 133, 134 4, 165, 179 4, 301, 309, 310, 317 1, 126, 137, 219, 223, 224, 225, 226, 227, 232, 233, 236; 4, 90, 93 4, 46, 56, 264, 268 1, 197, 199, 206 5,89 3, 51, 59 Chaplin, J. P. Chapman, R. N. Chart, J. J. 1, 155, 160 4, 283, 295 2, 779, 183 2, 166, 172 1, 43, 48 1,3, 10, II, 13 2, 7P7, 196, 198 . 2, 86, 96 . 5, 56 . I, 122, 123 5, 145, 146, 153, 175 1, 164, 169; 2, 216, 226, 231, 232, 236, 238 . 4, 755, 179 . 1, 7P5, 199 4, 193, 194, 195, 296 Chase, H. B. Chasis, H. . Cheek, D. B. Chesley, L. C. . Chevremont Chibnall, A. C. . Chipman, W. A. Chitty, H. Chow, B. F. Christensen, P, R Christie, R. V Chugaeva, M. Ciba Foundation Claesson, L. Clara, M. . Clark, F. N. Clark, H. W. Clark, J. . Clarke Clarke, J. M. Clarke, R. D. Clay, H. M. Clermont, Y. Cloette Cochran, W. Cockrum, E. L. Cohen, A. . Cohen, B. L. Cohen, D. M. Cohen, H. W. Cohen, J. . Cohn, J. E. Cohrs, P. Cole, D. F. CoUazos, Ch. J. I. Colonge, R. M. A Comfort, A Comroe, J. H Conn, J. W. Conrad, H. S. Conway, E. J. Cook, C. D. Cooke, R. E. 2, 191, 192, 194, 199 . 4, 230, 246 4, 105, 109, 113, 120, 121, 132, 133, 134 4, 82, 84, 89, 92, 93 2, 103 . 2, 204, 209 . 5, 170, 175 3, 10, 18; 5, 105, 296, 298 1, 198, 199; 3, 79, 91 ; 4, 245, 246 . 1, 40, 48 1, 58, 57, 62, 65, 65, 66, 67, 68, 246 . 5, 53, 54 . 5, 128, 132 . 2, 60, 66 . 1,54 5, 149, 155, 175 . 5, 272, 225 . 1, 239, 240 5, 266 . 5, 278, 281 5, 4, 5, 6, 7, 8, 10, 11, 14, 15 . 3, 155, 162, 168 . 2, 86, 96 . 2, 159, 160 4, 139, 141, 147, 149 . 3, 27, 29 2, 22, 23, 26 . 4,202,203,205 . 5, 757, 775 . 4, 57, 57 4, 132, 133, 135, 227 . 4,243,244,246 . 5, 770, 772 . 4, 193, 194 . 1, 109, 121, 122 2, 3, 13, 84, 168, 172 . 1, 28, 29, 30, 51, 52, 107, 138, 171, 204, 238, 241, 242, 243, 244, 245; 2, 215, 231; 3, 2, 7,11,12, 18, 29, 30, 36, 37, 38, 46, 68, 118, 124; 5, 17, 17, 32, 34, 35, 36, 37, 43, 44, 47, 52, 53, 54, 55, 56, 70, 71, 84, 85, 87, 88, 103, 104, 113, 114, 775, 777, 779, 752, 134, 136, 137, 759, 77i, 775, 178, 179, 208, 208, 209, 210, 277, 225, 111, 129, 230, 244, 262, 264, 265, 266, 286, 286, 289, 291, 294, 295, 298 . 4, 264, 268 . 4, 44, 56, 76, 280, 295 . 3, 777, 182, 186 4, 22, 23, 24, 31, 35, 99, 203, 205, 207, 226 . 4, 705, 775, 143, 149 . 4, 133, 134 Cumulative Author Index Cooke, W. T. Cooper, A. R. Cooper, E. L. Cooper, I. S. Cooper, K. E. Cope, O. . Copenhaver, J. H. Coquoin-Carnot, M Corbo, S. . Corey, E. L. Corkins, C. L. Cornaro, L. Combleet, T. Comer, G. W. Corsa, L. Cort, J. H. Cotes, P. M. Cotlove, E. Count, E. W. Court, D. M. Coumot, L. Courrier, R. Coville, F. E. Cowan, G. A Cowdry, E. V. . 4, 109, 112 4, 120, 121, 135 5, 144, 153, 176 4, 39, 40, 41, 43, 56 2, 120, 121, 125 4, 49, 56 4, 209, 219 4, 217, 219 1, 199, 200 2, 109, 113 5,245 1, 194, 200 2, 197, 198 2, 15, 55, 56,64, 66, 67, 246, 252 4, 105, 113, 114, 143, 149 4, 33, 45, 56, 214, 219, 222, 278, 283, 287, 288, 289, 291, 292, 295 . 2, 162, 172 . 4, 129, 135 5, 119, 123, 132 . 3, 95, 100 4, 262 2, 126, 168, 172 . 4, 133, 134 . 3, 55, 56 1, 24, 26, 27, 50, 53, 55, 88, 103, 108, 171, 172, 204, 205, 207, 242, 243 Cowie, D. B. . . 2, 111, 112, 113 Crabb6, J 4, 280, 298 Cramer, W. . 2, 23, 27, 122, 125 Crane, W. A. J. . . . 4, 260 Cravioto Munoz, J. . .1, 198, 200 Crawford, B. . . .4,51,56 Crawford, H. 2, 216, 226, 231, 237, 238 Crawford, J. D. . 4, 139, 140, 141, 146, 147, 149 Cresseri, A. . . .3, 137, 142 Crevier, P. H. Crosley, A. P. Crowell, M. F . 4,39,40,43,56 . 4, 289, 297 1, 197, 200; 3, 25, 29; 5, 169, 176, 251, 253, 264 Crozier, W. J. Cruickshank, D. H. Grumpier, H. R. Cruz, W. O. Csapo, J. . Cummins, J. Cuny, G. . Currie, C. Curtis Curtis, R. H. Curzon, E. G. Cutbush, M. Cuthbertson, D. P. Dacie, J. V. Dack, S. . Daines, M. C. Dalton, K. 5,267 2, 203, 210 2, 112, 113 2, 236, 238 4, 155, 161 4, 46, 56 4, 242, 246 1, 198, 201 . 5,87 4, 292, 297 4, 131, 135 2, 216, 226, 231, 237, 238 . 4, 130, 135 . 4, 199, 205 . 3, 75, 90 . 4, 266, 268 . 4, 84, 93 Dalton, N. N. Daly, C. . Danielli, J. F. Danowski, T. S. Darrow, D. C. , Dasmann, R. F. Davey, W. P. Davidson, F. J. Davidson, J. N. Davies, B. M. A. Davies, C. Davies, D. Davies, H, E. Davis, D. E. Davis, G. K. Davis, J. E. Davis, J. M. Davis, J. O. Davis, W. S. Davson, H. Dawes, G. S. Dawson, A. B. Dean, R. F. A Deane, H. W. Deason, H. J. De Duve, C. Decider, C. L. Deevey, E. S. Dejdar, R. De Jongh, S. E. De la Blaze, F. A Del Castillo, E. B. . 2, 228, 232 . 5,21,31 . 3, 39, 43, 45, 46, 47, 48,49,50,68,95,129,130, 144, 148, 185; 5, 1, 15, 17, 19, 32, 33, 34, 70, 103, 104, 114, 133, 136, 137, 177, 178, 209, 229, 230, 281, 282, 298 . 4, 289, 295 4, 106, 113, 133, 134 5, 92, 94, 103 5, 282 3, 107 . 1, 191, 193 1, 192, 193, 209, 219 . 4, 266, 268 1, 220, 236; 3, 88, 90; 5, 72 . 4, 150, 266, *268 . 5, 95, 102 . 2, 148, 159 1, 206 4, 102, 106, 108, 113 . 4, 277, 295 . 5, 192, 207 4, 15, 20, 28, 29, 31, 32, 33, 34, 35, 58, 59, 73, 96, 99, 100, 101, 114, 162, 206, 207, 208, 269, 309, 312, 315, 316 2, 104, 115, 125, 126 . 2, 81, 83 . 1, 198, 200 2, 23, 26, 37, 53, 60,66 . .5, 151, 176 . 2, 136, 143 . 5, 159, 175 1, 10, 15; 5, 91, 102 . 4, 291, 296 . 2, 64, 66 . 2, 86, 96 1, 151, 160 Delea, C. 4, 89, 92, 292, 293, 294, 297 Deming, Q. B. . . 4, 280, 293, 297 Dempsey, E. W. . 1, 103; 2, 41, 54, 54, 57, 60, 66, 67, 68, 100, 103, 104, 106, 107, 109, 110, 111, 113, 114, 129, 144, 146, 147, 186, 191, 192, 198, 241, 243, 246, 248, 250, 251, 253 Dennis, W. H. . . .4, 24, 31 Dent, C. E. . . .2, 112, 113 Deringer, M. K. . 3, 121, 123, 124 Desaulles, P. A. . 4, 59, 60, 76, 94, 180, 181, 187, 188, 194, 195, 196, 197, 198, 206, 227, 228, 262, 316 Detkens . . .5,67, 69 Dettmer . . . .3, 102, 103 Devoid, F. ... 5, 149, 175 Cumulative Author Index Dewar, M. M. Deyrup, I. . Diamond, I. Dice, L. R. Dickerson, J. W. T Dickie, L. M. Dieckmann, W. J. Dietrich, H. Dilley, W. E. Dinkhauser, F. Dlouhd, H. Dobriner, K. Dobson, E. L. Dodson Doisy, E. A. Dole, V. P. Doljanski, F. Donohue, D. Donohue, D. M. Dorfman, R. I. Dornhorst, A. C. Dorst, J. Douglass, P. M. Downing, D. F. . Doyle, A. E. Dreizen, S. Drescher, A. N. Droop, M. R. Drummond, J. K D'Silva, J. L. Dublin, L. I. Dubois, A. B. Dubois, E. Duckworth, J. Duesberg, J. Duetz, G. H. Duncan, L. E Dunewitz, A. Dunham, L. J. Dunn, Th. B. Dunning, M. F. Dupret, L. Dyke, H. B. van Dymond, J. R. Dyrenfurth, I. 4, 10, 11 4, 24, 31 4, 24, 31 3, 10, 18 4, 217, 219 5, 151, 175 2, 112, 113 5, 60, 69 1, 197, 200; 5, 169, 176 . 5, 63, 69 4, 9, 10, 11, 165, 779 . 1, 133, 137 . 4, 289, 295 3, 185 2, 50, 53 4, 67, 73, 77 5, 208 . 4, 46, 56 . 2, 237, 238 1, 126, 130, 134, 137; 4, 90, 93, 187, 195 . 2, 234, 238 . 3, 27, 29 . 1, 156, 161 . 1, 198, 200 . 4, 278, 295 . 1, 198, 200 . 4, 145, 149 . 1, 187, 193 3, 107 2, 148, 148, 149, 150, 158 1,195,200, 203; 3, 85, 90, 137, 142 . 4, 264, 268 . 5, 123, 132 2, 184, 187; 4, 309 2, 86, 93, 96 . 5, 85 1, 139; 3, 78, 90; 4, 89, 92, 226, 280, 292, 293, 294, 297 . 1, 198, 200 1, 151, 158, 160 3, 121, 122, 124 . 4, 46, 56 . 2, 136, 143 . 4, 92, 93 5, 175 4, 89, 93, 293, 294 Eadie, G. S. Earle, D. P. Eberlein, W. R. Economou-Mavrou, Eddy, M. . Edelman, I. S. Edmonds, S. J, . Edmunds, T. R. . Eeg-Larson, N. . Eichelberger, L. . 2, 234, 238 4, 275, 295 4, 80, 92 4, 218, 219 1, 160 4, 36, 47, 56, 109, 113, 280, 295 . 5, 169, 175 . 1, 196, 200 . 1, 111, 114, 115, 120, 122, 123 . 4, 119, 135 Eichna, L. W. Eidrigevits, E. V. Einstein, A. Eisenberg, S. Ek, J. El-Deeb, A. L. A. Eliasch, H, Elkin, R. . Elkinton, J. R. . Elliott, R. H. EUis, R. S. Elmadjian, F. 225, Elrick, H. . Elsdon-Dew, R. Elvehjem, C. A. Ely, R. S. . Emerson, A. E. Engel, S. L. Engeler, W. Engfeldt, B. Engle, E. T. Engstrom, W. W Enzmann, E. V. Epstein, F. H. Erickson, D. W. Ershoff, B. A. Escher, D. J. W. Eschmeyer, P. H. Essen-Moller, E. Etheridge, J. Evan, J. V. Evans, B. M. Evans, C. A. Evans, H. . Evans, H. M. Evans, W. A. Evenius, C. Everett, N. B. Evermann, B. W Ewing, G. Fahmy Fahr, H. O. Failla, A. . Fainstat, T. D. Falbriard, A. Falek, A. Falk, G. . Falzone, J. A. Farber, S. J. Fargo, W. C. Earner, D. S. Farr, W. . 4,5, . 4, 275, 295 . 5, 53, 54 . 2, 227, 231 . 4, 283, 298 4, 273, 279, 280, 283, 295, 298 . 5, 232, 234, 241 4, 273, 279, 280, 283, 295, 298 . 5, 168, 176 4, 38, 46, 56, 289, 295, 297 . 2, 120, 125 . 5, 248, 264 I, 219, 219, 223, 224, 226, 227, 230, 232, 233, 234, 236 . 1, 199, 200 3, 188 1, 197, 200; 4, 133, 136 . 4, 105, 113 . 1, 10, 14 4, 92, 93, 257, 259 . 5, 59, 69 . 2, 162, 172 1, 151, 160; 2, 60, 66 . 4, 42, 56 2, 36, 37, 54 4, 46, 57 5, 196, 197, 208 . 1, 197, 201 4, 277, 280, 295, 297 . 5, 197, 207 . 3, 134, 142 . 4, 109, 113 4, 201, 202, 203, 205, 221 4, 265, 268 5,297 5, 92, 102 1, 29: 2, 41, 53 4, 243, 246 5, 232, 241 2, 34, 53 5, 212, 226 4, 288, 296 2,128 5, 110, 112 5,289 2, 19, 26, 168, 172 4, 292, 293, 295 3, 132, 139, 143 II, 12, 13, 167, 169, 172, 179 . 3, 77, 90 . 4, 275, 295 . 1, 198, 201 5, 97, 101, 102, 103 . 5,9 Cumulative Author Index Farran, G. P. . . .5, 194, 207 Farrar, C. L. . . 5, 222, 234, 241 Fawcett, D. W. . 2, 86, 87, 96, 97, 98, 99, 103, 104, 109, 114, 129, 144, 177, 183, 250 Fazekas .... 3, 147 Feaster, J. P. . . .2, 148, 159 Feingold, L. . 1, 42, 48; 3, 137, 143 Feingold-Jarvik ... 3, 139 Fejfar, Z. . 4, 32, 33, 115, 163, 248, 261, 271, 271, 272, 273, 279, 281, 282, 284, 286, 291, 295, 296, 298, 299, 300, 317 Fejfarovd, M. . 4, 281, 284, 286, 291, 295, 296 Fekete, E. . 1, 146, 150, 156, 160 Feldman-Muhsam, B. . .5, 259, 264 Fell, H. B, Fencl, V. . Femgold, L. Fielding, U. Figuerson, W. G. Filer, L. J. Finch, C. A. Finck, M. A. von Findlay, G. H. Finkel, M. P. Finnell, J. Firket, H. Fischer, A. Fischer-Piette, E. Fisher, A. J. Fiske, C. H. Fister, H. J. Fitch, H. S. Fitch, J. E. Fitzgerald, M. G Flade, J. E. Fletcher, H. M. Flexner, L. B Flink, E. B. Fhnt, M. H. Flipse, R. J. Flower, S. . Floyer, M. A. Fliickiger, E. Foa, C. Foldi, M. . Foerster, R. E. Fogg, G. E. Folin Folkes, B. F. Folkes, J. P. FoUis, R. H. Forbes, G. B. Forbes, R. M. Forster, R. E. Forwell, G. D. 2, 4, 12 4, 278, 289, 295 . 3, 173, 182 . 2, 50, 53 . 4, 200, 205 . 4, 133, 134 2, 228, 229, 231, 232, 237, 238 4, 211, 212, 219 3, 102 . 2, 148, 158 . 5, 168, 176 . 2, 62, 66 . 3, 32 . 3, 25, 29 . 4, 98 . 4, 120, 135 . 4, 120, 135 5, 96, 97, 102 . 5, 755, 775 4, 301, 308, 309 . 5, 66, 69 . 4, 63, 73 2, 109, 111, 113, 118, 125, 148, 158, 159; 4, 26, 31 . 4, 301, 308, 309 . 3, 70 . 4, 200, 205 5, 77P, 128, 132, 168, 175, 217, 225 . 4, 258, 259 . 3, 64, 68 1, 757, 755, 160 4, 45, 57, 284, 286, 296 5, 757, 755, 775 . 1, 186, 193 1, 184 . 2, 208, 209 1, 757, 192, 193 4, 257, 259 4, 709, 775, 727, 755, 143, 149 4, 720, 727, 755 . 4, 264, 268 . 4, 65, 73 Foulds, G. ... 3, 750, 168 Fountaine, M. E. . . 3, 25, 29 Fouracre Barns, H. H. . 2, 774 Fourraan, P. . 4, 36, 44, 57, 59, 60, 95, 97, 115, 138, 151, 163, 164, 196, 795, 197, 206, 221, 224, 225, 262, 263, 507, 308, 309, 309, 575 Fowell, D. M Foy, H. . Fox, H. . Frahck, R. L. . Franca, P. da Franke, F. Frankland, H. M FrankUn, K. J Eraser, D. . Eraser, E. A. Eraser, F, C. Fraser, R. Franz, V. . Frederic Freeman, H. 236 Freudenberg, F. . Freudenstein, H. . Freudenstein, K. Freydberg, V. Friedemann, T. E Friedman, M. Friedman, R. Friedman, S. M. 4, 275, 277, 295 3, 189 1, 88; 5, 705, 709, 772 2, 5, 70, 75 . 5, 755, 775 . 5, 62, 69 T. . .5, 53, 54 1, 31, 65, 76, 77, 79, 202, 205, 214, 242, 245; 3, 51, 52, 59, 71, 72, 126 2, 174 . 5, 759, 207 2, 19, 26, 168, 172 . 4, 507, 308 . 5, 202, 207 2, 103 1, 36, 50, 51, 52, 137, 138, 139, 219, 279, 223, 224, 225, 226, 227, 230, 232, 233, 234, 236, 237, 238; 4, 90, 93 . 5, 62, 69 5, 235, 236, 237, 241 5, 232, 241 3,57 Friis-Hansen, B. J. Friley, M. Froesch, E. R Frontali, G. Frost, W. E. Funaro, R. Fry, D. H. Fry, F. E. J. 2, 750, 143 3, 755, 142 3, 75, 90 3, 45, 128, 146, 184, 192 4, 102, 705, 705, 709, 775, 114, 143, 149 2, 757, 772; 3, 77, 75 . 4, 280, 298 . 1, 799, 200 . 5, 755, 775 . 1, 799, 200 5, 755, 770, 775 . 5, 792, 207 Gaarenstroom, J Gabrio, B. W. Gakkel, L. B. Gale, E. F. Galvan, R. R. Gamble, J. L. Gans, B. . Gardiner, E Gardner, E. Gardner, G. Gardner, W. U Garrod, O. Gatenby, J. B. Gauer, O. H. 4, M. H. . .2, 64, 66 2, 228, 229, 231, 232, 237, 238 . 3, 755, 142 1, 757, 792, 795 . 1, 795, 200 138, 144, 149, 216, 219 4, 79, 93, 97, 98 . 5, 27, 57 . 5, 248, 264 . 3, 70, 75 1, 755, 154, 160, 161 4, 292, 293, 296 . 2, 55, 96 4, 255, 284, 296, 297 8 Cumulative Author Index Gaunt, R. Geiser, S. W. Gele, P. . Gellerstedt, N. Gellhorn, A. Genest, J. . Geoghegan, H G6rard, P. Gerking, S. D. Gemer, K. Gerritsen, T. Gianferrari, L. Giardini, A. Gibson, J. G. Gibson, J. R. Gilbert, C. Gilbert, C. S. Gilbert, J. C. Gilligan, D. R Gilman, A. Gillman, J Gillman, T. Ginsburg, J. Ginsburg, M. Girard, A. Giroud, A. Giroud, C. J. Glass, B. . Godden, W. Godet, R. Godwin, H. Gomori, P. Gofman Goldblatt, H. Golden, J. B. Goldfarb, N. Goldhamer, H Goldner, F. Goldring, W. Goldstein, K. G6mez, F. Gompertz, B. Gonse, P. . Gontarski, H. Goodbody, M, Goodman, L. Gordon Gordon, E. B. Gordon, E. E. Gordon, E. S. Gordon, G. L. Gordon, J. 4, 193, 194, 195, 296 . 5, 167, 175, 177 . 4, 273, 278, 296 . 1, 35, 48; 3, 147 2, 111, 113, 118, 125, 148, 158 . 4, 277, 296 4, 24, 31, 35 2, 39, 40, 42, 53, 56 5, 18, 79, 171, 175, 179, 180, 181, 183, 207, 209,210,211,228,229, 230, 265, 284, 289 . 5, 63, 69 . 3, 110, 114 . 3, 137, 142 . 4, 75 . 4, 243, 246 . 5, 29, 31 3, 126 5, 245 . 3, 160, 168 . 1,90, 103 . 4, 37, 57 3, 20, 26, 104, 105, 106, 107, 108, 109, 114, 126, 189 . 3, 33, 36, 48, 49, 70, 70, 72, 96, 100, 104, 104, 105, 106, 107, 108, 109, 114, 126, 145, 146, 188, 189, 190 . 2, 148, 158 . 4, 46, 57 . 1, 129, 137 . 2, 765, 172 . 4, 293, 294 . 3, 136, 142 4, 309 . 2, 16, 17 . 2, 206, 209 4, 284, 286, 296 1,89 4, 253, 258, 259 1, 143, 153, 160 . 3, 173, 182 . 1, 43, 48 4, 38, 40. 43, 44, 57 . 4, 230, 246 . 3, 180, 182 . 1, 198, 200 5, 2, 15, 72, 79, 302, 311 . 2, 21, 26 . 5, 232, 241 . 4, 245, 246 . 1, 151, 160 1, 138 . 4, 49, 56 2, 142, 143, 144 . 4, 292, 296 4, 38, 40, 43, 44, 57 4, 226, 257, 259 Gordon, J. E. Gottlieb, E. Gotzche, H. Gould, R. G. Gowenlock, A. H. Graber Graham, I. Grainger, E. H. . Gram, M. R. Granick, S. Gray, M. J. Greaves, M. S. . Green, C. V. Green, H. H. Green, R. G. Green, S. H. Greenberg, D. M. Greene, R. Greene, R. R. Greenfield, A. D. M. Greenwood, M. Greep, R. O. Gregory, F. G. Gresson, R. A. R. Gribetz, D, Griew, S. Griffin, G. E. Griffith, L. Griffiths, J. T. . Groot, A. P. de Grosch, D. S. Gross, F. Gross, J. . Grossman, M. I. Grossman, R. Grosz, S. . Gruber, G. B. Griineberg, H. Grueninger, R. M Gruenwald, P. Guild, W. R. Guilford, J. P. Gumbel, E. J. Gunter, G. Guzmdn, M. A. Haddow, A. Hagel, L. Hagerman, D. D Hahn, P. . Hahn, P. F. Hafter, E. Haigh, L. D. Hain, A. M. Halberstaedter, L Hald, A. . Haldane, J. B. S Haldane, J. S. Haley, H. B. . 5, 21, 31 . 5, 149, 175 . A, 41, 56 . 2, 217, 232 4, 60, 193, 195 4, 151 . 3, 98, 100 5, 147, 153, 175 . 1, 122, 123 2, 238 . 4, 89, 93 . 4, 265, 268 3, 9, 18 . 4, 200, 205 5, 297 2, 32, 33, 53 4, 301, 307, 308, 309 . 4, 84, 93 2, 3, 12, 17 2, 120, 121, 125 5, 311 . 2, 63, 66 2, 204, 207, 209 . 2, 86, 96 . 4, 143, 149 . 3, 166, 169 . 4, 109, 113 . 4, 109, 113 . 5, 277, 281 5, 235, 236, 237, 238, 241 5, 255, 264 1,242 . 3,48 1,191, 193 4, 280, 295 2, 179, 183 2, 181, 183 238; 3, 118, 124; 5, 19, 20, 79, 80, 283 . 1, 198, 201 . 2, 16, 17 4, 45, 56, 57 . 1, 40, 48 5,117, 132, 134 . 5, 167, 175 . 1, 198, 200 . 3, 68 . 5, 59, 69 . 2, 142, 144 . 4, 165, 179 . 2, 236, 238 5, 196, 208, 218, 225, 226 . 4, 131, 135 . 2, 162, 172 . 2, 46, 53 . 5, 120, 133 1,10, 15, 238; 3, 5, 12, 18; 5,53, 54 . 4, 63, 73 . 4, 109, 113 h Cumulative Author Index 9 Hall, D. A. . . . 1, 104: 3, 65, 66, 67, 96, 97, 98, 100, 102, 103; 5, 113 . 2, 120, 125 . 4, 256, 259 4, 274, 277, 297 4, 201, 202, 203, 205 4, 279, 296 4,262 Hall, F. G Hall, K. . Hall, P. E. Hallman, N. Halperin, M. H. Hamburger, J. Hamilton, B, Hamilton, H. B. Hamilton, J. A. Hamilton, J. B. Hamilton, T. S. Hamilton, W. F, Hamilton, W. J. Hamlett, G. W. D. Hammond, J. Hancock, W. Hanley, T. . Hanon, F. . Hansen, E. Hansen, J. D. L. Hansen, W. H. Hanyu, I. . Harmison, C. R. Harris, B. A. Harris, E. J. Harris, G. W. Harris, H. . Harrison, J. H. Harrison, J. L. Harrison, M. F. Harrison, R, J. Harrison, T. R. Hartman, C. G Hartmann, W. Harvey, G. F. Hartwig, W. Hart, J. L. Hart, J. S. . Hass, G. M. Hastings Hastings, A. B Hatai, S. . Hatcher, J. D. , Hatey, J. . Hathom, M. Hatton, H. Haugen, G. E. . Havighurst, R. J. Hawkins, D. F. , Haydak, M. H. , Hayman, J. M. , 4, 10, 11 1, 227, 236 4, 242, 246 1, 227, 236 4, 120, 121, 131, 135 4, 275, 277, 295 2, 114, 115, 116, 117, 755, 156, 158, 159, 160 2, 38, 40, 53, 56 . 1, 189, 193 . 4, 63, 73 4, 269 . 4, 217, 219 . 5, 59, 69 . 4, 133, 135 . 3, 123, 124 . 5, 757, 775 . 5, 73, 79 2, 729, 132, 143 . 4, 20, 31 . 1, 752, 760 4, 202, 203, 205 . 4, 45, 57 5, 95, 96, 102 . 1, 7P7, 193 2, 37, 54, 68, 81, 83, 83, 84, 85, 115, 148, 148, 149, 150, 155, 156, 158, 159, 160, 199 4, 283, 287, 288, 296, 298 . 1, 146, 160 . 5, 59, 69 3, 9, 18 . 5, 56, 57, 65, 66, 67, 69, 70, 71 . 5, 144, 151, 155, 175, 176 . 5, 770, 775 . 1, 88, 103 1,207 . . 2, 141, 143, 144; 3, 792; 4, 779, 720, 7i7, 135 5, 248, 264 . 4, 279, 296 . 2, 24, 27 3, 70^, 705, 77^ . 1, 70, 75 . 2, 7iO, 143 . 1, 43, 48 . 4, 90, 93 5, 232, 236, 237, 238, 239, 241, 242, 253, 264 . A, 271, 288, 297 Heady, J. A. Hebb, D. O. Hecht, H. H. Hechter, O. Hediger, H. Hegsted, D. M. Heidenhain, R. Heilbrunn, L. V. Heim, A. W. Heinbecker, P. Hejtmanek, J. Helikson, J. Heller, H. . 5, 27, 31 . 3, 77^, 182 . 4, 266, 268 1,139; 4, 262 . 2, 777, 183 1, 709, 727, 722 . 4, 63, 73 . 1,55 . 3, 163, 168 4, 256, 257, 259 5, 235, 237, 242 . 1, 198, 201 4, 6, 7, 11, 12, 72, 13, 13, 97, 98, 114, 137, 163, 755, 757, 755, 759, 178, 179, 195, 196, 247, 575 Hellman, L. M. 2, 777, 772, 775, 729, 752, 144; 4, 82, 85, 93, 251, 259 Hems, J. Henly, A. A. Henry, J. P. Henschel, A. Herald, E. S. Herbeuval, R. Herbst, E. J. Herkel, W. Herlant, M. Herlitzka, A. Herrington, W. C. Hers, H. G. Herscheimer, A Hertig, A. T. Hertz, R. Hervey, G. F, Hervey, G. R. Hess, A. Hess, G. Hess, J. H. Hesselberg, C. Heston, W. E. Heusler, K. Hevesy, G. . Hickey, J. J. Hickling, C. F. . 3, 9, 18; 5, 227 . 1, 750, 757 4, 38, 57, 283, 284, 296, 297 4, 757, 757, 752 . 5, 755, 775 . 4, 242, 246 . 1, 797, 200 . 2, 757, 158 2, 39, 40, 42, 53 . 2, 33, 53 . 5, 757, 777 . 2, 755, 7^5 4, 750 1,745, 161; 2, 110, 113 1, 143, 151, 160 5,227 4, 40, 57 Higgins, G. Higginson, J. Hilden, T. . Hile, R. . Hill, Hill, A. V. Hill, S. Hillarp, N. A Hiller, A. . Hillman, D. Hilton, J. G. Himbert, J. Himwich . Hines, B. E. Hingerty, D. . 3,745 5, 235, 242 4, 750, 755 2, 80, 83, 84 3, 775, 727, 124 . 4, 181, 195 . 2, 204, 209 5, 97, 101, 102 5, 795, 795, 205, 207 4, 36, 41, 43, 57 . 3. 770, 114 . 4, 41, 56 . 5, 757, 775 2, 727 1, 77, 76, 77, 78 . 2, 757, 158 . 2, 60, 66 . 4, 272, 298 4, 740, 747, 745, 749 4, 245, 246 4, 273, 278, 279, 296 . 1,57 . 4, 274, 279 4, 35, 99, 99, 101, 113, 207, 207, 226, 226, 247, 310, 570 10 Cumulative Author Index Hinshelwood, C. Hinton, H. E. Hinton, S. . Hippocrates Hisaw, F. L. Hitchcock, M. W Hoagland, H Hobbs, G. E. Hobson, W. Hodge, C. F. Hoelzel, F. Hoffman, F. G Hoffman, H. Hoffman, J. Hoffman, K. Hogreve, F. Holland, B. C. Hollander Hollander, W. Holliday, M. A Holmes, E. G. Holmes, S. J. Holmgren, A. Holt, S. J. . Honmia, H. Honsova, H. Hoogendoom, D. Hooker, C. W. . Hope, J. M. 3, 5, 18 5, 34, 112, 243, 245, 246, 284 5, 191, 217, 219, 225 2, 129 1, 143, 151, 160', 2, 81, 83 S. . .2, 122, 125 . Iy219,223,224,225, 226, 227, 230, 232, 233, 234, 236 1, 223, 232, 236 . 3, 94, 100 . 5, 248, 264 1, 197, 199, 206 . 2, 39, 54 . 5, 237, 243 . 1, 153, 160 . 1, 153, 161 5, 64, 65, 69 . 4, 51, 57 2, 251 . 4, 279, 296 . 4, 129, 135 . 4, 130, 135 . 5, 21, 31 4, 279, 280, 295 5, 86, 134, 142, 149, 151, 155, 158, 167, 175, 176, 178, 179, 209, 210, 226, 227, 228, 229, 246, 290 . 2, 188, 191, 192, 199 . 4, 291, 296 . 5, 21, 31 . 2, 91, 96 1, 219, 223, 224, 225, 226, 227, 232, 233, 236 Hopwood, H. H. . . 1, 198, 201 Horn, G. Howell, T. H. Hoy, P. A. Huant, E. . Hubble, D. Hubbs, C. L. Hudson, P. Huggett, A. St. G. Hughes, A. Hughes, J. S. Hughes Jones, N. Hugin, F. . Hull, T. Z. Hultquist, G. T. Hunmiel, K. P. . Humphries, E. M Hungerland, H. Hunt, J. N. 2, 189, 199 . 1,18,23; 3, 93, 100 4, 3, 4, 5, 6, 9, 11 . 2, 167, 172 . 4, 78, 93 5, 146, 155, 176 . 1, 136, 137 2, 27, 29, 56, 58, 84, 116, 117, 118, 119, 120, 121, 123, 125, 126, 127, 144, 145, 146, 159, 160, 174, 175, 186, 200, 201, 210, 214, 240, 242, 243, 246, 247, 251, 252 2, 251 . 4, 200, 205 2, 234, 235, 238; 4, 265, 268 3, 63, 64, 67 . 4, 131, 135 . 2, 162, 172 . 1, 156, 160 2, 34, 51, 53 . 4, 212, 219 . 4, 210, 219 Hunter, W. R. Hurley, T. H. Hurst, H. . Hurst, J. G. Huseby, R. A. Huxley, J. Hyett, A. R. Hynes, N. H. B. Hytten, F. E. Ickowicz, M. Ikkos, D. . Ingle, D. J. Ingle, L. Ingram, D. L. Innes, L. R. Inukai, T. lob, L. V. Irving, J. T. Irwin, J. O. Iseri, L, T. Iske, B. Isupov, A. P. Ito, T. Itoh, S. Ivlev, V. S. Ivy, A. C. Iwashige, K. Izzo, M. J. Jackson, B. H. Jackson, H. Jackson, L. Jackson, W. P. U Jacob, H. . Jacob, M. Jacobsohn, D. Jacquot, R. Jaczewski, Z. Jaeger, L. M. Jalavisto, Eeva James, A. H. James, D, W. James, W. . James, W. O. Jelinek, J. . Jenkins, R. Jennings Jenson, R. L. Jepson, R. P. Jeter, M. A. Jewell, P. A. Johansson, A. S. John, M. . Johnsen, S. G. Johnson, B. B. Johnson, L. C. Johnston, M. E. Jolliffe, N. Jolly, W. A. Jones, A. R. 5, 179 . 2, 235, 238 . 3, 10, 18 . 1, 155, 161 . 1, 755, 160 5, 104 . 2, 755, 759 5, 755, 756, 775 . 5, 33 . 2, 46, 53 . 4, 709, 775 4, 257, 259, 260 1, 797, 200; 3, 24, 29 2, 35, 44, 53 . 2, 755, 772 . 5, 284, 264 . 4, 750, 755 . 1, 770, 725 . 5, 37, 54 . 4, 289, 296 . 1, 775, 185 . 5, 53, 54 . 2, 189, 199 . 4, 24, 32 5, 186, 207, 208 . 2, 12 2, 759, 797, 799 . 2, 236, 239 . 5, 89 . 1, 34 . 1, 722, 725 . 2, 757, 772 . 3, 75^, 142 . 4, 252, 259 . 1, 752, 750 . 2, 24, 27 2, 181, 182, 183 . 3, 28, 29 3, 757, 142; 5, 16, 17,21,27,57,31,32, 33, 81, 113,266 4, 709, 775 . 3, 70 . 1,43 2, 205, 209 4, 165 5, 168, 176 2,252 4,222 4, 750 2, 148, 159 4, 37, 57 5,243 5, 237, 242 1, 69, 75, 76 4, 187, 195, 280, 293, 294, 297 1, 145, 161 4, 289, 298 1, 799, 200 2, 33, 54 2,225 Cumulative Author Index 11 Jones, E. C. Jones, E. R. Jones, H. E. Jones, H, S. Jones, J. W. Jones, R. V. H. Jordan, D. S. Josephson, B. Josimovich, J. I Jost, A. Judson, W. June, F. C. . 5, 277, 281 . 4, 130, 135 3, 160, 168, 171, 182, 186 . 4, 64, 73 5, 153, 156, 176 . 4, 54, 57 . 5, 212, 226 4, 279, 280, 295 . 2, 23, 26 2, 3, 11, 12, 13, 13, 14, 15, 16, 17, 18, 18, 19, 20, 21, 23, 24, 26, 27, 27, 28, 29, 30, 84, 116, 126, 127, 146, 167, 172, 173, 174, 175, 252 . 4, 279, 296 . 5, 193, 208 Kadid, M. . Kagan, B. M. Kahlenberg, O. J. Kalabouchov, N. I. Kallas, H. . Kallmann, F. J. . 5, 43, 54 . 4, 130, 135 . 4, 127, 135 . 5, 96, 102 . 1, 144, 160 1,42, 48; 3, 37,101, 131, 132, 133, 137, 139, 142, 143, 143, 144, 145, 146, 147, 148, 173, 182 Kalter, H 2, 168, 172 Kaplan, S. A. . . .4, 287, 296 Karnovsky, M. L. . . 2, 141, 143 Karvonen, M. J. 4, 74, 76, 199, 201, 202, 203, 205, 206, 207, 208, 221, 316 Karzinkin, G. S. Kassila, E. Katersky, E. M. Kattus, A. Katz, J. Katz, M. . Kaufman, N. Kauzman, E. F. . Kay, H. . 1, 213; 3, Keech, M. K. . Kelley, R. B. Kellner, G. Kellogg, V. L. . Kennedy, G. C. 5, 186, 208 . 4, 74 . 1,90, 103 . 4, 277, 296 4, 257, 259, 261 5, 196, 197, 208 3, 189 . 2, 228, 231 162, 169, 182, 183 3, 97, 100, 103 . 3, 10, 18 2, 36, 42, 43, 53 1, 196, 200 1, 189, 193; 4, 136, 136, 137, 150, 164, 197, 226, 227, 250, 251, 253, 254, 257, 259, 260, 261, 262, 263, 313, 314, 317 Kennedy, W. A. . 5, 145, 151, 176 Kerpel-Fronius, E. . 4, 39, 57, 96, 105, 113, 154, 155, 157, 161, 162 Kerr, S. E. . . .4, 202, 205 Kerrigan, G. A. . 4, 139, 140, 141, 146, 147, 149 Kershaw, W. E. . 5, 56, 56, 70, 105, 114, 245, 267,285 Keston, A. S. . . .2, 206, 209 Keszler, H. . . .4, 291, 296 Kety, S. S. . 3, 147; 4, 286, 296 Keys, A. . 1, 89, 207; 3, 75, 91; 4, 157, 161, 162 . 2, 206, 209 4, 63, 65, 68, 73 . 4, 266, 268 4, 253, 257, 259 . 5, 10 . 3, 20, 29 . 5, 56 4, 202, 203, 205 3, 189 . I, 227, 236 , 69, 75, 76, 76, 77, 78, 79 4, 63, 73 2, 197, 198, 199, 200 3, 139 . 3, 21, 29 3, 189 . 4, 46, 57 2, 127 . 1, 154, 160 . 4, 133, 135 , 33, 287, 288, 295 3, 134, 136, 143 4, 133, 135 5, 99, 102 Kidd, F. . Killmann, S-A. Kilpatrick, J. A, Kimball, G. C. King, G. . King, H. D. King, J. O. L. King, J. W. B. Kinney, T. D. Kirk, E. . Kirk, J. E. 1, 69 Kittsteiner, C. Klaar, J. Klaber Klauber, L. M. Klavins, J. Kleeman, C, R. Klein Klein, M. . Kleinman, A. Kleinzeller, A. Klopfer, H. W. Klug, H. L. Kluijver, H. N. Knobil, E. Knowles, H. C. Knox, G. . Koch, A. . Kocher, v.. Koczorek, K. R Koehler, A. Kohler, M. Konig, K. . Koltay, E. . Kondi, A. . Konopinski Koonig, T. Kopec, S. . Korenchevsky, V Kosterlitz, N. Kovach, A. Kovach, A. G. B Kovach, I. Kowalewski, K. Kracke, R. R. Kratky, E. Krause, C. Kfecek, J. Kfeckova, J. Krogh, A. . Krohn, P. L. Kruesi, O. R. Krumholz, L. A. 2, 166, 172 4, 36, 57 5, 21, 31 5, 236, 238, 242 . 5, 237, 242 4, 280, 292, 293, 298 . 5, 232, 242 . 4, 292, 298 5, 61, 62, 69 4, 284, 286, 296 3, 189 . 5, 67, 69 . 4, 301, 308 1, 197, 200; 3, 25, 29 . 4, 256, 259 . 2, 166, 172 4, 284, 286, 296 4, 45, 57, 284, 286, 296 . 4, 157, 161 . 1, 227, 236 .2, 112, 113 5, 232, 236, 242 . 5, 110, 112 4,9, 10, 11, 113, 165, 179, 195, 196, 316 . 4, 9 10, 11, 165, 179 1, 77; 4, 75 . 1, 30, 139, 141, 156, 160, 162, 111, 203, 242, 245;2, 57, 55, 67, 84, 211, 239, 240, 243, 249, 250, 252; 3, 123, 124, 125, 126, 127; 5,271,281 4, 245, 246 5, 171, 176, 187, 188, 208 12 Cumulative Author Index Kruse, H. D. Kuechle, M. E. Kuiken, A. Kuhn, W. . Kun, K. . Kuno, Y. . Kurzrok, R. Kuznetsova, G. Lacassagne, A. Lack, D. . Ladman, A. J. Lambdin, E. Lament, N. Lanciano, G. Landau, B. R. Landowne, M. 4, 301, 308 1, 198, 201 5, 237, 243 5, 240, 242 4, 161, 162 4, 64, 73 1, 145, 160 5, 53, 54 2, 34, 46, 53 1, 4, 15; 5, 97, 101, 102, 104 2, 23, 26 4, 46, 57 3, 105, 188 1, 199, 200 2, 141, 143 1, 66; 3, 33, 35, 36, 46, 47, 69, 69, 73, 75, 80, 84,88,89,90,91,94, 129; 4, 235, 236, 246 Lane, C. E. . 1, 152, 160; 2, 63, 66 Lane-Claypon, J. E. . 2, 36, 37, 39, 53 Lang, C. A. Langley, J. N. Langmaid, C. Langstroth, G. O. Lansing, A. L 3, 79, 91 4, 63, 73 . 4, 44 4, 64, 65, 73 1, 25, 26, 27, 30, 50, Laragh, J. H. Lardy, H. A. Larkin, P. A. Larsen, C. D. Larsen, E. E. Lasch, W. . Laser, H. . Laszio, D. . Laursen, T. J. S. Lavoipierre, M. M. J Law, L. W. Lawson, H. D. Lea, E. Leaf, A. Leavenworth, C. Leblond, C. P. Leduc, E. H. Lee, S. van der Leeson, P. M. Lehman, H. Leitch, L . Leiter, L. . Lell, W. A. Le Magnen, J. Le Marquand, H Len^gre, J. Lenhossek, M. Leonard, J. A. Leppanen, V. Lesher, S. . 51, 55, 78, 88, 91, 99, 103, 104, 105, 106, 107, 108, 148, 160, 203, 206, 241, 242, 243, 244; 2, 109, 113, 247; 3, 97, 98, 100, 102; 5, 259, 264 4, 53, 57, 292, 293, 296 . 2, 208, 209 . 5, 182, 208 . 3, 121, 124 . 4, 95 . 4, 97 . 4, 218, 219 1, 110, 122, 123 . 1,69 . 5, 245, 285 . 1, 147, 160 . 1, 143, 160 5, 145, 149, 176 4, 42, 46, 57, 145 . 2, 203, 210 . 2, 86, 96 2, 251 . 3, 117, 125 . 4, 36 1, 39, 40, 48, 51 . 1, 196, 199 277, 280, 295, 297 . 2, 109, 113 3, 129 . 1, 197, 200 4, 273, 278, 296 . 2, 86, 96 . 1, 211, 213 4, 200, 202, 203, 205 . 5, 80 4, Leslie, P. H. . 3, 10, 18; 5, 296 Leutscher, J. A. . . . 4, 294 Lever, J. D. . . . 2, 102 Leverton, R. M. . . 1, 122, 123 Levin, M. D. . . 5, 232, 236, 242 Levine, A. D. . . . 4, 228 Levine, J 4, 228 Levinson, Z. H. . . 5, 255, 259, 264 Lewin, W. . .4, 36, 41, 43, 57 Lewis .... 2, 103 Lewis, A. J. . 1, 32, 49, 50, 51, 52, 138, 214, 218, 238, 246; 3, 143, 144, 146, 147, 185 Lewis, C. S. . . .4, 266, 268 Lewis, G. T. . . .4, 127, 134 Lewis, H. B. . . .1, 94, 103 Lewis, K 1, 33, 48 Lewis, W. H. . 4, 242, 244, 246, 248 Li 2,174 Li, C. H 1, 29 Li, M. H 1, 154, 160 Liber, K. E. . .2, 109, 113 Liddle, G. W. . 4, 89, 92, 280, 292, 293, 294, 297 Lieb, C. W. . . . 4, 262 Lieberman, H. M. . 5, 250, 251, 255, 256, 257, 258, 260, 264 Liebman, A. . . .4, 42, 56 Liener, I. E. Light, A. E. Lightbody, H. D Lilly, D. M. Lindan, O. Lindauer, M. Linderstr0m-Lang, K Lindgren, A. Lindop, Patricia J. Lippi, B. . Lipschutz, A. Little, C. G. Liu, C. H. . Ljunggren, H. Llaurado, J. G Lloyd, C. W. Lobitz, W. C. Locke, W. . Lochhead, J. Loeb, J. Loeb, L. Loeschcke, H. Loeser, H. Logan, M. A. Logethetopoulos, J. Lombardo, A. J. Lombardo, T. A. Long, C. N. H. . Longley, L. P. Lorenz, E. 1, 192, 193 4, 120, ni! 128^ 131, 135 . 4, 133, 135 . 1, 187, 193 . 2, 112, 113 . 5, 233, 242 . 2, 204, 209 . 3, 134, 143 3, 52, 59; 5, 113, 136, 138, 141 . 4, 242, 246 1, 151, 160; 2, 34, 53 . 1, 156, 160 . 1, 110, 122 . 4, 109, 113 . 4, 293, 297 . 4, 193, 195 2, 191, 192, 194, 199 . 4, 64, 73 2, 23, 27, 122, 125 3, 24, 29 1, 146, 157, 158, 160; 1,41, 53, 54, 55, 69, 70, 73, 75, 80, 83, 84, 161, 172 2, 196, 197, 199 2, 129, 143 1, 91, 103 4, 265, 268 1, 198, 200 4, 283, 298 4, 82, 93 4, 271, 288, 297 . 5, 89 Cumulative Author Index 13 Lorge, I. Lotka, A. . Lotmar, M. Loutit, J. F. Louveaux, J Lovelock, J. E. Lowry, O. H. Luall, A. . Luck, J. M. Ludemann, H. Luecke, R. W, Luetscher, J, A Luft, R. . Lundberg, A. Lundeen, E. Lutwak-Mann, C Lutz, R. . Lwoff, A. . 3, 32, 33, 36, 37, 72, 147, 148, 170, 170, 182, 183, 184, 185, 186, 189 . 3, 137, 142 5, 232, 233, 234, 242 2, 221, 232, 236, 238 . 5, 236, 242 . 1, 163, 164, 165, 168, 169; 2, 215, 217, 218, 221, 224, 232, 240, 241, 242, 243 1,90, 103; 3, 192; 4, 120, 131, 135 . 1, 110, 123 . 4, 132, 135 . 4, 57, 56 . 5, 237, 243 4, 280, 292, 293, 297 . 4, 109, 113 . 4, 69, 70, 73, 76 . 4, 130, 135 . 4, 218, 219 . 1, 122, 123 . 3, 46 MacArthur, J. W. 3, 24, 29; 4, 141, 149 McCall, J. T. . .2, 148, 159 McCance, R. A. . 1, 67, 76, 77, 103, 186, 186, 193, 197, 200, 203, 205, 207; 3,792; 4, 1, 6, 11, 13, 38, 40, 41, 44, 56, 57, 59, 76, 95, 96, 96, 97, 100, 115, 120, 121, 128, 136, 137, 138, 150, 157, 160, 162, 162, 180, 195, 197, 799, 203, 205, 207, 208, 209, 277, 272, 27^, 275, 277, 218, 219, 222, 222, 223, 257, 259, 262, 263, 269, 298, 305, 306,308, 309, 310, 315 MacCarty, C. S. 4, 41, 56 Macaskill, E. H. . . 4, 200, 205 Macaulay, D. . . .4, 46, 57 McCay, C. M. . 1, 27, 29, 30, 722, 72J, 124, 125, 173, 174, 181, 185, 188, 189, 193, 194, 197, 200, 204, 206, 245 ; 2, 248; 3, 8, 18, 25, 29; 5, 759, 775, 257, 252, 253, 264, 265, 268 McCoIlum, E. V. McCormack, J. L McCracken, F. D. McCrory, W. W. MacDonald, A. M. McDonald, R. K. 4,22^ 4, 24, 31, 35 . 5, 757, 775 . 4, 46, 57 4, 301, 307, 308, 309 1, 220, 236; 4, 2J7, 233, 235, 237, 246, 247 McFadyean, J. . . -5, 50, 54 McFarland, R. A. . .3, 133, 143 McGaughey, R. S. . .4, 289, 296 McGee, L. E. McGregor, E. A. McGregor, J. H. McHugh, T. McIIroy, M. B. Mcintosh, B. J. Maclntyre, L McKeown, T. Mackintosh, J. Maclaren, A. MacMahon, L. McMaster, P. D. McMurrey, J. McMurrey, J. D. MacPhee, G. McRae, D. R. Maar, A. . Mach, R. S. Mackler, B. Macy, L G. Madden, S. C. Maddock, W. G. Magendie . Maggioni, G. Magnus, I. C. Mahler, R. F. Mahoney, J. J. Maizels, M. Makeham, W. M Malamud, N. Malerba, G. Malm, O. J. Mamby, A. R. Manca, P. V. Manciaux, M. Mancini, R. E. Mandel, L. Mandel, P. Mandelstam, J. Mandl, A. M. Manery, J. F. Mangel, M. Mannes, A. Mantz, F. R. Marcus, S. Marden, W. G. R Marescaux, J. Margetts, A. R. Markee, J. E. Marks, H. . Marks, L N. Mario w, H. Marmer, L L. Marmorston, J. Marr, J. C. Marsden, J. L. 3, 52, 59 5, 192, 208 2, 87, 96 5, 705 1, 57, 65 4, 24, 31 A, 115 5, 29, 31 5, 27, 31 3, 7, 18 5, 27, 31 4, 29, 32 4, 102 4, 702, 705, 108, 113 . 4, 90, 93 4, 64, 65, 73 . 5, 797, 208 4, 292, 293, 295, 297 . 4, 40, 57 . 4, 143, 149 4, 7i2, 133, 135 . 4, 39, 57 1, 174 . 1, 799, 200 4, 702, 705, 108, 113 . 4, 193, 195 2, 14, 15, 17 . 4, 799, 205 5, 2, 4, 15, 302, 311 . 3, 134, 143 . 4, 242, 246 1, 109, 777, 775, 720, 722, 725 4, 42, 46, 57 2, 797, 792, 799 . 4, 242, 246 . 2, 86, 96 . 4, 252, 259 . 4, 252, 259 . 3, 108, 114 1, 144, 160; 2, 32, 33, 34, 35, 49, 50, 51, 53, 54, 60, 66; 3, 727; 4, 257, 259 . 4, 705, 772 . 1, 722, 725 5, 59, 60, 69 . 4, 709, 775 . 4, 187, 195 . 1, 7^5, 750 2, 36, 37, 38, 39, 41, 42, 46, 47, 53, 62, 66 . 5, 757, 775 . 1, 752, 760 . 3, 85, 90 4, 301 . 5, 63, 69 . 4, 725, 755 4, 257, 259, 261 . 5, 755, 775 . 3, 56 14 Cumulative Author Index Marshall, F. H. A. 2, 33, 54, 148, 159 1, 67, 65 3, 171, 182 4, 301, 308 5, 59, 64, 69 4, 165, 179 . 4, 75 4, 280, 294 3, 75, 85, 90 2, 113 5, 43, 54 4, 289, 292, 295 2, 37, 54, 56, 81, 83, 85, 186, 187 . 5, 232, 234, 242 5, 231, 232,234,235, 236, 238, 242, 243, 244, 245, 246, 248, 254, 264 May, R. M. . . .1, 151, 160 Mayer .... 2, 775 Maynard, L. A. . 1, 188, 193; 3,8, 18, 25, 29; 5, 183, 208, 251, 264 Maynard-Smith, J. . 1, 239, 240, 241 Marshall, R. Marston, M. Martin, H. E. Martin, W. Martinek, J. Massart Massie, E. . Master, A. M. Masters, W. H. Matsumoto, K. Matthews, H. L. Matthews, L. H. Mauermayer, G. Maurizio, Anna Maxwell, E. L. Maxwell, M. H. Meara, P. . Medawar, P. B. Meduna Meggendorfer, F Megyesi, K. Mehl, J. . Meier, R. . Mellen, I. . Mendel, G. Menon, M. D. Menzel, D. W. 5 Mercier-Parot, L Merendino, K. A Merkel, F. Merrill, A. J. Merrill, J. P. Metcalf, E. V. Metchnikoff, E. Meves, F. . Meyer, H. . Michael, G. Michie, D. Mickelsen, O. Middlesworth, L Miescher, K. Miley, J. F. Milinsky, G. I. Millard, A. Miller, F. J. W. Miller, G. E. Miller, J. F. 2, 3, 10, 11, 13 . 4, 274, 297 . 4, 145, 149 1, 4, 5, 12, 15, 24, 54, 78, 105, 139, 169, 170, 171, 201, 245; 2, 199, 229, 23 1;3, 5, 6, 18; 5, 271, 281 1, 237 . 3, 137, 143 . 4, 286, 296 . 4, 301, 308 4, 181, 187, 194, 195 5, 277, 226 4,208 5, 149, 176 183, 184, 185, 208, 210 2, 161, 163, 168, 170, 172 . 4, 46, 56 . 4, 62, 73 4, 271, 277, 279, 297 4, 45, 56, 57 1, 223, 232, 236 1,175, 181, 185 . 2, 86, 96 . 4, 130, 135 . 2, 202, 209 3, 7, 18 4, 757, 757, 762 van . . 3, 55, 59 1, 56, 57, 138, 202; 3, 194 4, 24, 31 5, 201, 208 2, 208, 209 3, 95, 100 4, 288, 297 4, 301, 308 Miller, J. H. Miller, L. L. Miller, R. B. Milne, M. D. 257, 206, 227, Miner, C. S. Miner, J. R. Miner, R. W. Minot, C. S. Mitchell, H. H. Mitchison, M. J. Moffitt, G. R. Moise, T. S. Mokotoff, R. MolUson, P. L. 276, 233, 238, Moltoni, E. Monakow, von . Montagna, W. 103, 792, 201, Moon, H. D, Moore, C. R. Moore, F. D. Moore, H. L. Moore, R. A. Moore, W. W. Morant, G. M. Moreschi, C. Morgulis, S. Morris, C. J. O. R Morris, J. N. Morrison, A. Morrison, J. L. Morrison, S. Moscoso, I. Moskovljevic, V. Moskowitz, M. Mossman, H. W, Mothes, K. Moulton, C. R. Mounib, M. S. Muehrcke, R. C. Miihlbock, O. . 118, 125, Mussbichler, A. Muggleton, A. Muir, H. M. Muller, A. F. Munro, H. N. Munson, P. L. 1, 220, 236; 4, 12, 233,238,240,241,246 . 4, 133, 135 3, 26, 29; 5, 168, 176 4, 93, 96, 138, 197, 221, 222, 224, 225, 248, 262, 268, 280, 297, 299, 300 . 2, 228, 232 5, 267 . 5, 53, 54 1,6; 4, 250, 259 4, 120, 121, 131, 135 . 2, 223, 232 . 4, 289, 297 . 4, 254, 259 . 4, 227, 280, 295, 297 . 1, 164, 169; 2, 215, 221, 226, 231, 232, 234, 235, 236, 237, 239, 239, 240, 241, 242, 243, 244, 245 . 3, 9, 18; 5, 103 . 1, 50 2, 86, 91, 96, 96, 99, 160, 175, 188, 797, 79^, 799, 199, 200, 211, 240, 247, 247, 248, 249, 251 1, 29 2, 3, 13, 18, '27! 34, 54 4, 102, 108, 110, 113, 114, 291, 297 5, 755, 776 4, 238, 246, 251, 259 . 2, 81, 83 1, 195, 196, 200, 201 . 1, 777, 185 . 5, 186, 208 . 1, 130, 137 . 5, 27, 31 4, 226, 257, 259 . 4, 279, 297 2, 159 1, 109, 121, 122 . 5, 234, 242 . 2, 223, 232 . 2, 105, 114 . 2, 204, 209 . 4, 131, 135 . 4, 202, 205 . 4, 280, 297 3,99, 106, 115, 777, 779, 720, 72-^, 725, 126, 127, 128, 129, 130; 5, 33, 71, 80, 287 . 5,235,238,242 . 3, 42 . 2, 217, 232 4, 292, 293, 295, 297 . 4, 130, 135 . .2, 762, 772 Cumulative Author Index 15 Muntwyler, E. . . 4, 109, 113 1 Olbrich, O. 1, 30, 51, 52, 138, Murie, A. . 3,15,18:5,54,92, 201, 203, 223, 236, 237; 102 3, 31, 32, 35, 46, 125, 144, Murie, O. J. . 2, 184, 187 147, 148, 183, 190, 191, Murphy, D. P. . . 5, 262, 264 192; 4, 777, 772, 77i, 180, Murray, J. E. . 4, 45, 57 195, 248 Murrell, K. F. H . 3, 166, 169 Olesen, K. H. . 4, 102, 110, 113,113, Myers, G. B. . 4, 289, 296 114,299, 313,315 Myers, H. L. . 2, 41, 54 Oliver, B. B. Oliver, J. . . 4, 283, 298 4, 224, 250, 253, 254, 259 Nabarro, J. D. N . 4, 308 Oliver, J. A. . 3,26,29 Nadal, J. W. . 4, 39, 57 Olney, J. M. 4, 109, 113, 114 Nadell, J. . . 4, 265, 268 Olsen, A. G. . 2, 81, 83 Naeslund, J. . 1, 148, 159 Olsen, C. . . 1, 204, 209 Nagorniy, O. V. . 3, 31 Olsson, V. . . 5, 99, 103 Nagy, Z. . 4, 284, 286, 296 O'Neil, G. C. . 1, 198, 200 Nalbandov, A. V . 2, 81, 83 Opfinger, E. 5, 235, 236, 241 Nail, G. H. 5, 153, 165, 176 Ophuls, W. . 1,89 Natelson, S. 4, 40, 41, 57 Opie, E. L. 4, 23, 24, 32 Neher, R. . 4, 60, 292, 293, 295 Oppenheimer, E.T. . 3, 75, 90 Neil, N. W. 1, 148, 149, 158 Orent, E. R. . 4, 301, 308 Nelson, J. B. . 1, 179, 185 Orent-Keiles, E. '. 4, 224 Nelson, K. R. . . 4, 82, 93 Orr, R. T. . . 3, 10, 18 Nelson, N. 1, 130, 133, 143 Ortiz, E. . . 1, 143, 160 Nelson, W. P. . . 4, 45, 57 Orton, J. H. '. 5, 190, 191, 193, 198, Nesbett, F. B. . . 2, 141, 143 200, 201, 208 Neuberger, A. 1, 100,103,217, Osterlund, K. . 4, 203, 205 232 Outler, J. C. . 2, 148, 159 Neukirch, F. . 4, 41, 56 Owen, E. C. . 1, 770, 123 Neuman, R. E. . 1,91, 103 Owens, W. A 3, 163, 169, 172, 174, 182 Newman, E. A. . 1, 191, 193 Newman, E. V. 4, 277, 296, 297 Pacaud, S. . .3, 160, 169 Neyzi, O. . . 4, 146, 147, 149 Paccagnella, B. . . 5, 153, 176 Nichols, M. P. 1, 110, 120, 121, 122, Pace, N. . , . 4, 289, 298 139; 3, 78, 90 Padykula, H. A. 1, 106, 107, Nickel, H. K. . 5, 232, 242 108, 109, 111, 114 Nicolaysen, R. 1, 109, 111, 114, 115, Page, E. . . 1, 112, 114 12C ), 122, 123, 123, 124, Page, E. W. . 1, 129, 143 12t ), 205, 236, 237, 243; Pain, J. 5',235, 237, 238, 242, 243 3, 70, 147, 193 Palade, G. E. . 1, 93, 96 Nieberle, K. . 5, 110, 112 Pallaske, G. . 5, 110, 112 Nielson, P. E. . 2, 148, 159 Palmer, L. S. . 5,237,242,243 Nieto, D. . . 5, 107, 112 Pannabecker, R.F. 1, 3, 3, 11, 13 Nieuwkoop . 2, 56 Papanicolaou , G.N. . 1, 41, 54 Nigrelli, R. F. . 5, 113, 138, 147, 176, Papper, S. . . 4, 57, 57 175 i, 210, 211, 212, 213, Parets, A. D. . 3, 136, 142 21( 5, 217, 226, 226, 227, Park, O. . . 1, 10, 14 228, 229, 244, 245 Parker, F. . . 1, 112, 113 Nitsche, H. . 2, 779, 183 Parker, H. V. '. 4, 102, 102, 106, 108, Nixon, W. C. W. 2, 162, 172, 110, 113 173; 4, 90, 93 Parker, J. B. . 4, 130, 135 Nolazco, J. . 2, 86, 96 Parker, R. L. . 1, 112, 113 Nordmann, R. 4,220 Parker, R. R. . 5, 182, 208 Norris, A. H. 1, 66; 3, 77, 90, 91 Parkes, A. S. . 1, 139, 143, 154, 160, Northrop, J. H. 1, 196, 200; 3, 24, 29, 162, 163, 164, 165, 167, 101 168, 169, 170, 171, 172, Norton, H. W. . 1,81,83 203, 215, 216, 217, 245; Norton, R. . I, 198, 200 1, 16, 34, 49, 50, 53, 54, Nunes, J. P. . 1,36,54 55, 56, 57, 58, 60, 65, 66, 66, 127, 215, 216, 232, 239, O'Brien, J. R. P. 4, 36, 41, 43, 57 240, 241, 242, 243, 244, O'Connor 1, 138 245,246, 248, 249; 3,30, Ogbom, M. E. . . 5,5,75 99,116,124,126, 127,130, Ohlson, M. A. . 1, 122, 123 185; 4, 755 16 Cumulative Author Index Parmington, S. L. . . 4, 228 Parrish, D. B. . . .4, 200, 205 Parrott, D. M. V. . 4, 257, 259 Patterson, H. D. . 2, 35, 54 Pearce, J. W. . .4, 28, 57 Pearcy, M. . . . . 4, 75 Pearl, R. . \, 243\3, 24, 29, 137, 143; 5, 267, 283 Pearl, R. de Witt . 3, 137, 143 Pearson, E. G. . . . 5, 288 Pearson, J. A. . . 2, 206, 208, 209 Pearson, K. 1, 52, 238; 3, 8, 18; 5, 3, 15, 21, 31, 53, 54, 288 Pearson, O. P Pearson, P. B. Pedersen, S. Pemberton, J. Penkow Penn, J. Pepler, W. J. Perks, W. . Perley, A, . Perrone, J. C. Perry, S. M. Persike, E. C. Peters, J. P. Petrovitch, A. Petter, J. J. Petter-Rousseaux, A. Pettinari, V, Pfeiffer, C. A. Phillips, E. F. Phillips, E. W. PhiUips, J. B. PhiUips, J. C. Phillipson, A. T. Piaget Pickens, M. Pictet, A. . Piel, H. Pienaar, U. de V Pierce, E. C. Piggot, A. P. Pignard, P. Pilcher, C. Pincus, G. von Pirquet Pirre, G. D. Pitt, F. Pitts, R. F. Piatt, R. . Platts, M. M. Plentl, A. A. 3, 27, 28, 29 . 5, 237, 242 . 4, 39, 57 . 3, 94, 100 1, 145 3, 70, 96, 100, 106, 114 3, 102 5, 3, 4, 5, 14, 15, SI, 288, 289, 290, 291, 292, 293, 294, 302, 303, 311 4, 143, 149 1, 100, 103; 2, 217, 232 , 4, 200, 205 . 4, 257, 259 4, 38, 40, 57 . 2, 36, 37, 38, 39, 41, 42, 46, 47, 53, 62, 66 . 3, 28, 29 3, 17, 18, 28, 29 . 2, 33, 54 . 1, 151, 152, 160 . 5, 232, 234, 242 5, 4, 5, 15 . 5, 149, 176 . 1, 755, 160 . 4, 202, 205 . 1,47 . 4, 131, 135 . 1, 196, 200 5, 59, 63, 64, 69 . 3, 34, 36 . 1, 76 1, 18, 23; 3, 93, 100 .4,217,219 . 4, 287, 288, 296 1, 36, 48, 126, 126, 128, 130, 134, 136, 137, 219, 219, 223, 224, 225, 226, 227, 229, 230, 232, 233, 234, 236; 2, 36, 37, 54; 4, 90, 93, 187, 195, 262; 5, 267 1,201 1, 198, 199 3, 10, 18 4, 265, 268 4, 50, 56 4, 265, 268 4, 89, 93 Pohl, H. A. Polge, C. . Pollock, W. F. Polyakov, E. V. Ponder, E. Ponomareva, L. Pooler, N. R. Popjak, G. Pospelov, S. P. Potter, V. R, Power, M. H. Prader, A. . Prankerd, T. A. Price, D. . 1 Primrose, T. Pritchard, J. A. Pritchard, J. J. Pritchard, W. H Probst, R. T. Pucher, G. W. Puetter Pycha, R. L. Pyrkosch, W. Qasim, S. Z. Quinton, A. Quiring, D. P. . 2, 109, 113, 148, 158 1, 163, 164, 169; 2, 215, 232 . 2, 98 . 5, 53, 54 2, 217, 223, 232 I. . .5, 53, 54 3, 707 . 2, 277, 232 . 5, 53, 54 . 4, 133, 135 4, 41, 56 . 4, 60 J. 2, 238, 239; 4, 208 , 143, 160; 2, 3, 3, 4, 13, 14, 15, 17, 28 . 4, 89, 93 . 2, 235, 239 . 2, 123, 125 4, 277, 288, 297 5, 144, 153, 176 2, 203, 209, 210 . 1, 56, 57 . 5, 75i, 775 . 3, 134, 142 5, 755, 755. 775 . 4, 709, 772 . 5, 779, 133 Radhakrishnan, N Ragan, M. S. Rahm, M. P. G. Rahn, O. . Raiha, N. . Raitt, D. S. Rakowicz, M. Ramasarma, G. B Rand, R. W. Ranson, R. M. Rao, C. R. N. Rapoport, S. Rascoff, H. Rasmussen, A Rasquin, P. Raven, J. C. Ray, L. A. Raynaud, A. Reardan, J. B. Rector, F. C. Reed, R. . Regan Rehm, W. S. Reibisch, J. Reid, A. F. Reifenstein, E, Reinke, F. Relman, A. S. 5, 755, 176 1, 797, 200 3, 25, 29 1, 775, 185 4, 204, 205 . 3, 22, 30; 5, 7^9, 775, 198, 199, 200, 201, 202, 205, 208 . 5, 755, 775 . 1, 97, 94, 103; 3, 97, 100 . 2, 81, 83 3, 70, 18; 5,295 . 1,39,40,42,54 2, 237, 239; 4, 40, 46, 57, 287, 296 . 1, 198, 200 . 2, 95, 96 5, 196, 208, 218, 226 . 3, 160, 168 . 1, 777, 185 2, 3, 11, 13, 167, 172 4, 45, 56 . 4, 209, 219 3, 65, 66, 67, 96, 97, 100, 102, 103 . 2,56 4, 24, 31 5, 202, 208 4, 109, 113 1, 109, 122 2, 92, 96 4,222 Cumulative Author Index 17 Remington, J. W Renold, A. E. Renton, R. M. Renzi, A. A. R6v, J. Reynolds, S. R. M Ribbands, C. R. Ribot Rice, T. R. Rich, C. O. Richards, D. W. Richards, F. J. Richardson Richardson, I. M Richdale, L. E. Richet, G. . Richie, R. Richmond, J. Richter, C. P. Richter, W. Ricker, W. E. Rickham, P. Riesen, W. H. Riesman, D. Riess, B. F. Rigby, B. . Rijssel, Th. G. van Rinsler, M. G. Riondel, A. M. Ripke, G. . Rittenberg, D. Roberts, E, 1 Roberts, J. M. Roberts, J. R. E. Roberts, P. H. Roberts, R. B. Robertson, J. A. Robertson, J, J. Robertson, O. H, Robertson, R. N, Robertson, T. B. Robinson, C. V. Robinson, J. R. Robinson, O. J. Robinson, S. Robson, J. M. Rockstein, M. 4, 275, 277, 295 . 4, 280, 298 . 5, 189, 207 4, 296 . 4, 286, 296 2, 80, 83, 129, 143, 148, 159 5, 232, 233, 242 . 1, ii 5, 170, 175 5, 4, 15 3, 79, 91 2, 207, 209 2,200 3, 156, 169 5, 100, 101, 103 4, 99, 228, 262 4, 139, 141, 149 4 34 \ 4, 166,' 179 . 2, 197, 199 3, 11, 19; 5, 145, 149, 151, 155, 176 . 4, 98 . 1, 197, 200 1, 47, 48, 49 . 3, 171, 182 4, 89, 90, 93 3, 120, 121, 125 4, 89, 90, 93 4, 293, 297 5, 59, 69 2, 204, 209 91, 94, 103; 3, 97, 100 Rodgers, C. Rodolfo, A. Rosch, G. A. Roessle, R. Rottgerman, W. Rogers, J. B. Roguski Rokaw, S. N. Rolf, D. . 2, 81, 83 . 4, 75 1, 122, 123 2, 148, 158 5, 149, 176 1, 755, 166 2, 234, 239 2, 206, 208, 209 1, 777, 185; 5, 7i7 2, 130, 143 4, 23, 24, 32, 214, 219 5, 59, 69 . 4, 76 2, 168, 172 . 5, 15, 31, 55, 83, 84, 112, 114, 179, 210, 226, 228, 232, 234, 243, 244, 245, 246, 247, 247, 248, 249, 250, 251, 254, 255, 256, 257, 258, 264, 265, 267,268,284,287,289,291 4, 289, 298 2, 109, 114, 118, 125 5, 243 . 4, 257, 259 5, 59, 62, 69 3, 77, 19, 20, 30 . 4, 33 . 4, 280, 294 . 4,256,257,259 Rollason, H. D. . Rollefsen, G. Rollins, R. Z. Romanoff, A. J. Romanoff, A. L. Romanoff, L. P. . 4, 252, 259 . 5, 149, 176 . 5, 258, 264 3, 20, 21, 30 3, 20, 21, 30 1,725, 136, 137,219, Rominger, E, Rook, J. A. F. . Rorig, A. . Rorschach, H. Rosenbaum, J. D. Rosenberg, E. K. Rosenberger, C. R. Rosenman, R. H 27P, 223, 224, 225, 226, 227, 230, 232, 233, 234, 236; 4, 90, 93, 187, 195 4, 130, 135 4, 301, 307, 30, 3098 2, 179, 183, 184, 185, 187 . 3, 181, 182 . 4, 45, 51, 56, 57 . 1, 139; 3, 78, 90 . 5, 255, 264 . 3, 137, 142 Rosenthal, T. B. 1, 91, 99, 103; 3, 97, 100 Ross, G. Ross, M. A. Rotblat, J. Roth, C. B. Rothman Rothschild, P. Roulet, F. . Roulet, F. C. Rounsefell, G. A Rous, P. . Roux, M. . Rowlands, I. W Rowntree, L. Rowson, L. E. A. Rubin, B. L. 7i7, Rubini, M. E. Rubner, M. Rudzinska, M. A Ruger, H. A. Rumbauer Runnstrom, S. Rupple, B. Russell, E. S. Russell, W. I. Russell, W. L. Ryberg, C. Rynearson, E. H 4, 277, 297 . 4, 256, 259 . 5, 15, 19, 20, 86, 87, 103, 135, 177, 209, 210, 283, 289, 290, 291, 292 2, 765, 772 2,207 2, 757, 755 4, 257, 259 1, 16,23; 3,93, 100 . 5, 797, 208 2, 234, 236, 239 3, 70, 96, 100 1, 142, 159; 2, 29, 66, 69, 81, 83, 83, 84, 85, 214, 260 4, 95 '. 1, 164, 'l69 1, 54, 126, 130, 134, 138, 139, 140, 237; 4, 90, 93 . 4,^5,57 3, 23, 30; 5, 124, 125, 126, 127, 129, 130, 133, 134 1, 187, 193, 197. 200; 3, 4, 19 ' . 3, 777, 752 5, 59, 63, 69 . 5, 149, 176 1, 160 5, 190, 199, 208 1, 155, 156, 161 . 3, 775, 725 . 4, 209, 219 . 4, 41, 56 Sabine, J. C. Sacher, G. A. . 2, 235, 239 5, 75, 18, 31, 52, 54, 80, 87, 88, 104, 115, 775, 727, 133, 133, 134, 135, 755, 136, 137, 138, 138, 139, 140, 141, 147, 176, 245, 266, 282, 284, 289, 292, 295, 296, 29a 18 Cumulative Author Index Sagild, U. Sakagami, S. F Salazar Samokhvalova, G Samuels, A. J. Sander, G. Sandulesco, G. Santos, R. F. Sapsford, C. S. Sarkar, B. C. R Sato, R. . Sawyer, W. H. Saxl, H. . Saxon, L. . Saxton, J. A. Sc6bat, L. Schaffer, J. Schaumann, O. Scheerer, M. Scheidegger, S. Scheinberg, P. Scherer, H. J. Scheyer, W. J. Schiefferdecker, P Schieren, J. Schilling, A. Schloerb Schloerb, P. R. Schmidlin, J. Schmidt, C. F. Schmidt, L G. Schmidt, K. P. Schneider, C. L. Schneider, F. Schoenheimer Schoenheimer, R Scholz, W. Schotterer, A. Schroder, G. Schroeder, H. A. Schuler, W. Schultze, M. O. Schultze, W. Schuster, D. Schutzman, F. L Schwartz, I. Schwartz, I. L. Schwartz, R. Schwartz, W. B. Schwarz Schwarz, J. Schwarz, W. Scott, J. P. Scott, W. C. Scribner, B. H. 96, Scrimshaw, N. S. Segar, W. E. Seldin, D. W. . Selkurt, E. W. . 4, 109, HI, 113 5, 233, 243 2, 64, 66 5,209 4, 266, 268 3, 137, 143 1, 130, 137 4, 46, 57 2, 86, 96 5, 237, 243 5, 151, 176 4, 167, 179 3, 97, 98, 100, 103; 5, 113 . 4, 51, 57 . 4, 253, 257, 259 4, 273, 278, 279, 296 . 2, 196, 199 . 4, 167, 179 . 3, 180, 182 5, 103, 106, 112, 113, 226 3, 147', 4, 286, 297 . 5, 108, 112 . 5, 89 . 2, 188, 199 . 5, 59, 69 1, 110, 122, 123 4, 269 . 4, 109, 113 . 4, 181, 195 . 4, 286, 296 . 2, 39, 54 . 1, 10, 14 . 2, 112, 114 . 5, 240, 243 . 3, 48, 49 . 2, 204, 209 . 5, 707, 112 . 5, 43, 54 . 5, 65, 69 . 4, 288, 297 . 4, 187, 195 . 1, 192, 193 1, 28, 67, 217 . 1, 42, 48 . 4, 301, 308 . 4, 274, 297 4, 24, 32, 67, 73, 77 4, 129, 132, 133, 135 4,222 3, 702, 70i 5, 238, 242 4, 29, 32 2,171, 182 4, 709, 772 4, 46, 56, 93, 100,101,114,222,249, 269 . 1, 198, 200 . .4, 133, 134 . 4, 209, 219 . 4,274,277,297 Sellers, A. L. Selwyn, J. G. Selye, H. . Sen, P. K. . Serane, J. . Serpen, G. Severinghaus, E. Seville, R. Seymour, W. M. Shaeflfer, K. Shannon, C. Sharaf, A, A. Sheldon, D. S. Sheldon, J. H. 4, 257, 259, 261 . 4, 199, 205 4, 256, 259; 5, 88 2, 204, 207, 209 4, 262 . 2, 30 1, 143, 153, 160 . 3, 96, 100 B. . 4,271,288,297 . 4, 130, 135 . 2, 230, 232 . 2, 168, 172 . 4, 109, 113 3, 94, 100, 112, 114; 4, 264, 268 Shell, H. M. . . .5, 237, 243 Sherman, H. C. . . 1, 197, 200; 4, 132, 133, 135 Shiraishi, Y. . . .5, 757, 775 Shirley, R. J. . . .2, 148, 159 Shock, N. W. . 1, 52, 53, 55, 66, 66, 110, 120, 121, 122, 139, 139, 206, 207, 207, 213, 214, 217, 220, 223, 224, 229, 230, 231, 232, 236, 237, 242, 244; 3, 77, 78, 79, 84, 88, 89, 90, 91, 151, 168; 4, 11, 72, 777, 772, 77i, 114, 115, 164, 207, 225, 226, 229, 230, 231, 232, 233, 235, 236, 237, 238, 240, 241, 242, 243, 244, 245, 246, 247, 247, 248, 249, 250, 264, 268, 277, 295, 314, 315 Shohl, A. T. . . .4, 143, 149 Shorr, E 4, 280, 295 Sieker, H. O. . 4, 283, 284, 296, 297 Silberberg, M. . . .1, 189, 193 Silberberg, R. . . .1, 189, 193 SilUman, R. P. . . .5, 149, 176 Silver, H. M. . . 3, 75, 80, 90, 91 Silversten, I. . . .1, 180, 185 Silverstone, F. A. . . 1, 237; 3, 77, 80, 90, 91 Simms, H. S. . 5, 72, 73, 74, 75, 76, 79, 266, 268, 293 Simon .... 3, 775 Simon, E. R. . . .4, 200, 205 Simon, E. W. . . .2, 208, 209 Simonson, E, . . .3, 75, 91 Simpson, A. C. . 5, 203, 204, 205, 208 Simpson, M. E. . . 1, 29; 2, 33, 54 Simpson, S. A. . . 4, 292, 293, 296 Simpson, T. . . .4, 267, 268 Sinclair, H. M. . 1, 194, 195, 197, 200, 201, 202, 203, 205, 207 Sinclair, Sir J. . . .1, 194, 201 Sinclair-Smith, B. . .4, 277, 296 Singer, B. . . .4, 181, 195, 297 Singer, M. . 2, 775, 178, 183, 185, 186, 187 Singer, R. B. . . .4, 289, 297 Cumulative Author Index 19 Sirota, J. H. Sjogren, H Sjogren, T. Skjelkvale, L. Slack, H. G. B. Slater, P. . Sloviter, H. A. Sloviter, J. A. Sluiter, J. W. Smith Smith, A. D. Smith, A. H. Smith, A. U. Smith, A. V. Smith, G. S. Smith, G. van S Smith, G. V. Smith, H. . Smith, H. L. Smith, H. W. Smith, I. H. Smith, J. Maynard Smith, K. R. Smith, L. E. Smith, L. L. Smith, M. Smith, O. C. Smith, P. E. Smith, P. K. Smyly, W. J. Sniffen, R. C. Snyder, D. P. Snyder, F. F. Sobel, E. H. Soberman, J. S0bye, P. . Solomon, A. K Solomon, D. H. Solti, F. . Sommerson, W. H Somerville, I. F. . Sonneborn, T. M. 4, 288, 295 3, 134, 143 3, 134, 143 1, 109 1, 100, 103 1, 37, 48 1, 164, 169 2, 215, 232 5, 97, 103 5, 59, 69 2, 215, 216, 232 4, 120, 121, 128, 131, 135, 254, 259 1, 163, 164, 165, 167, 168, 169; 2, 215, 232 2, 33, 53 . 1, 18 1, 145, 161 2, 112, 114 4,227 3, 74, 75, 76, 91 4, 230, 246, 277, 298 . 1, 198, 201 5, 16, 32, 33, 55, 86, 87, 104, 134, 137, 141, 266, 267, 268, 271, 278, 279, 281, 281, 282, 283, 284, 285, 290 3, 151, 169 1, 230, 236 5, 755, 174 3, 9, 19 2, 112, 114 1, 145, 160 4, 120, 121, 128, 131, 135 P. 5, 145,153,169,175,176 2, 91, 96 5, 95, 103 2, 109, 112, 113, 114 . 4, 141, 149 4, 295 . 3, 137, 143 . 4, 109, 112 1, 139, 223, 224, 229, 232, 236; 3, 78, 79,90,91; 4, 235, 237, 246 4, 286, 296 Sontag, L. W. Soudek, S. Southern, H. N Spacek, B. Spahr, A. . Spector, H. Spence, J. . Spencer, A. G. Spencer, M. P. Sperhng, G . 2, 130, 143 . 2, 112, 114 2, 252; 3, 16; 5, 32, 35, 53, 54, 262, 264 2, 162, 172 5, 235, 243 5, 101, 103 4, 291, 296 . 4,60 4, 131, 135 3, 95, 100 4,308 4, 274, 277, 297 1, 188, 193, 197, 200; 2,189,199;3,8,18;S,251 264 Sperry, W. M. Spiegelman, M. Spielmeyer, W. Spies, T. D. Spitskaya, T. D. Spohde, H. Sprague, P. H. Spray, C. M. Sproul, E. E. Spurway, H. Squires, R. D. Stanbury, S. W. Stanier, M. W. Stark, W. . Stavraky, G. W. Stead, E. A. Stearns, G. Steenburg, R. W Steggerda, F. R. Steinbach, H. B. Stephenson, R. P Stevens, A. R. Stevens, C. F. Stevenson, J. A. F Steward, F. C. Stewart, J. M. Stewart, W. B. Stigall, C. . Stockard, C. R. Stockklausner, F Stoerk, H. C. Stoesigger, B. Stolbova, A. StoU, G. . Stowers, J. M. Strangeways, W. Strauss, F. Strauss, M. B. Strawn, K. Strong, L. C. Strube, H. . Stuart-Harris, C. H Suau, P. . Subbarow, Y. Sumner, F. B. Sumner, J. B. Summers-Smith, D Sundnes, G. Sutow, W. W. Sutter, J. . Svardson, G. Svoboda, J. Swanson, M. A. Swanson, P, P. Swanson, W. W, Sward, K. Swezy, O. . Swift, R. W. 4, 23, 32 3, 137, 142 1, 33, 48 1, 198, 201 5, 53, 54 5, 64, 69 3, 74, 75, 76, 91 4, 120, 121, 128, 136, 306, 308 . 4, 253, 259 . 5, 276, 281 . 4, 289, 297 4, 50, 56, 193, 195, 267, 268 4, 128, 130, 135, 216 . 1, 194, 201 4, 64, 65, 73 4, 51, 57, 271, 277, 298 4, 130, 135 .4,114 4, 120, 121, 131, 135 4, 22, 32 4, 167, 179 2, 228, 232, 237, 238 . 5, 21, 31 1, 223, 232, 236 . 2, 207, 209 . 2, 236, 239 . 2, 236, 239 . 4, 130, 135 . 2, 41, 54 . 5, 59, 69 4, 292, 293, 296 .3, 171, 182 . 5, 53, 54 . 4, 109, lis 4, 308 M. B. . 4, 211, 219 2, 56, 66, 67, 68, 160, 174, 185, 200 4, 45, 51, 56, 57 5, 755, 775 1, 747, 767 3, 134, 142 4,269 5, 149, 177 4, 720, 7i5 3, 9, 19 1,177 5, 99, 101, 103 5, 169, 176 4, 105, 113 5, 21, 31 5, 777, 775, 182, 190, 208 . 5, 235, 243 . 2, 136, 143 . 1, 722, 72i- . 4, 130, 135 1,48; 3, 159, 169 2, 35, 41, 53, 54, 55 . 4, 127, 135 20 Cumulative Author Index Swyer, G, I. M. de Sylva, D. Szabo Szabo, G. . Szafran, J. Szasz, J. Szilard, L. . 2, 774, 7 75; 4,13,78,93, 94,95, 97, 98, 115, 197, 198,227, 260, 313, 316 . 5, 755, 775, 228 3, 103 . 4, 284, 286, 296 1,211,213; 3, 760, 162, 169 . 4, 286, 296 5, 19, 129, 133, 289 Tabah, L. . Taber, R. D. Taffel, M. Tainter, M. L Tait, J. F. Takacs, L. Talbot, F. G. Talbot, N. B. Talke, L. . Tallqvist, H. Tanaka, T. Tandler, J, Tang, C-T. Tannenbaum, A Tanner, J. M . 5, 21, 31 5, 92, 94, 103 . 4, 38, 56 5, 136 4, 292, 293, 296 4, 45, 57, 284, 286, 296 3, 9, 18 4, 13, 32, 64, 73, 76, 94, 97, 100, 705, 77i, 137, 139, 139, 140, 141, 143, 146, 147, 149, 150, 151, 152, 162, 163, 226, 260, 261, i77 2, 7P7, 799 4,227 3, 24, 30 2, 779, 183 2, 747, 143 1, 797, 207 4,80, 93; 5, 84, 85, 86, 88, 266, 290, 296, 297 Tanquary, M. C. . .5, 232, 242 Tattersall, R. N Tauber, O. E. Tayler, R. Q. C. Taylor, A. . Taylor, C. C. Taylor, D. J. Taylor, H. L. Taylor, W. H. . Ted, H. M. Templeman, W. . Templeman, W. G. Templeton, H. A. Tener, J. S. Tengbergen, W. van E Tenney, B. Terao, A. . Terman, L. M. Terpenning, J, G Terry, M. , Tester, A. L. Thaysen, J. H. Theard, A. Thelander, H. Thiele, T. N. Thomas, W. A 3, 96, 100 5, 277, 281 4, 40, 57 4, 116 5, 149, 167, 177 3, 188 4, 757, 161, 162 4, 36, 41, 43, 57 . 2, 762, 772 5, 146, 151, 177 . 2, 207, 209 . 1, 797, 207 . 3, 70, 18 . 3, 720, 725 2, 770, 772, 114 . 3, 24, 30 1, 40, 48; 3, 772, 174, 182 . 5, 182, 208 4, 7i9, 140, 141, 145, 146, 147, 149 . 5, 149, 177 4, 62. 63, 64, 65, 67, 68, 69, 72.73, 73, 74, 74, 75, 76, 77, 77, 94, 99 4, 273, 278, 279, 296 2, 720, 727, 725, 727 5, 3, 15 4, 81, 93, 262 Thomasson, B. . Thompson, J. C. Thompson, J. F. Thompson, W. F. Thomsen, A. C. . Thomsen, K. Thomson, A. M. Thorn, D. W. . Thorn, G. W. . Thorn, N. A. Thorndike, E. L. Thorndike, R. L. Thung, P. J. Thurstone, L. L. Tibbo, S. N. Tiews, K. . Timoner, J. Tizard, Todd, F. E. Topper, Y. J. Toro, G. . Tosteson, . Townsend, C, H. Tracey, K. M. Trowbridge, P. F Tschertok . Tsuchiya, K. Tuchmann-Duplessis, H. 4, 273, 283, 298 4, 79, 93, 97, 98 . 2, 207, 209 . 5, 757, 777 . 4, 41, 56 2, 110, 112, 114 . 5, 33 . 4, 82, 93 4, 82, 93, 280, 298 4, 63, 65, 67, 68, 73, 76 3, 762, 769, 772, 182 . 3, 178, 182 3, 779, 727, 725 . 3, 181, 182 . 5, 149, 177 . 5, 755, 777 . 4, 266, 268 1, 138 . 5, 237, 243 . 4, 200, 205 1, 770, 727, 722 4, 206 . 5, 277, 226 . 5, 262, 264 . 4, 757, 755 1, 145 . 2, 189, 199 2, 84, 127, 161, 765, 168, 172, 173, 174, 175, 245 Tucker, D. W. . .5, 174, 177 Tucker, W. A. . . .3, 766, 769 Tudvad, F. . . .4, 726, 756 Tufts, E. V. . 4, 507, 307, 308, 309 Tunbridge, R. E. 1, 1, 23, 48, 50, 52, 53, 54, 57, 67, 68, 104, 105, 107, 108, 125, 137, 138, 205, 206, 238, 241, 245; 3, 46, 65, 66, 67, 69, 71, 72, 92, 96, 97, 100, 102, 103, 128, 182, 183 Turner, CD.. . .2, 18, 27 Turner, J. R. Ueberwasser, H. Ullmann, E. A. . Unna, P. G. Vaas, K. F. Vacek, Z. . Valach, A. Vallois, H. V. . Van Cleave, H. J. Vanderlinde, R. J. Van-Eck, G. J. V. Van Heerdt, P. F. Van Slyke, D, D. Van Wagenen, G. Vara, P. . Varga, F. . Varnauskas, E. . . 2, 236, 239 . 4, 757, 795 . 3, 52, 59 . 3, 96, 100 . 5, 770, 777 4, 165, 755, 187 . 4, 291, 296 5, 134 5, 179 . 4, 709, 772 2, 36, 44, 45, 46, 54, 55 . 5, 97, 103 4, 248, 272, 298 . 1, 755, 767 . 4, 203, 205 . 4, 767, 762 4, 273, 283, 298 Cumulative Author Index 21 Velardo, J. T. . . .2, 81, 83 Verney, E. B. . . .4, 37, 57 Venning, E. H. . 4, 89, 93, 293, 294 Verschuer, O. v. . . 3, 137, 431 Verzdr, F. 1, 24, 57, 67, 68, 108, 171, 204, 238; 3, 30, 60, 63, 64, 66, 67, 68, 68, 69, 70, 94, 128, 194; 5, 82, 85, 86, 112,133,137,138,141,298 Verzar, J 1, 215, 216 Verzdr-McDougall, J. 3, 64, 68, 187 Vair, W. N. . . .4, 283, 298 Vickery, H. B. . . 2, 203, 209, 210 Videbaeck, A. . . .4, 241, 246 Villee, C. A. 2, 17, 28, 107, 112, 114, 117, 129, 136, 141, 142, 143, 144, 144, 145, 146, 147, 212, 213, 242, 243, 245, 249, 250, 253 F. . . . 1,37,48 L. . 1, 16. 23, 25, 49, 66; 3, 93, 100 B. . . hl79, 180, 185 5, 35, 43, 46, 49, 54 . 5, 149, 177 5, 237, 238, 242, 243 . 3, 10, 18 . 5, 149, 177 1, 35, 48, 49 . 2, 184, 187 1, 35, 48, 49; 3, 147 . 4, 161, 162 . 5, 237, 243 2, 111, 112, 113, 148, 159 . 2, 34, 53 Vincent, D. Vischer, A. Visscher, M. Vitt, V. o. Vives, F. . Vivino, E. Vizoso, M. VNIRO Vogt. C. . Vogt, F. . Vogt, O. . Vonoczky, J. Voogd, S. . Vosburgh, G. Voss, H. E. V, Wackwitz, J. D Waelsch, L. Wagenen, G. van Waggoner, R. W Wagman, I. H. Wahl, O. . Wakman, A. J. Walaas, E. Walaas, O. Waldeyer, L. Waldo, C. M. Walford, L. A. Walker, A. R. P. Walker, E. P. Walker, W. G. Wallace, B. Wallace, J. G. Wallace, W. Wallace, W. M Wallart, J. Walliker, C. Walton, W. S, . 3, 165, 169 . 2, 197, 199 2, 33, 54 . 3, 134, 143 . 3, 77, 91 . 5, 235, 243 . 2, 203, 210 . 2, 133, 144 . 2, 133, 144 . 1, 145, 161 2, 777, 178, 179, 180, 182, 183, 184, 187 5, 159, 177, 214, 226 3, 110, 114, 189 . 5, 119, 133 . 2, 148. 159 . 5, 20 . 3, 163, 168 . 5, 189, 208 4, 34, 58, 59, 74, 75, 98, 116, 120, 129, 132, 133, 134, 135, 136, 137, 138, 152, 164, 224, 225, 226, 227, 316 . 1, 145, 161 .1, 197,200 . 3, 95, 100 Wang, H. . Wang, H. W. de Wardener, H. Ward, B. . Warkany, J. Warming-Larsen Warner, F. G. Warnock, W. M. Warren, F. L. Warren, J. V. Watchorn, E. . 2, 34, 54 . 2, 129, 144 4, 54, 57, 150 . 2, 30 . 2, 165, 172 . 4, 75 . 4, 289, 298 4, 309 4, 202, 203, 205 . 4, 271, 298 4,100 Watkin, D. M. 1, 206, 207; 3, 79, 88, 91; 4, 226, 231, 233, 244, 246 Watson, M. Watts, R. M. Waymouth, C. Weatherford, H Weaver, N. Weaver, W. Webb, C. S. Weber, R. . Weidenreich, F. Weil, W. B. Weir, J. F. Weisman, R. Weismann . Weiss, J. M. Weitnauer, E. Weizel, N. C. Welford, A. T Wells, H. G. Wells, L. J. Welt, L. G. Wendt, W. E. Wener, J. . Wenner, R. Werko, L. . Werther, M. Wertman, M. Wesman^ A. G. Wesson, L. G. West, A. S. West, C. . West, C. D. Westman, A. Weston, H. C. Weston, R. E. Wettstein, A. Wetzel Weyer, F. . Weygand, F. Wheeler, N. C. Whipple, G. H. White, A. G. White, H. L. White, P. White, V. K. Whitehouse, A. 2, 86, 96 1, 151, 158, 160; 2, 162, 172 1,191, 193 L. . .2, 176, 183 5, 237, 243 2, 230, 232 5, 104 5, 43, 54 5, 134 4, 116 . 4,95 2, 235, 238 . 3,4 2, 101, 103 5, 104 1, 198, 201 1,41, 48, 209, 210, 213, 215; 3, 71, 149, 160, 162, 169, 182, 183, 184, 186, 187 . 1,89 2, 3, 10, 11, 12, 13 . 4, 75 . 4, 283, 296 4, 297 . 1, 151, 161 4, 273, 279, 280, 283, 295, 298 . 4, 62, 73 . 4, 301, 308 . 3, 174, 182 . 4, 277, 298 . 5, 258, 264 . 2, 206, 209 4, 40, 46, 57, 120, 121, 132, 134 2, 49, 54, 145, 161 3, 162, 164, 169 . 4, 280, 295 4, 94, 181, 195 .3,184 . 5, 234, 243 . 5, 237, 243 4, 275, 277, 295 4, 132, 133, 135 . 4, 99 4, 256, 257, 259 5, 267 . 2, 141, 144 G. R. . 4, 63, 73 22 Cumulative Author Index Whitney, L. F. . . .1, 755, 161 Widdas, W. F. . . 2, 118, 119, 120, 125;4, 201,203, 205 Widdowson, E. M. . 1, 186, 195, 198, 201; 4, 13, 96, 97, 113, 114, 120, 121, 128, 136, 136, 137, 160, 162, 163, 164, 180, 195, 199, 203, 205, 208, 209, 215, 217, 219, 221, 222, 305, 306, 308, 309 Wieland, P. . . .4, 181, 195 Wiesner, B. P. . Wigglesworth, V. B. 243, Wilde, E. Wilde, W. S. Wilens, S. L. Wilkes, A. Wilkins, L. Wilkinson, C. F. Wilkinson, E. Williams, I. T. D. Williams, J. N. . WilUams, P. C. Williams, R. G. Willis, K. . Willius, F. A. Wilson, C. Wilson, C. P. Wilson, I. . Wilson, J. W. Wilson, R. C. Wiltner, W. Wiltshire, G. H Wimsatt, W. A. Winiwarter, H. de Winnigstedt, R. Winsatt, W. A. Winslow, J. A. Wise, G. H. Wislocki, G. B Witschi, E. Wittwer, S. H. . Wohlschlag, D. E. Woke. P. A. Wolfe, J. M. Wolff Wolff, H. P. Wolkoff, K. Woll, E. . 1, 152, 161 5, 87, 133, 240, 243, 245, 267, 285 . 1, 198, 201 2, 111, 112, 113, 148, 159 4, 253, 259 5, 258, 264 4, 79, 93 3, 137, 143 4, 6, 11 4, 283, 295 4, 133, 136 2, 58, 59, 60, 63, 66, 66, 67, 68, 145, 174, 175, 213, 242, 243, 247, 249, 252 2, 91, 95, 96 . 4, 283, 298 3, 74, 75, 76, 91 . 4, 258, 259 . 4, 287, 296 . 5, 21, 31 2, 251 . 1, 40, 48 4, 284, 286, 296 . 2, 204, 209 2, 155, 158, 159 . 2, 42, 54 . 5, 59, 69 . 2, 106, 114 4, 275, 277, 295 . 4, 200, 205 2, 27, 98, 103, 104, 105, 106, 107, 108, 109, 110, 112, 113, 114, 115, 116, 129, 144, 147, 159, 176, 176, 177, 178, 179, 180, 181, 182, 183, 184, 184, 185, 185, 186, 187, 191, 192, 198, 241, 243, 246, 248, 249, 251; 4, 217, 219 2, 14, 15, 17 . 5, 237, 243 5, 145, 151, 177 . 5, 255, 264 . 1, 146, 161 . 2,56 4, 280, 292, 293, 298 5, 110, 112 1, 145, 161 Wolstenholme, G. E. Wood, G. C. Wood, J. G. Wood, M. J. Wood, T. R. Woodford- Williams, Woodger, J. H. . Woollard, H. H. Worcester, J. Wright, A. W. Wright, N. C. Wright, P. L. Wrong, O. Wurzel, W. Wussow, W. Wyburn, G. M. Wynn, V. . Wynne-Edwards, V. W. . .5, 71 3, 66, 68, 98, 100 . 1, 203, 210 3, 97, 100, 103 . 3, 24, 29 4, 111, 112, 113, 180, 195 . 2, 86, 96 . 2, 189, 199 . 4, 64, 73 . 1, 146, 161 . 5, 59, 69 . 2, 81, 83 4, 46, 57, 266, 268 . 5, 59, 69 5, 66, 67, 69 . 2, 41, 53 . 4, 145, 149 5, 193, 194, 195, 208 Yannet, H. Yemm, E. W. Yerkes, A. Yerkes, J. . Yerkes, R. Yerushalmy, J Yiengst, M. J. Yockey, H. P. Yokota, T. Young, F. G. Young, I. M. Young, L. E. Young, P. T. Young, W. C. Young, W. F. Yudkin, J. . 4, 106, 113 . 2, 145, 202, 203, 204, 208, 209, 210, 210, 211, 212, 213, 214, 241, 242, 249 . 5, 119, 133 . 3, 179, 182, 185 . 5, 119, 133 5, 21,31, 53, 54 1, 206, 207; 3, 77, 88,91;4,242,244,245, 246, 247 . 5, 129, 133 . 5, 155, 163, 177 162, 166, 167, 172, 174 2, 221, 232, 236, 238 . 2,236,238,239 . 4, 166, 179 1, 149, 161; 2, 41, 54 4, 38, 40, 41, 56, 57, 59, 59, 96, 162 1, 192, 193; 4, 209, 219 5, 267 . 5, 53, 54 . 5, 232, 243 M. . .2, 179, 183 3, 147 5, 59, 60, 64, 69 3, 190 . 3, 133, 142 . 2, 86, 96 . 3, 95, 100 2, 191, 192, 199 1, 144, 145, 149, 160, 161; 2, 15, 28, 31, 31, 32, 33, 34, 35, 38, 49, 50, 51, 53, 54, 55, 56, 57, 58, 60, 65, 66, 66, 67, 98, 126, 184, 187, 201; 3, 127; 4, 82, 93, 251, 259 Zweifach, B. W. . . 4, 280, 295 Zweymiiller, E. . .4, 220, 221, 246 Zahl, P. A. Zamyatin, N Zander, E. Zawadowsky, M. Zeman, F. D. Ziegenhagen, G Zilh, A. . Zinina, N. V. Zlotnik, J. . Zonneveld, R. J Zorzoli, G, Zuckerman, S SUBJECT INDEX Figures in heavy type indicate volume number. Figures in ordinary type indicate page number. Acid-base balance, changes in due to age, 4, 224-245 development of, 4, 209-223 during menstrual cycle, 4, 93 in foetal life, 4, 217-219 in old age, 4, 242-243 Accipitres, arteriosclerosis in, 5, 109 Achievements in later life, 1, 39, 40, 50, 51 Achilles' tendon, age changes in structure of, 3, 68, 69 Acid phosphatase, in placenta, 2, 107, 108 Acipenser fulvescens, survival curves of, 5, 143, 144 Acipenser ruthenus, oldest age of, 5, 191 Acipenserifonnes, lifespan of, 5, 220, 222 lifespan and size of, 5, 152 ACTH, administration of in aged schizo- phrenics, 1, 220-236, 237 effect on adrenals, 4, 175 effect on glycogen storage, 2, 24-25, 29 effect on hair, 2, 175 effect on potassium excretion, 4, 176, 177, 178 effect on schizophrenic patients, 1, 229 effect on sodium excretion, 4, 176, 177, 178 effect on water loss, 3, 176, 177, 178 response to in old age, 1, 139 Actuarial aspects of human lifespan, 5, 2-20 Adaptation, decHne of, 3, 62, 63 definition of, 3, 61 in study of ageing, 3, 60-67, 71 Adenosine triphosphate, in placenta, 2, 107, 108 in red cells, 2, 238 Adjustment in old people, 1, 41, 42, 43, 46, 47 Adolescence, water and electrolyte changes during, 4, 80-81 Adrenal cortex, activity of, in elderly schizophrenic patients, 1, 219-238 mitochondria in, 2, 101-102, 103 relationship with pituitary gland, 2, 22 Adrenal corticosteroids, excretion of, changes due to age, 4, 91 Adrenal glands, control of sodium intake by, 4, 166 effects of ACTH and cortisone, 4, 175- 176 effect of castration, 4, 197-198 effect on diuresis, 4, 1 3 mitochondria in, 2, 101, 102 Adrenal hyperplasia, effect on water and electrolytes, 4, 79-80 Adrenal steroids, biosynthesis of, 1, 126- 128, 129, 133 effect of age on influence of, 4, 192-194 effect on kidney, 4, 257, 262 effect on water and electrolyte excre- tion, 4, 180-194, 196-198 excretion of, age changes in, 1, 126-140 relationship with urinary ketosteroids, 1, 128-130, 133-139 Adrenalectomy, effect on rats, 2, 58 Adrenaline, effect on water diuresis, 4, 9, 14 in emphysema, 1, 67 Adrenosterone, administration of in old age, 1, 134-136 Adults, water in body of, 4, 106-110 Aedes aegypti, effect of diet on, 5, 255 Africans, ageing in, 3, 34, 104-114 carcinoma in, 3, 104, 106, 113 liver disease in, 3, 188-190 malnutrition in, 3, 104-114, 145, 188- 190 mental disease in, 3, 145 nutrition in, 3, 188-190 Age, body water changes due to, 4, IIQ- 112, 114, 115 causing changes in acid-base balance, 4, 224-225 causing changes in effect of aldosterone on urine, 4, 182-187 causing changes in effect of pitressin, 4, 239-240 causing changes in extracellular water, 4,31, 110-112, 114-115 causing changes in glomerular filtration rate,4, 231,238, 246 causing changes in haemoglobin, 4, 203, 206, 207 causing changes in hormonal control of homeostasis, 4, 168-179 causing changes in intracellular water, 4,110-112,114,115 causing changes in nitrogen excretion, 4, 243 causing changes in oestrogen excretion, 4,91 causing changes in steroid metabolism, 4, 90-92 cellular changes due to, 4, 199-205 changes due to in fish, 5, 218 23 24 Cumulative Subject Index Age changes in ketosteroid excretion due to, 4,91 effect of fecundity in fish, 5, 198-200, 201 effect on blood volume, 4, 243 effect on cells, 4, 199-205 effect on deer antler growth, 2, 183-184 effect on diuresis, 4, 6-10 effect on fecundity in fish, 5, 201 effect on homeostasis, 4, 139-153 effect on influence of adrenal steroids, 4, 192-194 effects on insects, 5, 247-268 effect on kidney, 4, 11-12, 227-228, 229-249, 253-254 effect on oestrogen excretion, 4, 91 effect on renal disease, 4, 250-263 effect on reproduction, 5, 181-182 effects on reproduction in fish, 5, 186- 206 effect on starvation, 4, 226 effect on water diuresis, 4, 238-240 electrolyte changes due to, 4, 241, 311- 312 erythrocyte changes due to, 4, 199-205, 207 haemoglobin changes due to, 4, 203, 207 parental, and lifespan, 5, 21-34 pulmonary effects of, 4, 264 renal effects of, 4, 11-12, 227-228, 229- 249, 253-254 Age factor in prenatal endocrine events, 2, 18-30 Ageing, actuarial measurement of, 1, 7-14 adaptation in, 3, 60-67, 71 and endocrine glands, 2, 161-162 biological approach to, 3, 2-17 clinicopathological tests of, 3, 92-100 definition of, 1, 4-14, 55, 56, 57, 242, 243, 244, 245; 2, 246-248, 249; 3, 73, 93, 131-132, 191 due to rapid maturation, 1, 194-199 effect of environment, 3, 35, 37, 45, 48, 171, 183-184 effect on skill, 1, 209-218 emotional changes during, 3, 170-182, 184 genetic factors in, 1, 26, 27, 238-241 ; 3,131, 132-133, 144; 5, 137 genetics of, in Drosophila, 5, 278-280, 283 twin data in, 3, 131-148 horizontal studies of, 3, 95 in Africans, 3, 104-114 in Drosophila, 5, 269-285 in leaves, 2, 202-214 in red cells, 2, 233-238 intelligence changes in, 3, 170-187 longitudinal studies of, 3, 95 measurement of, 1, 4-15; 3, 6-8 mental aspects of, 1, 32-52 molecular changes in, 5, 129-131 Ageing neglected areas in, research on, 1, 173- 185 nutritional aspects of, 1, 186-193 pathological basis of, 1, 16-31 performance changes in, 3, 149-169 personal measurement of, 1, 5-7 psychological aspects of, 1, 209-218 physiological approach to, 3, 20-29 physiological changes due to in fish, 5, 181-211 theory of, 5, 1 29-1 3 1 , 269-272, 28 1 , 282, 297 use of inbred strains in research into, 3, 115-130 use of various animals in research into, 1, 177-181 variability in, 3, 32, 33 Alanine, in elastin, 1, 95, 96 Albatross, lifespan of, 5, 100, 103 Albumin, effect on bees, 5, 235 Aldosterone, 4, 59, 60 effect on potassium excretion, 4, 183- 184, 186, 192-194, 196-197 effect on sodium excretion, 4, 183, 185, 192-194, 196 effect on sodium/potassium ratio, 4, 184-185, 186-187, 192-194, 196 effect on urinary output, 4, 182, 192- 194, 196 excretion, in congestive heart failure, 4, 280, 292-293, 298, 300 in pregnancy, 4, 89-90 Algyria, 2, 100, 101 Allantoic fluid, 4, 217, 218 Allometry of lifespan, 5, 125 Alosa sapidissima, fecundity of, 5, 191-192 Alveolar air, contact with pulmonary blood, measurement of, 1, 64-65, 66 Alzheimer's disease, 3, 134, 135-136, 143- 144, 146 Amide N, in elastin, 1, 96 Amiiformes, lifespan of, 5, 222 Amino acids, effect on bees, 5, 238 in aorta, 3, 97 in cells, 3, 47, 49 in elastic tissue, 1, 91, 94-96, 104 in leaves, 2, 203, 206, 207, 212 in plants, 2, 212 in pollen, 5, 237 Amoeba, growth and development of, 3, 41-42 Ammonia, excretion, 4, 209-210, 213-215 in respiratory acidosis, 4, 266 Ammonium salts, in metabolism, 4, 209- 210 Ammotragus lervia, lifespan of, 3, 12, 13, 14, 15 Amphibia, lifespan of, 3, 9 Amyloid, effect of diet, 5, 79 Amyloidosis, in mice, 3, 121, 122, 125, 126; 5,80-81 Anaemia, erythrocytes in, 4, 199 in animals, 3, 33, 34, 36 Cumulative Subject Index 25 Androgens, effect on development of re- productive tract, 2, 10, 11 effect on foetal growth, 2, 173 produced by ovary, 2, 15-16 Androsterone,!, 128, 129, 130, 131, 133, 134 excretion of in old age, 1, 134 Aneurysm, in old people, 1, 22 Angelfish, lifespan of, 5, 221 protein metabolism in, 5, 183-185 Angiomata, 1, 82-84, 86 Anglia cattle, lifespan of, 5, 60 Anguilliformes, lifespan of, 5, 223 Anguilloidei, lifespan and size of, 5, 152 Animal populations, control of, 5, 296-297 Animals, ageing process in, 3, 20 lifespan of {see also under names of animals), 3, 3, 8-11 cause of decrease in, 3, 4—6 methods of obtaining data, 3, 11-17 Anions, excretion of, 4, 209, 210 in infancy, 4, 211-213 Anseres, arteriosclerosis in, 5, 109 Anseriformes, mortality rate of, 5, 101 Antibiotics, effect on lifespan, 3, 36 Antidiuretic hormone, 4, 12, 37, 46, 47, 53, 55, 92, 238-240 in congestive heart failure, 4, 280 Antlers, deer {see Deer, antlers of) Anuria, due to respiratory infection, 4, 268 Aorta, amino acids in, 3, 97 calcification of elastic tissue in, 1, 97-98 calcium in, 3, 97, 98 degenerative changes in, 3, 97 effect of elastase, 3, 98, 102 elastin content of, 1, 92-93, 94, 101 glycolysis in, 1, 76 Aortic disease, in cancer cases, 1,21 Aortic sclerosis, age incidence of, 3, 76 Aortic tissue, diffusion of solutes through, 1, 69, 72-74, 76-78 Apis mellifera, 5, 247 Apocrine sweat glands, ageing of, 2, 188- 201 control of by endocrine glands, 2, 198 distribution of, 2, 188 effect of age on, 2, 194-198, 199 effect of menstruation on, 2, 196-198, 199-200 effect of pregnancy, 2, 197-198, 199-200 glycogen in, 2, 195 in children, 2, 198 in young adults, 2, 190-194 iron content of, 2, 191-193, 195 lipid content of, 2, 191, 195 mitosis in, 2, 196, 199 phylogeny of, 2, 188 pigment in, 2, 190-194, 195 ribonucleic acid in, 2, 194 structure of, 2, 189 Apodiformes, mortality rate of, 5, 101 Apples, ageing in, 2, 206, 208, 210 Aqueous humour, concentration of ions in, 4, 25-26, 28, 29 Arctic char, lifespan of, 5, 147 Arcus senilis, 3, 109 Ardea cinerea, lifespan of, 5, 99 Argentine, lifespan and size of, 5, 1 50 Arginine, in elastin, 1, 96 Armadillo, oogenesis in, 2, 39, 40 Arterial spiders, 1, 81, 82 Arteries, age changes in, 1, 53 cholesterol in, 3, 101 degenerative changes in, 3, 97 pressure waves in, 3, 69 Arteriosclerosis, 1, 56; 3, 74-77, 80, 97-99, 102, 134, 136, 137, 191, 193-194 aetiology of, 5, 107-108 biochemistry of, 1, 28 brain in, 5, 107 calcification of elastic tissue in, 1, 97, 98, 102 cholesterol in, 1, 78-79, 89 ; 5, 1 10-1 1 1, 113 elastase in, 1, 99 elastic tissue in, 1, 88, 89 in animals, 3, 23, 31, 32, 33 in birds, 5, 106-114 in cases of cancer, 1, 21 in centenarians, 1, 16 in dogs, 3, 32, 33 in fish, 5, 225 in mammals, 5, 109 in monkeys, 5, 108 incidence of, in old people, 1, 18, 19, 20, 21, 24, 25, 26 site of, 5, 110 Ash, in rat body, 4, 120-124 relation to body composition, 4, 118, 122 Aspartic acid, in elastin, 1, 91, 94, 95, 96 in plants, 2, 212 Asthma, effect on lung, 1, 65, 66 Barley leaves, protein in, 2, 203-204, 205 respiration in, 2, 206 Basal metabolic rate, variations with age, 1,206 Bass, hfespan of, 5, 220, 221, 223 Batrachoidiformes, lifespan of, 5, 221 Bats, Hfespan of, 5, 97, 103 metabolism and lifespan in, 3, 27 Bees, ageing in, 5, 248 brain cells in, 5, 248 effect of diet, 5, 254 effect of protein on, 5, 254 in tropics, 5, 246 lifespan of, caged, 5, 235-239 factors influencing, 5, 231-246 in free-flying colony, 5, 231-235 physiological condition of, free-flying, 5,231-235 caged, 5, 235-239 yearly life cycle in, 5, 239-240 Beryciformes, lifespan of, 5, 220 Bicarbonate, excretion of in respiratory acidosis, 4, 265-266 in pancreatic juice, 4, 64 in parotid saliva, 4, 64 26 Cumulative Subject Index Birds, ageing in, 3, 9 arteriosclerosis in, 5, 106-114 expectation of life in, 1,4 in research on ageing, 1, 177 lifespan of, 3, 9, 17, 37 in captivity, 5, 103 in Nature, 5, 90-105 metabolism and lifespan in, 3, 28 Bison, lifespan of, 5, 105 Blackfish, lifespan of, 5, 221 Black-pied cattle, lifespan of, 5, 60 Bladder, cancer of in old people, 1, 21 Blatella germanica, in research on ageing, 1, 174 Blennioidei, lifespan and size of, 5, 1 52 Blermius pholis, growth of, 5, 156, 157, 158 lifespan and size of, 5, 152 Blenny (see Blennius pholis) Blood, ventilation of, 1, 62, 63, 66, 68 Blood-brain barrier, 4, 26 Blood pressure, age changes in, 3, 32, 33, 64 changes in, 3, 6 Blood supply to brain, in old people, 1, 51 Blood volume, effect of age, 4, 243 Blue-head, lifespan of, 5, 221 Blue-striped grunt, lifespan of, 5, 221 Body composition, effect of protein and mineral intake on, 4, 116-138 Body weight, relationship to lifespan, 5, 115-139 Bone, magnesium in, 4, 309 molecular changes in, 3, 47, 48 Boselaphus tragocamelus, lifespan of, 3, 12, 13 Bowfin, lifespan of, 5, 222 Brain, blood supply to in old people, 1, 51 changes in, in Alzheimer's disease, 3, 135-136. 143-144, 146 in old age, 1, 34, 35, 36, 49, 51 ; 3, 145, 146, 147 in Pick's disease, 3, 135-136, 143-144, 146 in arteriosclerosis, 5, 107 internal environment of, 5, 133 weight of, relationship with lifespan, 5, 115-139 Brain cells, in bees, 5, 248 Breast cancer, in cattle, 5, 71 in centenarians, 1, 16 in mice, 3, 119, 120 Bronchiectasis, in aged rats, 1, 178-179 Bronchitis, and emphysema, 1, 65, 66 Bull frog, metabolism and lifespan, 3, 26 Bullhead, effect of diet on, 5, 169 lifespan of, 5, 152, 223 size of, 5, 152 survival curves of, 5, 143, 145 Bulls, causes of death in, 5, 65 lifespan of, 5, 64-65 Burbot, lifespan of, 5, 223 Buteo buteo, lifespan of, 5, 99 Butterflies, 3, 25 Butterfly fish, lifespan of, 5, 121 Buzzards, lifespan of, 5, 99 Calcium, absorption of, 1, 110, 111-114, 119-120, 124-125 effect of diet on, 4, 120, 121, 127-128, 132, 138 in aorta, 3, 97, 98 in body of rat, 4, 120, 121, 127-128, 132, 138 in elastin, 1, 97-98, 104, 105 in foetal urine, 4, 218 in placenta, 2, 110 requirements of, 1, 109, 120-122 Calcium balance studies, 1, 114-118 Calcium binding, age changes in, 3, 67 Calcium metabolism, in old age, 1, 109-125 in osteoporosis, 1, 109, 110, 116-118, 121, 122, 123, 125 in rats, 1, 111-114, 124-125 Callionymoidei, lifespan and size of, 5, 1 52 Callionymus lyra, mortality rate of, 5, 146 survival curves of, 5, 143, 145 Cancer, {see Carcinoma) Cancer eye, in cattle, 5, 71 Canis, lifespan of, 3, 12-13, 14, 17; 5, 134 Capelin, lifespan and size of, 5, 150 mortality rate of, 5, 146 Capreolus capreolus, lifespan of, 5, 92-94 Carbohydrate, in pollen, 5, 237, 238 Carbohydrate metabolism, endocrine fac- tors in, 2, 24-25 Carbon dioxide, diffusion of through aortic tissue, 1, 72-74, 77 diffusion through tentorium cerebelli, 1, 74-75 Carbon dioxide production, age changes in, 3, 77, 84-90 Carbonic anhydrase, 4, 218, 223 control of urinary pH, 4, 210 Carcinoma, arteriosclerosis in cases of, 1, 21 in Africans, 3, 104, 106, 113 in animals, 3, 3 1 in fish, 5, 226 in inbred strains, 3, 117, 118-119, 120, 121 incidence of, in old people, 1, 18, 19, 20, 21, 24, 25-26 sites of in old people, 1, 21-22 Cardiac output, 3, 84-86, 87, 191 effect on kidneys, 4, 234, 248, 267 in congestive failure, 4, 272, 276 Cardiac stroke index, age changes in, 3, 87,88 Caribou, lifespan of, 5, 92-94 Castration, 4, 227 effects of, 2, 10 effect on adrenals, 4, 197-198 Cat, hfetime energy expenditure of, 5, 127 lifespan of, 5, 1 34 Catfish, lifespan and size of, 5, 154, 220, 223 Cattle, breeding of, 5, 57, 58 lifespan of, 5, 57-65, 70-71 Caviar lesions, 1, 84, 86 Cavy, lifespan of, 3, 12, 13 Cumulative Subject Index 27 Cell(s), age changes in, 1, 52-53, 54, 55; 2, 250; 3, 39-50; 4, 199-205 changes in, 3, 43-44 cold storage of, 2, 215, 216 eflfect of environment on, 3, 45 effect of nutrition on, 1, 191 effect of temperature on, 5, 281 electrolytes and water in, 4, 15-35 electrolyte transfer in, 4, 19 enzymes in, 3, 47, 48 function of, 3, 41 in testis, 2, 86-99 lifespan of, 2, 216, 227; 3, 5 metabohsm, 2, 216, 217 mutation in, 3, 43, 45 nuclear transplantation in, 3, 46 nucleic acids in, 3, 47, 48 nucleus of, 2, 239-240, 241 osmotic equilibrium of, 4, 1 8 oxygen in metabolism, 2, 243 proliferation of, 3, 41-42, 46 proteins in, 3, 48, 49 replacement of, 3, 4, 5, 6, 8 Cell membrane, permeability of, 4, 16-31 Cell respiration, relationship with protein metabolism, 2, 207-208 Centenarians, 3, 93 accuracy of age of, 5, 13 emphysema in, 1, 16 memory in, 1, 25 mental state of, 1, 25 pathological lesions in, 1, 16-17 Central nervous system, control of vital functions by, 5, 128-129 Centrarchidae, protein metabolism in, 5, 182-183 Cerebral blood supply, in old people, 1, 51 Cerebral hypoxia, in congestive heart failure, 4, 286 Cerebral vascular disease, in cases of cancer, 1, 21 in old people, 1, 22 Cerebrospinal fluid, concentration of ions in, 4, 25-26 Cervus elaphus, lifespan of, 5, 92-94 ChaflSnch, lifespan of, 3, 9 Char, fecundity of, 5, 191-192 lifespan and size of, 5, 1 52 Characin, lifespan of, 5, 224 Charadriiformes, mortality rate of, 5, 101 Cherry angiomata, 1, 82-84, 86 Chickens, arteriosclerosis in, 5, 110, 113 Children, diet for spastic, 1, 183 heights and weights of, 1, 195, 201 malnutrition in, 1, 195 water in body of, 4, 103-106 Chiropodomys gliroides, lifespan of, 5, 96 Chlorides, effect of diet on, 4, 120, 121, 126, 132 excretion of, in congestive heart failure, 4, 276, 277, 278, 284-285 in babies' urine, 4, 21 1 in body of rat, 4, 120, 121, 126, 132 in erythrocytes, 4, 203 Chlorides in foetal urine, 4, 217 in sweat, 4, 64, 74 in tears, 4, 64, 71 loss of during labour, 4, 90 Chlorophyll, in nasturtium, 2, 203 Cholesterol, 1, 126, 127 in aetiology of arteriosclerosis, 5, 110- 111, 113 in arteries, 3, 101 in arteriosclerosis, 1, 78-79, 89; 5, 110- 111, 113 in elastin, 1, 97-98 in placenta, 2, 122 loss of from red cells, 2, 218, 219, 220 Cholesterol metabolism, 3, 191, 193-194 and ageing, 3, 133 fatty acids in, 3, 193-194 Cholinesterases, age activity of, 5, 247 Chorioallantoic membrane, 4, 218 Chorion, membranous, in goat, 2, 151, 153, 154, 156-158 Chrysops, lifespan of, 5, 105 Chub, lifespan of, 5, 150, 221 size of, 5, 1 50 Chymotrypsin, 3, 101 Ciconiformes, mortality rate of, 5, 101 Circulation, age changes in, 3, 77, 84-90 effects ofdeficiency of water, 4, 160, 163 nervous control of in foetus, 2, 29 Cirrhosis of liver, 3, 107, 108, 113 Cisco, lifespan and size of, 5, 150 Citellus pygmaeus, lifespan of, 5, 97 Citric acid, excretion of, 4, 217, 218, 221, 222 Cladocera, 1, 30 Cleft palates, produced by cortisone, 2, 18-19 Clinicopathological tests of ageing, 3, 92- 100 Clupea harengus, growth of, 5, 156, 157, 158 Clupea pallasii, fecundity of, 5, 196-197 Clupea pilchardis, 3, 26 Clupeiformes, lifespan of 5, 220, 222 Clupeoidei, lifespan and growth in, 5, 160- 166 lifespan and size of, 5, 148 Clupeoids, survival curves of, 5, 143, 145 Coalfish, lifespan and size of, 5, 148 Coates' knifefish, lifespan of, 5, 224 Cobalt deficiency, 3, 188-190 Cockroaches, effect of diet on, 5, 253-254 in research on ageing, 1, 174, 175 hfespan of, 5, 253-254 Cod, hfespan and growth of, 5, 160-166 lifespan and size of, 5, 148 Cold, effect on cell, 4, 24 effect on red cells, 2, 224-229, 236-237, 244 Collagen, 3, 96, 97, 98 age changes in, 3, 48, 65, 66, 68, 71 contraction of, 3, 66, 70 in elastin, 1, 91 28 Cumulative Subject Index Collagen in placenta, 2, 146 in scar tissue, 3, 70 Collagen fibres, 1, 88, 106, 107, 108 Colon, cancer of in old people, 1, 21 Columbiformes, mortality rate of, 5, 101 Comparative age studies, 3, 1-38 clinicopathological tests, 3, 92-100 Compensatory adjustment in ageing, 1, 25, 27,41,42,43,46,47 Congestive heart failure, aldosterone ex- cretion in, 4, 280, 292-293, 298, 300 cerebral hypoxia in, 4, 286 humoral factors, 4, 279-288 neural factors, 4, 279-288 renal changes in, 4, 275-279 renal function in, 4, 271-275 salt and water retention in, 4, 288-293 water and electrolyte metabolism in, 4, 271-300 Connective tissue, age changes in, 3, 65 water and electrolytes in, 4, 27 Coregonus clupeaformis, survival curves of, 5, 143, 144 Cormorants, arteriosclerosis in, 5, 109 Coronary disease, in cancer cases, 1, 21, 22 Coronary sclerosis, age incidence of, 3, 74, 75 Coronella laevis, fertility of, 3, 21 Cor pulmonale, renal function in, 4, 266- 267 Corpus luteum, effect on parturition, 2, 79, 85 effect on placenta, 2, 84 effect on pregnancy, 2, 78, 79, 84 effect on uterus, 2, 78-79 growth rate of, 2, 79, 80 of guinea pig, 2, 69-85 in guinea pigs, growth rate of, 2, 73, 75 in pregnancy, 2, 74, 75, 76, 84 ovulation changes in, 2, 71-72, 83, 84 lifespan of, 2, 80-82 Cortexone, effect on potassium excretion, 4, 174-175 effect on sodium excretion, 4, 174, 175, 177 effect on water loss, 4, 174, 175 Corticosterone, administration of in old age, 1, 134-136 Cortin, excretion of, in elderly and schizo- phrenics, 1, 221-222, 223-236 Cortisol, 1, 127, 128 administration of in old age, 1, 134-136 effect on potassium excretion, 4, 188 189, 192-194, 196 effect on sodium excretion, 4, 187-188 189, 192-194, 196 effect on sodium/potassium ratio, 4 190-192, 193-194, 196 effect on urinary output, 4, 187, 188 192-194, 196 Cortisone, effect on adrenal glands, 4, 175-176 effect on diuresis, 4, 1 3 Cortisone effect on hair, 2, 175 effect on placenta, 2, 141 effect on postnatal growth, 2, 168-171 effect on potassium excretion, 4, 171, 172, 176, 178 effect on sodium excretion, 4, 171, 172, 173, 176, 178 effect on water loss, 4, 171, 172 Cottoidei, lifespan and size of, 5, 152 Cottus gobio, survival curves of, 5, 143, 145 Cough, effect on lung, 1, 65 Cows, average age of different breeds, 5, 59 cause of death, 5, 63-64 lifespan of, 5, 58-64, 70 lifetime energy expenditure of, 5, 127 Cowfish, lifespan of, 5, 221 Creatinine excretion, 4, 249 in elderly and schizophrenics, 1, 221- 222, 223-236 Creatinuria, in rats, 5, 83 Cristivomer namaycush, mortality rates of, 5,145 Croaking gourami, lifespan of, 5, 224 Cross linking, 3, 66, 68, 69 in carcinogenesis, 3, 68 Cross-sectional studies, 3, 158-159, 171- 172, 173, 174, 179, 186 Crystalloids of Reinke, 2, 92, 94, 95 Cunner, lifespan of, 5, 221 Cyprinids, metabolism of, 5, 169 Cypriniformes, lifespan of, 5, 220, 222, 224 Cyprinodontiformes, lifespan of, 5, 1 52, 224 size of, 5, 1 52 Cyprinoidei, lifespan and size of, 5, 1 52 Cystine, in elastin, 1, 96 Dab, fecundity in, 5, 192, 200-202 Dace, lifespan of, 5, 223 Daphnia, effect of metabolic rate on life- span, 3, 24 effect of nutrition on lifespan, 3, 24-25 Dasyatis akajei, lifespan and growth of, 5,163 Dasypus, oogenesis in, 2, 56 Deaths, accidental, 5, 17 accuracy of age at, 5, 12-13 age at, 3, 7 Death curves, 5, 6-9, 15, 17, 286-296 of horses, 5, 56 Death rate, anticipated, 5, 4, 5, 7, 12, 15 formula for, 1, 8 laws governing, 5, 2-4 senescent, 5, 4, 5, 6, 7, 10, 11, 12 Decarboxylic amino acids, in elastin, 1, 94, 95, 105 Deer, antlers of, absence of, 2, 186 blood supply of, 2, 177, 185 effect of age on, 2, 183-184 effect of nutrition on, 2, 184 effect of testosterone on, 2, 179-180, 186 enervation of, 2, 178, 185-186 Cumulative Subject Index 29 Deer antlers of {continued) endocrine factors in growth of, 2, 179-181 growth cycle of, 2, 176-187 mechanism of growth, 2, 176-178 radio-phosphorus uptake in, 2, 177- 178 shedding of, 2, 179, 181, 184, 211 cyclic gonadal changes in, 2, 179 growth in, 1, 205-206 lifespan of, 3, 12; 5, 91-94 Dehydration, effect on water intake, 4, 4 in elderly, 1, 207 in labour, 4, 94, 95 Dehydration reaction, 4, 38-39, 47 Dehydroe7J/androsterone, 1, 127, 128 Dementia, genetics of, 3, 137, 147 senile, 1, 35, 36, 44, 45, 46, 214 de Morgan's spots, 1, 82-84, 86 11-Deoxycortisol, 1, 127, 128 Deoxypentose nucleic acid, 3, 47, 48, 49 Diabetes insipidus, causing loss of water, 4, 39, 42-43 Diarrhoea, causing hypematraemia, 4, 58 Dibenamine, effect on kidney, 4, 281, 282 Diet, effect on amyloid, 5, 79 effect on bees, 5, 254 effect on body composition, 4, 117-138 effect on cockroach, 5, 253-254 effect on electrolytes, 4, 116-138 effect on fertiUty, 5, 34 effect on fish, 5, 178, 215-216 effect on flies, 5, 249-255 effect on growth, 4, 116-138; 5, 84-85, 177, 178, 254-255 effect on homeostasis, 4, 143-144 effect on lifespan, 1, 30; 5, 78, 83-85, 87, 88, 169, 177, 178, 249-255, 265-267, 268, 282 effect on mice, 5, 79-80 effect on mosquito, 5, 255 effect on rat, 5, 78, 83-85, 87, 88, 251- 252, 254 effect on sexual maturity, 5, 84-85 effect on trout, 5, 253, 254 effect on wasps, 5, 255 protein in, 5, 252-254 Diffusion coefficients of solutes for mem- branes, 1, 69-79 Digestive disease, incidence of in old people, 1, 20 Dingoes, lifespan of, 3, 12 Dipodomys heermani, lifespan of, 5, 96 Disease, effect on lifespan, 3, 31, 32; 5, 72-89 Diuresis, effect of adrenaline, 4, 9, 14 effect of age, 4, 6-10, 238-240 effect of cortisone, 4, 1 3 effect of hypoxia, 4, 8 effect of pitressin, 4, 7-8, 1 1 effect on adrenal glands, 4, 1 3 in congestive heart failure, 4, 272-273, 275 DOCA (deoxycorticosterone acetate), effect on foetus, 2, 24 Doctor fish, lifespan of, 5, 221 Dog(s), in research on ageing, 1, 182 lifespan of, 3, 12, 13, 14, 17; 5, 134 lifetime energy expenditure of, 5, 127 oogenesis in, 2, 43 Dog snapper, lifespan of, 5, 221 Dolichotis patogona, lifespan of, 3, 12, 13 Domestication, effects of, 3, 15, 37 Dragonet, lifespan and size of, 5, 1 52 mortahty rate of, 5, 146 survival curves of, 5, 143, 145 Drosophila, effect of heredity on lifespan, 1, 239-241 Drosophila melanogaster, effect of tem- perature on lifespan, 3, 24 Drosophila subobscura, lifespan of, 5, 262, 266 rate of ageing in, 5, 269-285 Drum, hfespan of, 5, 221 Ducks, arteriosclerosis in, 5, 109 Dwarf top-minnow, reproduction and senescence in, 5, 189 Education, effect on performance, 3, 1 59- 160 Eel, lifespan of, 5, 152, 223 size of, 5, 1 52 natural death in, 5, 174 Eggs, effect on nutrition, 1, 205 Egg production, variation with age, 3, 21 Elastase, 1, 98-101, 107; 3, 97-98, 101- 103, 136 in arteriosclerosis, 1, 99 in fish, 1, 99 in pancreas, 1, 98, 99, 100 preparation of, 3, 101 production of, 3, 45 tracer studies with, 1, 100-101 Elastic tissue, ageing of, 1, 88-108; 3, 65, 66 amino acids in, 1, 91, 94-96, 104 in atherosclerosis, 1, 88, 89 in skin, 1, 100-102, 105, 106, 107, 108 preparation of, 1, 90-91 Elastin, calcium content of, 1, 97-98, 104, 105 cholesterol content of, 1, 97-98 in aorta, 1, 92-93, 94, 101 in pulmonary artery, 1, 93-94 polysaccharides in, 1, 104, 107 preparation of, 1, 90-91, 104 Elastosis, 3, 96, 97 senile, 1, 101-103, 106-108 Electric catfish, lifespan of, 5, 224 Electric eel, lifespan of, 5, 224 Electrolytes (see also under Sodium, potassium, etc.) cellular aspects of, 4, 15-35 changes in due to age, 4, 241, 311-312 deprivation of, 4, 144 effect of diet, 4, 116-138 30 Cumulative Subject Index Electrolytes effect of hormones, 4, 313-314 effect of hypercapnia, 4, 265 effect on mineral content of body, 4, 125 effect on protein body content, 4, 125 excretion of, in elderly schizophrenics, 1, 223-237 response to adrenal steroids, 4, 180- 194, 196-198 glandular secretion of, 4, 62-77 hormonal aspects of, 4, 78-98 in congestive heart failure, 4, 271-300 in muscle, 4, 164 in parenteral fluid therapy, 4, 146-148 in pregnancy, 4, 88-90 metabolism, in infancy, 4, 154-164 regulation of by kidney, 4, 229-249 total exchangeable in body, 4, 108 Elephants, lifespan of, 1, 242 Emotion, changes in with age, 3, 170-182, 184 Emphysema, aetiology of, 1, 65-66 antispasmodics in, 1, 67 bronchitis and, 1, 65, 66 diffusing capacity in, 1, 64-65, 66 in centenarians, 1, 16 pulmonary ventilation in, 1, 62, 63 sex distribution in, 1, 68 vital capacity in, 1, 58, 59, 60 Endocrine organs, age changes in, 2, 161- 162; 3, 128-129 control of apocrine sweat glands by, 2, 198 Endometrium, pregnancy changes in, 2, 115 Environment, effects of, 3, 171 effect on ageing, 3, 35, 37, 45, 48, 145, 148, 171, 183-184 effect of lifespan, 5, 167-168, 229 effect on onset of disease, 5, 86-87 Enzymes, in cells, 3, 47, 48 in rat placenta, 2, 107-108 in red cells, 2, 235, 241 Enzyme production, 3, 44, 45 Eosinophils, levels of, variations with age, 1,223,231,232 Ephedrine, in emphysema, 1, 67 Epinephalus guttatus, protein metabolism in, 5, 183-184 Erythema, palmar, 1, 82 Erythrocytes, electrolytes and water in, 4, 17-21, 199-208 in foetus, 4, 204, 205, 206 in sheep, 4, 200-203, 204, 206 Ewes, effect of nutrition on, 2, 184 Experience, effect of on performance, 3, 156-158 Extracellular fluid, equilibrium with plasma, 4, 15-16 volume of, changes due to age, 4, 244 Eyes, changes with age, 1, 214; 3, 109, 195 water content of, 4, 28, 29 Eyesight, 3, 182-183 Falconiformes, mortality rate of, 5, 101 Fats, in body, 4, 113, 114, 115, 129, 132 in pollen, 5, 237 in rat body, 4, 119 storage in bees, 5, 237, 238, 239 Fat body in bees, 5, 233, 235, 237, 238, 239, 244 Fatigue, in old people, 5, 150, 151 Fatty acids, in cholesterol metabolism, 3, 193-194 Fecundity, variation of in fish, 5, 191- 192 Female, longevity in, 3, 30 Fertility, effect of age on, 3, 20-22 effect of parental age on, 5, 56 Fibrin, in placenta, 2, 110 Fibrinoid, in placenta, 2, 110 Finch, lifespan of, 5, 99, 103 Fish {see also under common names) changes in due to ageing, 5, 218 effect of diet on, 5, 1 67 effect of fat diet on, 5, 215-216 effect of growth and size on lifespan, 5, 147-159 effect of toxic substances on, 5, 177 egg counting in, 5, 192-193 elastase in, 1, 99 fecundity in relation to age in, 5, 186- 206 fertility of, 3, 22 growth of, utilization of protein in, 5, 182-186 growth and senescence in, 5, 217 growth rate and lifespan, 5, 227, 229 in nutritional research, 1, 174, 175 infectious disease in, 5, 213-214, 228 lifespan of, 1, 28, 29; 3, 9, 15-16 characteristics of long, 5, 217-218 in captivity, 5, 212-230 in different species, 5, 219-224 in Nature, 5, 142-180 lipid metabolism in, 5, 227 metabolic disease in, 5, 216 metabolism of, 5, 169-170 metabolism and lifespan in, 3, 25-26 natural death and reproduction in, 5, 170-174 natural mortality of, 5, 142-147 neoplasia in, 5, 216 nutrition in, 5, 215 ovaries of, 5, 193 parasites in, 5, 213-216 physiological changes due to ageing, 5, 181-211 protein utilization in, 5, 210 relationship of age, mortality and growth in, 5, 160-166 survival curves of, 5, 143-147 variation in fecundity of, 5, 191-192 Flamingoes, effect of diet on, 5, 112 Flexure lines, 1, 12-13 Flies {see also Houseflies) lifespan of, 5, 105 Flounder, lifespan and size of, 5, 150 Cumulative Subject Index 31 Fluids, metabolic disturbances, reasons for, 4, 154-155 Foetal gigantism, 2, 112, 162, 164, 166, 167, 171 Foetus, acid-base balance in, 4, 217-219 blood vessels of, radio-potassium in, 2, 160 decapitation of, effects of, 2, 12, 20-30 effect of cortisone on, 2, 167-175 effect of DOCA on, 2, 24 effect of gonadotrophic hormone on, 2, 28 effect of growth hormone on, 2, 161- 167, 171, 173-175 effect of testosterone on, 2, 173 endocrine events in, 2, 18-30 haemoglobin in, 4, 203 modification of growth of, by growth hormone and cortisone, 2, 161-175 urine in, 4, 217 weight of, 2, 124, 125, 127 effect of growth hormone, 2, 164-165 Follicles, atresia of, 2, 59-65 changes of in guinea pigs, 2, 69 effect of irradiation on, 2, 67, 68 effect of oestrogens on, 2, 60-65 inhibition of in guinea pigs, 2, 78 redundant, history and fate of, 2, 59-68 ruptured, in guinea pigs, 2, 75-78 Food, conversion of in fish, 5, 186 Fowls, fertility of, 3, 20, 21 Freezing, resistance of tissue to, 1, 163- 165 Fructose, in placenta, 2, 120-122 production in placenta, 2, 120, 121 Fruit, ageing in, 2, 206 Function, studies in human, 3, 73-9 1 Functional residual capacity, variations with age, 1, 60 Gadiformes, growth of, 5, 160-166 lifespan of, 5, 148, 160-166, 223 metabolism of, 5, 169 size of, 5, 148 Galago, oogenesis in, 2, 56 Galli, arteriosclerosis in, 5, 109 Galliformes, mortality rate of, 5, 101 Gambusia a. affinis, reproduction and senescence in, 5, 187-189 Gannets, arteriosclerosis in, 5, 109 Gar, hfespan of, 5, 222 Gasterosteiformes, lifespan of, 5, 1 52, 223 Gasterosteus aculeatus, 3, 26 effect of environment of, 5, 168 Geese, arteriosclerosis in, 5, 109 Genes, relation with macromolecules, 3, 47 Genetical aspects of ageing, 1, 238-241 ; 3, 131, 132-133, 144; 5, 137 in Drosophila, 5, 278-280, 283 Germinal epithelium, proliferative powers of, 2, 36, 37, 38 Gibbs-Donnan equilibrium, 4, 15-18, 27, 28,30 Gigantism of foetus, 2, 112, 162, 164, 166, 167, 171 Gingival epithelium, effect of brushing on, 3, 55-57, 71-72 Glomerular filtration rate, changes in with age,4, 231,238, 246 effects of pyrogen, 4, 237 in congestive heart failure, 4, 275, 277 Glomerulonephritis, in man, 5, 77 in rats, 5, 74-76 Glucose, administration of in elderly schizophrenics, 1, 220-236, 237 diffusion of through aortic tissue, 1, 72-74 diffusion through tentorium cerebelli, 1, 74-75 in foetal liver, 2, 136, 145 in placenta, 2, 135-137, 144, 145 production in placenta, 2, 120, 121 Glutamic acid, in elastin, 1, 91, 94, 95, 96 in plants, 2, 212 Glutamine, in senescent leaves, 2, 203, 212 Glycerophosphatase, in placenta, 2, 107, 108 Glycine, in elastin, 1, 91, 94, 95, 96 Glycogen, in apocrine sweat glands, 2, 195 in placenta, 2, 27-28, 107, 108, 122, 123, 135-137, 144, 145, 159 Glycogen storage, effect of ACTH on, 2, 24-25, 29 effect of hormones on, 2, 24 inbees, 5, 237 238 239 in liver, age effects of, 2, 23-25, 27, 29 Glycolysis, in dog aorta, 1, 76 Goats, lifespan of, 3, 13, 14, 36 placenta of, 2, 120, 127, 154-158, 159 placentome of, development of, 2, 155- 158, 159 radio-potassium uptake during preg- nancy, 2, 148-160 Golden stiver, lifespan of, 5, 223 Goldfish, lifespan of, 3, 9 ; 5, 223 metabolism of, 5, 170 Golgi complex, 2, 88, 89 Gompertz' Law, 5, 2, 72, 302 Gompertz-Makeham equations, 5,117 Gonadotrophic hormone, effect on foetus, 2,28 effect on reproductive tract, 2, 1 5 produced by pituitary gland, 2, 21-22, 28 Gonadotrophin, effect on ovary, 1, 150, 151, 152, 153;2, 31,35, 55 effect on sterilized ovary, 2, 49, 51, 57 ovulation induced by, 2, 75, 79, 82 Grass, protein in, 2, 203 respiration in, 2, 206 Grayling, effect of environment of, 5, 168 metabolism and lifespan in, 3, 26 Grouper, lifespan of, 5, 220 Growth, 2, 251; 5, 296, 298 body water changes due to, 4, 103-106 effect of nutrition on, 1, 186, 188 32 Cumulative Subject Index Growth effect of diet, 4, 116-138; 5, 84-85, 177, 178,254-255 effect on electrolytes, 4, 160 effect on lifespan, 1, 29; 5, 147-159, 160-166 in mice, 4, 136 in rats, 4, 136 relationship with lifespan, 5, 160-166 utilization of protein in, 5, 182-186 Growth hormone, action of, 1, 29; 2, 161- 167, 171, 173-175 effect on lifespan, 1, 29 Growth rate, 1, 6 Grunion, lifespan and size of, 5, 154 Guinea pig, corpus luteum of, 2, 69-85 fertility of, 3, 20 follicular atresia in, 2, 59, 60 lifetime energy expenditure of, 5, 127 oocytes in, 2, 36, 39, 41, 42, 54-55 placenta of, 2, 110 uterus of, 1, 167 Gums, brushing of, 3, 51, 55-57, 71-72 keratinization in, 3, 72 Guppies, effect of diet on, 5, 1 69 fecundity of, 5, 209 growth of, 5, 229, 230 lifespan of, 3, 15-16, 30-31 regeneration in, 5, 208-210 reproduction in, 5, 171 Gynaecomastia, in malnutrition, 3, 108- 109 Habrobracon juglandis, 5, 255 Haddock, effect of age on fecundity, 5, 198-200 fertility of, 3, 22 lifespan and size of, 5, 148 relationship of fecundity and body weight, 5, 200 sex organs in, 5, 190 Haemoglobin, changes in due to age, 4, 203, 206, 207 foetal, 4, 203, 206, 207 in sheep, 4, 202-203 loss of from red cells, 2, 218, 219, 220 Hair, effect of cortisone on, 2, 175 Hake, lifespan and size of, 5, 148 Halibut, lifespan and size of, 5, 148, 150 Hamsters, hypothermia in, 1, 168; 2, 248 in research on ageing, 1, 177, 181 lifespan of, 1, 177 Health, effect on sampling, 3, 183, 184 Hearing, as measurement of ageing, 1, 6 Heart, effect of potassium on, 4, 95-96 output of, 3, 191 changes with age, 3, 84-87 stroke index of, 3, 87, 88 Heart disease, in Africans, 3, 104, 1 1 1-1 12, 113 in man and rat, 5, 74-77 Heart failure, congestive {see Congestive heart failure) in Africans, 3, 1 1 1 Heat regulation, decline with age, 3, 63-64 Hereditary haemorrhagic telangiectasis, 1, 85, 105 Heredity, effect on lifespan, 1, 238-241 role of in ageing, 1, 26, 27, 48 Herons, lifespan of, 5, 99 Herring, fecundity of, 5, 196-197 growth of, 5, 156, 157, 158, 160-166 lifespan, 5, 148, 156, 157, 158, 160-166 relation of gonad growth to body weight, 5, 193 relation of size and maturity, 5, 172 size of, 5, 148 survival curves of, 5, 143, 145 Hesperoleucus venustus, effect of environ- ment on, 5, 168 Meter andria formosa, reproduction and age in, 5, 189 Hibernation, effect on lifespan, 2, 249 ; 5, 103-104 Highland cattle, lifespan of, 5, 59, 61, 62 Hippocampus hudsonius, lifespan of, 5, 147 Hippoglossoides platessoides, fecundity in, 5, 192 Hippoglossus spp., lifespan of, 5, 147 Histidine, in elastin, 1, 96 Holocanthus bermudensis, protein meta- bolism in, 5, 183-185 Homeostasis, disturbances of in infants, 4, 154-157 effect of hormones on, 4, 165-179 of water and electrolytes, effect of age, 4, 139-153 Hormonal environment, effect on ovary, 1, 142-146 Hormones, effect on development of re- productive tract, 2, 3-18 effect on electrolytes, 4, 313-314 effect on glycogen storage, 2, 24 effect on homeostasis, 4, 165-179 ovarian, 2, 48-52 sensitivity of ovary to, age changes in, 1, 142-146 Horses, breeding of, 5, 55 causes of death in, 5, 68 coat colour, and longevity, 5, 49-50 death curves of, 5, 56 lifespan of, under various climatic con- ditions, 5, 65-69 lifespan of English thoroughbred, 5, 35-56 lifetime energy expenditure of, 5, 127 survival curves of, 5, 37-42 Housefly, ageing in, 5, 247-262 effect of diet on, 5, 249-255 lifespan of, 5, 249-262, 287-288 effect of paternal age, 5, 259-262 sex differences in, 5, 255-259 sex ratio in, 5, 249 Humming-birds, 3, 28 lifespan of, 5, 105 Hyalinization, in placenta, 2, 1 10 Hydrogen ion gradients, 4, 34 Cumulative Subject Index 33 17-Hydroxycorticosteroids, effect on water and electrolytes, 4, 79 1 7-Hydrox>T)rogesterone, 1, 127, 128 Hydroxyproline, in skin elastic tissue, 1, 88 Hypercapnia, renal effects of, 4, 265-266 Hypercholesterolaemia, 3, 136 Hypernatraemia, and cerebral disturb- ances, 4, 36-44 due to diarrhoea, 4, 58 due to water deficiency, 4, 38-44 Hypertension, renal aspects of, 4, 258 Hypertonic saline, effect on hyponatrae- mia, 4, 50 Hyponatraemia, 4, 44-55, 95 and cerebral disturbances, 4, 36-37 and steroid output, 4, 60 Hypophysectomy, effect on foetus, 2, 166 effect on ovary, 2, 60-65 Hypothalamus, effect on thirst, 4, 37 Hypothermia, in hamsters, 1, 168; 2, 248 Hypoxia, effect on diuresis, 4, 8 Ide, lifespan of, 5, 223 Inbred strains, 3, 6-8 and parabiosis, 3, 116, 124, 127-129, 130 carcinoma in, 3, 117, 118-119, 120, 121 cause of death in, 3, 120, 121 development of, 3, 115-119, 127 in experimental gerontology, 3, 115- 130 lifespan of, 3, 13, 119-120, 121 ovarian function in, 3, 119-120, 126, 127 survival of, 3,13 transplantation in, 3, 1 16, 120, 122-123, 126, 129 Index of cephalization, 5, 123, 135 Industrial performance, studies in, 3, 151, 153, 164-167 Infants, electrolyte metabohsm in, 4, 78- 79, 154-164 water metabolism in, 4, 78-79, 154-164 water retention in, 4, 96-98 Lifection, resistance to, as measure of ageing, 1, 6 Infectious disease, in fish, 5, 213-214 Insects (see also under names of species) ageing in, 5, 247-268 in nutritional research, 1, 174, 175 lifespan of, 5, 105 overwintering in, 5, 232-234, 240 Insemination, artificial, 1, 171 Insulin, effect on placenta, 2, 141 Insulin sensitivity, 1, 237 Intellectual behaviour, 3, 175-177, 182 Intelligence, changes in with age, 1, 36-39 ; 3, 170-187 Interstitial cells, in testis, 2, 91-96 Involutional melancholia, 1, 44-45 Involutional psychosis, 3, 134 Iodide, diffusion of through aortic tissue, 1, 72-74 12 Iron, content of apocrine sweat glands, 2, 191-193, 195 in heart muscle, 3, 111, 112 metabolism of, 3, 109-110, 112 Irradiation, effect on lifespan, 5, 19-20, 138-141,282,290,292 effect on onset of disease, 5, 87, 88 effect on ovary, 3, 127 effect on thyroid, 3, 55 Isoleucine, in elastin, 1, 91, 94, 96 Jack, lifespan of, 5, 220 Jacob-Creutzfeldt's disease, 3, 134 Kwashiorkor, 1, 190-191 Kar> oplasm, 2, 90, 91 Keratinization, in gums, 3, 72 Ketosteroids, in testis, 2, 98, 99 in urine, relationship with adrenal steroids, 1, 128-130, 133-139 17-Ketosteroids, changes in excretion due to age, 4, 91 daily excretion of, 1, 131-133 excretion of, age changes in, 1, 126-140 in elderly and schizophrenics, 1, 221- 222, 223-236 in elderly subjects treated with steroids, 1, 134-136 Kidney, age changes in, 4, 234-235 amyloidosis in, 3, 122, 125, 126 blood content of, 3, 52, 58 blood flow in, 3, 191-192; 4, 248 age changes in, 3, 88 changes in due to age, 4, 227-228, 229- 249, 253-254 concentrating ability of, changes in due to age, 4, 11-12,243-244 diseases of, effects of age, 4, 250-263 effect of Dibenamine, 4, 281, 282 effect of obesity, 4, 254, 255, 260 effects of potassium deficiency on, 4, 262-263 effects of pyrogen, 4, 235-238 effect of water deficiency of, 4, 43 en2ymes in, 4, 245 function, 4, 229 in respiratory failure, 4, 264-270 variations with age, 1, 220 glomerular filtration rate, changes due toage, 4, 231,248 in congestive heart failure, 4, 275, 277 growthof, 4, 251-252 hormonal damage to, 4, 256-258, 262 in congestive heart failure, 4, 272, 273, 276, 282, 283, 284, 287 in cor pulmonale, 4, 266-267 lesions of, causing loss of water, 4, 39 overloading producing senile changes, 4, 254-255, 260 plasma flow in, changes in due to age, 4, 229-231, 235 effect of pyrogen, 4, 237-238 regeneration of, 4, 252-253 34 Cumulative Subject Index Kidney role of in water and electrolyte regula- tion, 4, 229-249 tubular excretion, changes due to age, 4, 233, 247 Kidney disease, in man and rat, 5, 74-77 Killifish, lifespan of, 5, 220 Labidesthes, growth of, 5, 156, 157, 158 Labidesthes sicculus, mortality rate of, 5, 146 Labour, dehydration during, 4, 94, 95 Lactate, diffusion of through aortic tissue, 1, 72-74 Lactic acid, in placenta, 2, 137, 141 Lagonistica senegala, lifespan of, 5, 99 Lamniformes, lifespan of, 5, 220 Larynx, cancer of in old people, 1, 21 Laws of mortality, 5, 2-5, 302-311 Learning, adaptation in, 3, 62 in old people, 1, 41, 215 variation with age, 1, 215 Learning ability, in animals, 3, 62, 64-65 in human beings, 3, 175-180 Leaves, amino acids in senescent, 2, 203, 206, 207, 212 catabolism of protein in, 2, 202-205, 206, 207-208, 212, 214 effect of light on ageing of, 2, 211 metaboUsm in senescent, 2, 202-214 protein catabolism in, 2, 202-205, 206, 207-208, 212, 214 respiration in senescent, 2, 205-209, 212,213,214 Lebestes reticulatiis, eflFect of diet on, 5, 169 fecundity of, 5, 209 growth of, 5, 229, 230 lifespan of, 3, 15-16, 30-31 regeneration in, 5, 208-210 reproduction in, 5, 171 Lemur, oogenesis in, 2, 39-40 Lepidosteiformes, lifespan of, 5, 223, 224 Leucichthys kiyi, survival curves of, 5, 143, 145 Leucichthys sardinella, metabolism of, 5, 169 mortality rates of, 5, 145 Leucine, in elastin, 1, 91, 94, 96 Leuresthes tenuis, survival curves of, 5, 143 Leydig cells, in testis, 2, 91, 92, 93, 94, 95, 97, 98, 99 Lipid(s), content of sweat glands, 2, 191, 195 in aortic membrane, 1, 74, 78 in Leydig cells, 2, 93 in placenta, 2, 122 loss of from red cells, 2, 218, 219, 220, 223, 242 Lifespan, allometry of, 5, 125 and delayed maturation, 1, 27-30 and hibernation, 2, 249; 5, 103-104 and nutrition, 1, 186-193, 194-199 Lifespan and parental age, 5, 21-34, 43-47, 53 effect of antibiotics on, 3, 36 effect of basal metabolic rate on, 3, 31 effect of brain and body weight, 5, 115- 139 effect of diet, 1, 30; 5, 169, 249-255, 282 effect of disease on, 3, 31 ; 5, 72-89 effect of early maturation, 1, 201-208 effect of egg laying on in Drosophila, 5, 275-278 effect of environment, 5, 167-168, 229 effect of growth and size, 5, 147-159, 160-166 effect of growth hormone, 1, 29 effect of hibernation, 2, 249; 5, 103-104 effect of late maturation, 1, 196-197 effect of metabohsm on, 5, 103-104, 124-126, 129, 136-137, 169-170 effect of mitotic inhibitors, 5, 136 effect of nutrition on, 1, 186-193, 194- 199;3, 24, 25, 28, 34, 35, 36 effect of parental age, 5, 21-34, 43-47, 53, 259-262 effect of radiation on, 5, 19-20, 138- 141, 282, 290, 292 effect of reproduction, 5, 179, 275-279, 284-285 effect of temperature, 3, 24-25 effect of temperature on in Drosophila, 5, 271-279, 283, 284 effect of thyroid gland on, 5, 137 effect of weight on, 1, 203 genetic aspects of, 5, 33, 55, 137 hereditary aspects of, 1, 238-241 increase of, 1, 179, 217 in nineteenth century, 5, 23, 32 mathematical basis of, 5, 286-296 measurement of, 5, 133-134, 138-141 methods of study, 5, 21-26 of Africans, 3, 105 of albatross, 5, 100, 103 of Ammotragus lervia, 3, 12, 13, 14, 15 of Amphibia, 3, 9 of animals, 3, 3, 8-11 cause of decrease in, 3, 4—6 methods of obtaining data, 3, 11-17 of Arabian horses, 5, 42 of Ardea cinerea, 5, 99 of bats, 3, 27; 5, 103 of bees, factors influencing, 5, 23 1-246 ofbirds, 1,4;3,9, 17, 28 in captivity, 5, 103 in Nature, 5, 90-105 of bison, 5, 105 of Boselaphus tragocamelus, 3, 12-13 of bull frog, 3, 26 ofbuUs, 5, 64-65 of Buteo buteo, 5, 99 of buzzards, 5, 99 of Canis, 3, 12-13, 14, 17; 5, 134 of Capreolus capreolus, 5, 92-94 of caribou, 5, 92-94 ofcat, 5, 134 Cumulative Subject Index 35 Lifespan ofcattle, 5, 57-65, 70-71 of cavy, 3, 12, 13 ofceIls,2, 216, 227;3, 5 of Cervus elaphus, 5, 92-94 of chaffinch, 3, 9 of Chiropodomys gliroides, 5, 96 of Citellus pygmaeus, 5, 97 of cockroaches, 5, 253-254 of deer, 3, 12; 5,91-94 of dingoes, 3, 12 of dogs, 3, 12, 13, 14, 17; 5, 134 of Dipodomys heermani, 5, 96 of Dolichotis patagona, 3, 12, 13 of Drosophila subobscura, 5, 262, 266 of elephants, 1, 244 of EngHsh thoroughbred horses, 5, 35- 56 offinch, 5, 99, 103 offish, 1,28, 29; 3, 9, 15-16 characteristics of long, 5, 217-218 in captivity, 5, 212-230 in Nature, 5, 142-180 of different species, 5, 219-224 of flies, 5, 105 ofgoats, 3, 13, 14, 36 ofgoldfish,3, 9;5, 223 of guppies, 3, 15-16, 30-31 of Hafling mares, 5, 43 of hamsters, 1, 177 of herons, 5, 99 of Hokkaido ponies, 5, 43 of horses, 5, 35-56, 65-69 of housefly, 5, 249-262, 287-288 sex difi'erences in, 5, 255-259 of human beings, actuarial aspects, 5, 2-20 of humming-birds, 5, 105 of inbred strains, 3, 13, 119-120, 121 of insects, 5, 105 of Lagonostica senegala, 5, 99 of Lebistes reticulatus, 3, 15-16, 30-31 of Lipitsa horses, 5, 43 of lizards, 3, 26, 27 of lupus, 3, 12 of mammals, 3, 9, 10 of mammals in Nature, 5, 90-105 of man and woman compared, 5, 10-1 1, 16 of Megadyptes antipodes, 5, 100 of mice, 1, 177; 3, 9; 5, 95-96, 287 of Microtus, 3, 10 of mosquitoes, 5, 105 of Muntiacus muntjac, 3, 12 of Mus, 3, 9 of Myotis mystacinus, 5, 97 of nilghaie, 3, 12 of Odocoileus hemionus, 5, 92-94 ofowls, 5, 99 of Ovis dalli, 5, 92-94 of Ovis musimon, 3, 13, 14 of Paramecium, 3, 16 of parent and progeny correlated, 5, 47-49 Lifespan of Parus major, 5, 100 of Passer domesticus, 5, 99 of Passerines, 5, 99-100 of Perognathus, 3, 9, 10 of Peromyscus, 3, 9, 10 of Peromyscus leucopus, 5, 95-96 of rabbits, 1, 181 ; 3, 11 ; 5, 97 of Rangifer articus, 5, 92-94 of rats, 1, 177, 178, 179, 180; 5, 95-96, 251-252 eff'ect of diet, 5, 78 eff"ect of disease, 5, 72-89 of red cells, 2, 217, 233, 234, 238, 239, 241,245 of sheep, 3, 12, 13, 14, 15; 5, 91-94 of shrews, 5, 97 of Sorex araneus, 5, 97 of souslik, 5, 97 of sparrows, 5, 99 of Sterna hirundo, 5, 97-99 of Strix aluco, 5, 99 of swifts, 5, 100, 103, 104 of terns, 5, 97-99, 104 of tits, 5, 100 of Tokophyra, 1, 187, 191 of trout, 1,29; 5,222, 265 of twins, 3, 140, 141 of ungulates, 5, 91-95 of vertebrates, 3, 9, 10, 11-17 of wolfhound, 3, 12, 13 of wolves, 3, 12, 14 relationship to index of cephalization, 5, 123-125 relationship with growth, 5, 160-166 sex difi'erences in, 3, 30 sexual maturity and, 1, 29, 30 Life-tables, 1, 9-10; 5, 9-11 limitations of, 5, 11-12, 18 Light, effect on ageing of leaves, 2, 211 efi'ect on deer antler growth, 2, 181- 182 Liver, carcinoma of, 3, 106, 107, 113 foetal, glucose production in, 2, 136, 145 relationship with placenta, 2, 108, 109 glycogen storage in, age efi"ects of, 2, 23-25, 27, 29 in fish, 5, 227 mitochondria in, 2, 101 Liver disease, in Africans, 3, 107-111, 188-190 infish, 5, 215-216 Lizards, metabolism and lifespan in, 3, 26, 27 parasites in, 3, 33 Longevity (see also Lifespan) onset of disease and, 5, 72-89 Longitudinal studies, 3, 163-164, 172, 173, 174 Look-down, lifespan of, 5, 221 Lophius piscatorius, elastase in, 1, 99 Lovettia seali, mortality rate of, 5, 146 36 Cumulative Subject Index Lung, age changes in, 1, 65 blood flow in, 1, 67 cancer of in old people, 1, 21 diseases of, 3, 77-79 in rats, 5, 82 effects of age on, 4, 264 elastic resistance of, 1, 58, 60, 61, 63, 65 function of, 3, 77, 78 mixing of air in, 1, 68 volume of, variations with age, 1, 58, 60, 65, 66, 67, 68 Lungfish, lifespan of, 5, 224 Lupus, lifespan of, 3, 12 Lytnantria dispar, 3, 24 Lymphocytes, levels of, variations with age, 1,224,231,232 Lysine, in elastin, 1, 96 in plants, 2, 212 Mackerel, lifespan and size of, 5, 154 Magnesium, deficiency of, 4, 301-310 signs of, 4, 304, 307, 309-310 in bone, 4, 309 in plasma, 4, 99-100 in rat body, 4, 120, 121 Makeham's Law, 5, 2, 302 Malayan flying barb, lifespan of, 5, 224 MaUotus villosus, mortality rate of, 5, 146 Malnutrition, eft'ect of, 1, 205 effect on body fluids, 4, 156-157 in Africans, 3, 188-190 water metabolism in, 4, 156-157 Mammals, arteriosclerosis in, 5, 109 lifespan of, 3, 9, 10 in Nature, 5, 90-105 life-tables of, 5, 117-118 relationship of brain and body weight to Hfespan, 5, 115-139 Man, disease in, effect on lifespan, 5, 72-89 lifetime energy expenditure of, 5, 127 Mandibular glands, in bees, 5, 233 Manual work, performed by elderly people, 1, 216 Maternal age, effect on lifespan, 5, 23-27, 31,32 Maturation, definition of, 1, 202 delay of and lifespan, 1, 27-30 effect of nutrition on, 1, 189-190 effect on lifespan, 1, 27-30, 201-208 rapid, as cause of ageing, 1, 194—199 Mealworm, effect of paternal age on life- span, 5, 262 Megadyptes antipodes, lifespan of, 5, 100 Melanogrammus aeglefinis, fecundity of, 5, 198-200 sex organs in, 5, 190 Membranes, diffusion coefficients for solutes for, 1, 69-79 permeabilitv of, effect of enzyme in- hibitors, i, 78 Memory, 3, 187 age changes in, 3, 65 in centenarians, 1, 25 Memory in old people, 1, 40-41 in rats, 1, 215 Menarche, age of, 1, 202 Menopause, ovarian changes in, 1, 144- 146 Menstrual cycle, acid-base balance in, 4, 93 effect on water and electrolytes, 4, 81- 88 sodium/potassium ratios during, 4, 83-88 Menstruation, effect on apocrine sweat glands, 2, 196-198, 199-200 restoration of after cessation, 1, 144- 145 Mental ability, adaptation in, 3, 62 decline of with age, 3, 64—65 Mental aspects of ageing, 1, 32-52 Mental disease, genetics of, 3, 133-134, 137, 144 in Africans, 3, 145 in old people, 1, 45-46 Mental excitement, due to magnesium deficiency, 4, 304, 307, 309-310 Mental testing, 3, 176, 181, 182, 183-184, 186 Mercury poisoning, excretion of sweat in, 4,99 Mer tones libycus, 3, 51, 57 Metabolic disease, in fish, 5, 216 Metabolic disturbances, in infants, 4, 1 54- 164 Metabolism, comparison between infant and adult, 4, 157-159 effect of undernutrition, 1, 189 effect on lifespan, 5, 103-104, 124-126, 129, 136-137, 169-170 in senescent leaves, 2, 202-214 Metabolic rate, effect on lifespan, 3, 31 relationship with ageing, 3, 23-28 Methionine, in elastin, 1, 96 Mice, amyloidosis in, 5, 80-81 effect of diet on, 5, 79-80 eflfect of X-irradiation on ovary of, 2, 49-50, 57 growth in, 1, 204 lifespan of, 1, 177; 3, 9; 5, 95-96, 287 ovarian changes in, 1, 146 white-footed, lifespan of, 5, 95-96 Microtus, lifespan of, 3, 10 Miller's thumb, lifespan of, 5, 223 Minerals, in pollen, 5, 237 intake of, effect on body composition, 4, 116-138 Minnow, lifespan and size of, 5, 152 Mitochondria, appearance of, 2, 100 division of, 2, 103, 104 in adrenal cortex, 2, 101-102, 103 in adrenal glands, 2, 101, 102 in different physiological states, 2, 100- 104 in Leydig cells, 2, 93 Cumulative Subject Index 37 Mitochondria in liver, 2, 101 in pancreas, 2, 101 in placenta, 2, 107 in yolk-sac, 2, 101 Mitosis, 2, 250, 251 in apocrine sweat glands, 2, 196, 199 Mitotic inhibitors, 5, 136 Moina macrocopa, 3, 24 Molecular changes in ageing, 5, 129-131 Mollusca, 3, 25 Monkeys, arteriosclerosis in, 5, 108 in research on ageing, 1, 182 menstruation in, 1, 145-146 oogenesis in, 2, 44, 55 ovulation in, 2, 67 Moonfish, lifespan of, 5, 221 Mortality, relation with age and growth, 5, 160-166 theory of, 5, 127-129 Mortality rates, 5, 286-296 laws governing, 5, 2-5 mathematical models for, 5, 302-311 sex differences in, 5, 81 Mosquito, effect of diet on, 5, 255 lifespan of, 5, 105 Mosquito fish, mortality rates of, 5, 167 reproduction and senescence in, 5, 187- 189 Mothers, age of, eflfect on lifespan, 5, 23- 27, 31, 32 Motivation, 3, 154-156, 185, 186, 187 Mugiloidei, lifespan and size of, 5, 1 54 MiiUerian ducts, normal development of, 2,3,13-14 Muntiacus muntjac, lifespan of, 3, 12 Mus, lifespan of, 3, 9 Musca domestica {see Housefly) Muscle, analysis of, in sodium deficiency, 4, 49-50 composition of, 4, 23 electrolytes in, 4, 21-22, 164, 224-225 potassium in, 4, 289-291 power of, 1, 137 as measure of ageing, 1, 6 water in, 4, 21-22, 113, 163-164 Muscular degeneration, 5, 77 in rats, 5, 74-76 Muskallunge, lifespan of, 5, 222 Mutation, 3, 43-44, 45 Myleran, effect on oogenesis, 2, 43 Myocardial degeneration, 5, 77 in rats, 5, 74-76 Myotis mystacinus, lifespan of, 5, 97 Naevi, spider, 1, 81-82 Nasturtium, proteins in, 2, 203 Matrix natrix, fertility of, 3, 2, 22 Naturalists, longevity in, 1, 52 Neothinmus macropterus, fecundity of, 5, 193 Nephrectomy, effects of, 4, 252, 255, 257 Nephrosis, in man, 5, 77 in rats, 5, 74-76, 83 Nerve cells, 2, 250, 251 in bees, 5, 234, 244 pigment in, 3, 45 Nervous disease, incidence of in old people, 1,20 Nervous system, adaptation in, 3, 61 Nilghaie, lifespan of, 3, 12 Nitrogen, diffusion of through aortic tissue, 1, 72-74 excretion of, changes due to age, 4, 243 in pollen, 5, 236 inrat body, 4, 120, 121 Normal, definition of, 3, 74, 94, 190-191 Nucleic acids, in cells, 3, 48, 49 Nutrition, and ageing, 1, 186-193 and disease in Africans, 3, 104-1 14, 145, 188-190 effect on cell, 1, 191 effect on deer antlers, 2, 184 effect on growth, 1, 186, 188 eff"ect on lifespan, 1, 186-193, 194-199; 3, 24, 25, 28, 34, 35, 36 effect on pregnant ewes, 2, 184 in Africans, 3, 188-190 neglected research areas in, 1, 173-185 Obesity, effect on kidney, 4, 254, 255 Odocoileus hemionus, lifespan of, 5, 92-94 Oesophagus, cancer of in old people, 1, 21 Oestradiol, effect on development of reproductive tract, 2, 11, 17 Oestrogen, effect on ovary, 2, 60-65 effect on placenta, 2, 142 effect on water retention, 4, 79, 84, 86 excretion of, changes in due to age, 4, 91 secretion of, 2, 49-50, 51-52 Oestrus, in X-irradiated mice, 2, 49-50 Oncorhynchus nerka, fecundity in, 5, 197- 198 growth of, 5, 156, 157, 158 Oncorhynchus spp., mortality rates of, 5, 145 Oocytes, atresia of, 2, 44-45 decline of with age, 2, 32 effect of X-irradiation on, 2, 49, 50, 51 formation of in ovary, 2, 32-48 number of, 2, 43, 47, 54-55 Oogenesis, 1, 147-149; 2, 32-48, 55-56 effect of Myleran on, 2, 43 effect on of hypophysectomy, 2, 25, 41, 44 in armadillo, 2, 39, 40 in Dasypus, 2, 56 in dog, 2, 43 in galago, 2, 56 in guinea pig, 2, 36, 39, 41, 42 in lemur, 2, 39-40 in monkey, 2, 44, 55 in rabbit, 2, 36-37 in rats, 2, 42-43 in seal, 2, 37 Ophiocephaliformes, lifespan of, 5, 224 Orange chromide, lifespan of, 5, 224 38 Cumulative Subject Index Organ culture studies of foetal reproduc- tive tract, 2, 3-17 Osier's disease, 1, 85, 105 Osmotic diuresis, 4, 40 Osteoporosis, calcium metabolism in, 1, 109, 110, 116-118, 121, 122, 123, 125 senile, incidence of, 1, 125 Ova, age changes in, 3, 126-127, 149-150 age of, 1, 172 age of at time of ovulation, 1, 147-149 ageing of, transplantation techniques in study of, 1, 150-159 lifespan of, 1, 148, 149 transfer of, 1, 150 Ovaries, age changes in, 3, 1 19, 126 transplantation techniques in study of, 1, 150-159 amyloidosis of, 3, 125 androgen production in, 2, 15-16 changes in, 1, 148 changes in at end of reproductive life, 1, 144-146 changes in at puberty, 1, 143-144 changes in before puberty, 1, 142-143 effect of age on in fish, 5, 201 effect of gonadotrophin on, 1, 150, 151, 153;2, 31,35, 55 effect of hypophysectomy on, 2, 60 effect of irradiation on, 2, 34, 49-51; 3,127 effect of oestrogen on, 2, 60-65 formation of oocytes in, 2, 32-48 grafting of, 1, 150-152, 154-157 oocyte survival in, 2, 33 growth of in herring, 5, 195 hormonal secretions of, 2, 48-52 hypertrophy of, effect on oocytes, 2, 34-35 in bees, 5, 233, 235, 237, 238, 239, 243 in Drosophila, 5, 275-277, 285 in fish, 5, 193 changes due to age, 5, 182 life of, 1, 142-147 lifespan of cells of, 1, 142 low-temperature storage of, 1, 165 regenerative capacity of, 2, 31-58 sensitivity of to hormonal environment, 1, 142-146 sensitivity to hormones, age changes, 1, 142 transplantation of, 1, 165-166; 3, 116, 122-123, 126 tumours of, development, 1, 153-154 Overwintering, 5, 232-234, 240 Ovis dalli, lifespan of, 5, 92-94 Ovis musimon, lifespan of, 3, 13, 14 Ovogenesis, 1, 172 Ovulation, age of ova at time of, 1, 147-149 in guinea pigs, 2, 71, 72, 75, 83, 84 induced by gonadotrophin, 2, 75, 79, 82 Owls, lifespan of, 5, 99 Oxygen, diffusion of through aortic tissue, 1, 72-74, 77 in cell metabolism, 2, 243 Oxygen consumption, in placenta, 2, 131- 132 Oxygen pressure, adaptation to changes, 3,64 Oxygen requirements of tissue, 1, 76, 77 Palmar erythema, 1, 82 Palometa, lifespan of, 5, 220 Pancreas, elastase in, 1, 98, 99, 100; 3, 101-102 mitochondria in, 2, 101 Pancreatic juice, bicarbonate in, 4, 64 sodium excretion in, 4, 63, 65, 71 urea in, 4, 68-69 Parabiosis, 1, 30, 170; 3, 116, 124, 127-128 Paramecium, lifespan of, 3, 16 Parasites, in fish, 5, 213-216, 228 Parental age, effect on fertility of offspring, 5,56 effect on lifespan, 5, 21-34, 43-47, 53, 259-262 "Parental death", in fish, 5, 189 Parenteral fluid therapy, 4, 144, 146-148, 151 Parotid saliva, bicarbonate in, 4, 64 potassium excretion in, 4, 63, 64, 65, 74, 75 sodium excretion from, 4, 62-63, 65, 66, 69,71 urea in, 4, 67-69, 75 Parrots, arteriosclerosis in, 5, 109, 110 Parturition, effect of corpus luteum on, 2, 79,85 Par us major, lifespan of, 5, 100 Passer domesticus, lifespan of, 5, 99 Passeres, arteriosclerosis in, 5, 109 Passerines, lifespan of, 5, 99-100 Pelecaniformes, mortality rate of, 5, 101 Pelicans, arteriosclerosis in, 5, 109 Penguin, mortality rate of, 5, 100 Peptic ulcers, incidence of in old people, 1,19 Perca fluviatilis, survival curves of, 5, 143, 144 Perch, effect of diet, 5, 178 lifespan of, 5, 220, 223 lifespan and size of, 5, 154 survival curves of, 5, 143, 144, 145 Perciformes, lifespan of, 5, 220, 223, 224 Percoidei, lifespan and size of, 5, 1 54 Performance, changes in with age, 3, 149- 169 experiments in study of, 3, 150-152 in old people, 1, 42 peak of, 1,50-51 problems of study of, 3, 152-153, 156- 159 Perognathus, lifespan of, 3, 9, 10 Peromyscus, lifespan of, 3, 9, 10 Peromyscus leucopus, lifespan of, 5, 95-96 Personality changes, in old age, 1, 42-44 Pharyngeal glands, in bees, 5, 233, 235, 237, 238, 239, 244 Pharynx, cancer of in old people, 1, 21 Cumulative Subject Index 39 Phenylalanine, in elastin, 1, 96 Phosphate, excretion of, in elderly and schizophrenics, 1, 221-222, 223-236 excretion in respiratory acidosis, 4, 266 in babies' urine, 4, 211, 213, 215,216 Phospholipsin, in placenta, 2, 122 Phosphorus, effect of protein intake, 4, 121, 127-128 excess of, 4, 144, 145-146 in body of rat, 4, 120, 121, 127-128 intake of, 4, 142 Phosphorylation, in plant cells, 2, 208 Physiological function, stability of, 5, 145 Pick's disease, 3, 134, 135-136, 143-144, 146 Pigmentation, role of in ageing processes, 1, 53, 54 Pigments, in nerve cells, 3, 44 Pigs, effect of nutrition on growth, 1, 188 Pike, lifespan of, 5, 222 Pike-killie, lifespan of, 5, 224 Pilchards, 3, 26 Pilot fish, lifespan of, 5, 22 1 Pitressin, effect on hyponatraemia, 4, 53- 55 effect on water diuresis, 4, 7-8, 1 1 variation of effect due to age, 4, 239-240 Pituitary gland, control of by testis, 2, 20- 21,28 control of on embryo, 2, 165, 166, 174 effect on electrolytes, 4, 166, 167 excision of, effect on oogenesis, 2, 35, 41,44 function of, age factor in, 2, 20-23 gonadotrophic hormone produced by, 2, 21-22, 28 relationship with adrenal cortex, 2, 22 relationship with thyroid, 2, 21 thyrotrophic hormone produced by, 2, 21-22 transplantation of, 1, 152-153 Pituitary hormones, effect on ovary, 1, 142, 143, 144-145 Placenta, acid phosphatase in, 2, 107, 108 adenosine triphosphate in, 2, 107, 108 ageing in, biochemical evidence, 2, 129- 147 morphological aspects, 2, 105-117 biochemical changes in, 2, 107-109 blood flow in, 2, 125, 126 calcium in, 2, 110 cholesterol in, 2, 122 collagen in, 2, 146 cytological changes in, 2, 107 effect of corpus luteum on, 2, 84 effect of cortisone on, 2, 141 effect of hormones on glucose metabol- ism, 2, 141-142, 146 effect of insulin on, 2, 141 effect of oestrogens on, 2, 142 enzymes in, 2, 107-108, 141, 145 fibrin in, 2, 110 fibrinoid in, 2, 110 fructose in, 2, 120, 121, 122 Placenta function of, chronological changes in, 2, 118-128 glucose in, 2, 135-137, 144, 145 glucose production in, 2, 120, 121 glycerophosphatase in, 2, 107, 108 glycogen in, 2, 27-28, 107, 108, 122, 123, 130, 133-134, 137, 159 effect of progesterone, 2, 123, 126 hyalinization in, 2, 110 in postmaturity, 2, 112-113 in toxaemia of pregnancy, 2, 112 lactic acid in, 2, 137, 141 length of gestation and, 2, 110, 111 lipids in, 2, 122 mitochondria in, 2, 107 of goat, 2, 154-158 of guinea pig, 2, 110 of rabbit, 2, 108 of rat, 2, 110 oxygen consumption in, 2, 131-132 permeability of, 2, 118-119, 126 permeability of barrier, 2, 109, 116 phosphohpsin in, 2, 122 pregnancy changes in, 2, 110-111, 115 pyruvate in, 2, 137 radio-potassium uptake of in preg- nancy, 2, 148-160 relationship with foetal liver, 2, 108, 109 sodium transfer in, 2, 111 structure of, 2, 105-106, 116 succinic dehydrogenase in, 2, 107, 108 surface area of, 2, 115 villi of, in goat, 2, 155-158 weight of, 2, 120, 124, 125, 127, 146 Placentome, in goat, uptake of radio- potassium in, 2, 151-154, 159 Plaice, fecundity of, 5, 202-206 growth of, 5, 157, 158, 159 lifespan and size of, 5, 1 50 mortality rates of, 5, 167 reproduction and growth in, 5, 189-190 Plants, amino acids in, 2, 212 Plasma, concentrations of ions in, 4, 25-26 magnesium in, 4, 99-100 potassium in, 4, 65 sodium in, 4, 65 urea in, 4, 67 Plasma proteins, in hver disease, 3, 112 Pleuronectes platessa, fecundity in, 5, 202-206 growth of, 5, 157, 158 reproduction and growth in, 5, 189- 190 Pleuronectoidei, lifespan and growth of, 5, 161-166 hfespan and size of, 5, 148 Pneumonia, incidence of in old people, 1, 19 Poeciliidae, effects of age on reproduction of, 5, 181, 186-189 Pollen, content of, 5, 237 effect on bees, 5, 236-240 40 Cumulative Subject Index Polysaccharides, in elastin, 1, 104, 107 Pompano, lifespan of, 5, 220 Population, cross-sectional surveys, 3, 158-159, 171-172, 173-174, 179, 186 horizontal surveys, 3, 94-95 longitudinal surveys, 3, 95, 158, 163- 164, 172-173, 174, 187 sampling, 3, 73-79, 81, 94-96, 156-159, 171-172, 174, 183-184 Porcupine fish, lifespan of, 5, 221 Porgy, lifespan of, 5, 221 Postmaturity, placenta in, 2, 1 12-113 Potassium, accumulation of in cell, 4, 32- 33 activity of, in pregnant rat and goat, 2, 148-154, 159 deficiency of, effects due to, 4, 140 effects on kidney, 4, 262-263 in congestive heart failure, 4, 289- 292, 299-300 deprivation of, causing cellular oedema, 4,32 effect ofprotein intake, 4, 121, 126, 129, 133 effect on heart, 4, 95-96 excess of, effects due to, 4, 140, 141, 144-146, 152 in adrenal hyperplasia, 4, 80, 95 exchangeable amounts in body, 4, 108, 109, 111, 114 exchange of in cell, 4, 18, 19, 20, 21, 22, 24, 30, 34 excretion of, after water loading, 4, 167-169 effect of ACTH and cortisone, 4, 176, 177, 178 effect of aldosterone, 4, 183-184, 186, 192-194, 196-197 effect of cortexone, 4, 174, 175 effect of Cortisol, 4, 188, 189, 192-194, 196 effect of cortisone, 4, 171, 172, 176, 178 effect of vasopressin, 4, 170 in elderly and schizophrenics, 1, 221- 222, 223-236 in parotid saliva, 4, 63, 64, 65, 74, 75 in respiratory acidosis, 4, 266 in sweat, 4, 63, 64, 65, 74, 76, 77 in body of rat, 4, 120, 121, 126, 129, 133 in erythrocytes, 4, 200-202, 204, 206, 207, 208 in foetal urine, 4, 218 in muscle, 4, 224-225, 289-292 in plasma, 4, 65 in saliva, during menstrual cycle, 4, 83-88 loss of, 4, 226 during labour, 4, 90 ranges of intake, 4, 142 Potassium chloride, effect on hyponatrae- mia, 4, 53 Potassium pump, 4, 204 Poultry, arteriosclerosis in, 5, 110, 113 Pregnancy, aldosterone excretion in, 4, 89-90 duration of and placenta, 2, 110, 111 effect of corpus luteum on, 2, 78, 79, 84 effect of cortisone on foetus, 2, 167-175 effect of growth hormone on foetus, 2, 161-167, 171, 173-175 effect of progesterone, 2, 126, 127 effect on apocrine sweat glands, 2, 197- 198, 199-200 endometrial changes in, 2, 115 placental changes in, 2, 110^-111, 115 placental uptake of radio-potassium during, 2, 148-160 prolongation of, 2, 163 sodium retention during, 4, 88-89 toxaemias of, 2, 1 1 1-1 12, 142, 144 uterine uptake of radio-potassium in, 2, 148-160 water and electrolyte changes in, 4, 88- 90 water retention during, 4, 88-89 Premenstrual oedema, 4, 81-83 Preservation of tissue, 1, 162-172 Procellariiformes, mortality rate of, 5, 101 Progesterone, 1, 127, 128 effect on placental glycogen, 2, 123, 126 effect on pregnancy, 2, 126, 127 Proline, in elastin, 1, 91, 94, 95, 96 Prostate gland, cancer of in old people, 1, 21 in female animals, 2, 14, 15, 16 normal development of, 2, 5-6, 9, 10, 11, 14,20 Proteins, age changes in, 3, 66 breakdown of, causing osmotic diuresis, 4,41 catabolism in senescent leaves, 2, 202- 205,206,207-208,212,214 effect of age on, 1, 103 effect on bees, 5, 254 effect on Drosophila, 5, 266-267 in barley leaves, 2, 203-205 in bees, 5, 234 in cells, 3, 48, 49 in children's diets, 1, 198 in diet, 5, 252, 253, 254 in elastin, 1, 94-96 in pollen, 5, 237, 238 in renal disease, 4, 260-262 intake of, effect on body composition, 4, 116-138 storage in bees, 5, 237, 238, 239 utilization of in growth offish, 5, 182- 186,210 Protein metabolism, relationship with cell respiration, 2, 207-208 Protoplasm, 1, 244 Protozoa, nutrition and growth, 1, 186-187 Pseudopregnancy, 3, 117, 126, 129-130 Psittaci, arteriosclerosis in, 5, 108-109, 110 Psychological aspects of ageing, 1, 209-218 Pudding wife, lifespan of, 5, 221 Cumulative Subject Index 41 Puffer, lifespan of, 5, 221 Pulmonary artery, calcification of elastic tissue in, 1, 98 elastin content of, 1, 93-94 Pulmonary oedema, 4, 284—285 Pygosteus pungitius, growth of, 5, 156, 157, 158 Pyrogen, effects of kidney, 4, 235-238 Pyruvate, in placenta, 2, 137 Pyruvic acid, in urine, 4, 221 Rabbit, adrenal cortex of, 2, 22 cottontail, lifespan of, 5, 97 elastin in aorta of, 1, 101 in research on ageing, 1, 181 lifespan of, 1, 181 ; 3, 11 ; 5, 97 oogenesis in, 2, 36-37 ovulation in, 1, 143 pituitary function in foetal, 2, 20-22 placenta of, 2, 108, 126 renal blood content of, 3, 52, 58 Radioiodine, deposition of in thyroid 3, 52-58 Radio-phosphorus, uptake of in deer antlers, 2, 177-178 Radio-potassium, uptake of in uterus and placenta, 2, 148-160 Rajiformes, lifespan and size of, 5, 1 54 lifespan of, 5, 220 Rangifer arcticus, lifespan of, 5, 92-94 Rasbora, lifespan of, 5, 224 Rat, adrenal cortex of, 2, 22 brown, lifespan of, 5, 95-96 calcium metabohsm in, 1, 111-114, 124-125 causes of death in, 5, 82-83 disease in, effect on lifespan, 5, 72-89 effect of diet on, 5, 78, 83-85, 87, 88, 251-252, 254, 265, 268 effect of somatotrophic hormone and cortisone on foetal growth of, 2, 161- 175 foetal, organ culture studies in repro- ductive tracts of, 2, 3-17 growth in, 1, 204 hypothermia in, 1, 168 in research on ageing, 1, 178-180 Kangaroo, hfespan of, 5, 96 lifespan of, 1, 177, 178-179, 180; 5, 95-96, 251-252 lung disease in, 5, 82 maze experiments with, 1, 215-216 memory in, 1, 215 nephrosis in, 5, 83 oestrus in, 2, 56-57 oogenesis in, 2, 42-43 ovarian changes in, 1, 146 placenta of, 2, 110 placental enzymes of, 2, 107-108 radio-potassium uptake during preg- nancy, 2, 148-160 redundant follicles in, 2, 59-68 response to stimuli, 3, 51, 57 thyroid function in, 3, 52-55 Rattlesnake, fertility in, 3, 22 Rat t us rat t us, lifespan of, 5, 95-96 Rays, growth of, 5, 163 lifespan of, 5, 154, 163,221 size of, 5, 154 Rectum, cancer of in old people, 1, 21 Red cells, adenosine triphosphate in, 2, 238 ageing in, 2, 233-238 cholinesterase activity in, 2, 235 dissolution, 2, 218-229 effect of cold on, 2, 224-229, 236-237, 244 enzymes in, 2, 235, 241 freezing of, 2, 215-216, 236-237 haemolysis of, 2, 222, 225, 226, 227 lifespan of, 2, 217, 233, 234, 238, 239, 241,245 low temperature storage of, 1, 164, 171 membrane of, 2, 222, 223 metabolism of, 2, 217, 227, 228, 240 ofinfants, 2, 233, 239 physical instability of, 2, 215-231 preservation of, 2, 215 self-repair in, 2, 234 structure of, 2, 223 Red hind, hfespan of, 5, 220 protein metabolism in, 5, 183-185 Regeneration, in guppies, 5, 208-210 Regenerative capacity of the ovary, 2, 31- 58 Regression, 1, 33-34 Reinke, crystalloids of, 2, 92, 94, 95 Renal artery disease, in cases of cancer, 1, 21 Renal blood flow, 3, 193-194 age changes in, 3, 88 Reproduction, effect of age on in fish, 5, 181-182, 186-206 effect of nutrition on, 1, 192 effect of ovarian transplantation on, 1, 158, 159 effect on hfespan, 5, 179, 275-279, 284- 285 infish, 5, 170-174 Reproductive organs, ageing of, tissue transplantation techniques in, 1, 141- 161 Reproductive tract, organ culture studies in, 2, 3-17 Respiration, in barley leaves, 2, 206 in senescent leaves, 2, 205-209, 212, 213,214 Respiratory acidosis, 4, 265-266 Respiratory disease, incidence of in old people, 1, 20 Respiratory failure, renal function in, 4, 264-270 Respiratory function, effects of ageing on, 1, 58-68 Retirement, effects of, 1, 49-50 Rhinolophus, 3, 27 Ribonucleic acid, in apocrine sweat glands, 2, 194 42 Cumulative Subject Index Roach, lifespan of, 5, 222 Rockfish, lifespan of, 5, 220 Rodents, ageing in, 3, 9-10, 11, 12-13 Rosy tetra, lifespan of, 5, 224 Rubner's theory of ageing, 5, 125-126, 128 Rudder fish, lifespan of, 5, 221 Sailfish, lifespan and size of, 5, 154 Saliva, bicarbonate in, 4, 63, 64, 65, 74, 75, 83-88 sodium in, 4, 62-63, 65 66, 69, 71, 83- 88 sodium/potassium ratios in, 4, 94 urea in, 4, 67-69, 75 Salivary glands, in bees, 5, 233, 235, 237, 238, 239 Sahnon, fecundity of, 5, 192, 197-198 growth of, 5, 156, 157, 158, 160-166 lifespan of, 5, 150, 152, 160-166, 222 mortality rates of, 5, 145, 166 natural death in, 5, 174 relation of size and maturity, 5, 172 reproduction in, 5, 179 size of, 5, 150, 152 Salmo gairdneri, growth rates of, 5, 182 Salmo salar, fecundity in, 5, 197-198 Svlmonoidei, lifespan and size of, 5, 1 50 lifespan and growth of, 5, 161-166 Sahelinus alpinus, lifespan of, 5, 147 Sampling, 3, 74, 77, 79, 81, 171-172, 174, 183 Sand dab, lifespan and size of, 5, 148 Sardina pilchardis, effect of environment on, 5, 168 Sardines, eflfect of environment on, 5, 168 Sanger, lifespan and size of, 5, 1 54 Scandinavians, lifespan of, 5, 21-26 Scar tissue, collagen in, 3, 70 Scat, lifespan of, 5, 221 Sceloporus, 3, 36, 27 Schizophrenics, adrenal cortex activity in, 1, 219-238 cortin excretion in elderly, 1, 221-222, 223-236 creatinine excretion in elderly, 1, 221, 222, 223-236 effect of ACTH on, 1, 229 17-ketosteroid excretion in elderly, 1, 221-222, 223-236 phosphate excretion in elderly, 1, 221- 222, 223-236 potassium excretion in elderly, 1, 221- 222, 223-236 sodium excretion in elderly, 1, 221-222, 223-236 uric acid excretion in elderly, 1, 221- 222, 223-236 Scombroidei, lifespan and size of, 5, 1 54 Sea anemone, 1, 244 Sea horse, lifespan of, 5, 147 lifespan and size of, 5, 154 Seal, oogenesis in, 2, 37 "Second childhood", 1, 34 Seeds, ageing of, 2, 211 Seminal vesicles, normal development of, 2, 5-6, 8-9, 10 Senescence, {see also Ageing) definition and measurement of, 1, 4-15, 31 "Senile death", in fish, 5, 189 Senile elastosis, 1, 101-103, 106-108 Serine, in elastin, 1, 96 Serotonin metabolism, 3, 145, 146 Sex differences, in body water, 4, 107-1 10, 113 in growth offish, 5, 159 in lifespan, 3, 30 in mortality among fish, 5, 145, 166-167 in reactions to environment in fish, 5, 178 Sex distribution, in emphysema, 1, 68 Sex organs, changes in due to age in fish, 5,218 in fish, 5, 229 relationship to body weight in fish, 5, 193-196 Sexual maturity, age of, 1, 202-203 and lifespan, 1, 29, 30 effect of diet, 5, 84-85 infish, 5, 170-171 Shad, fecundity of, 5, 191-192 Shark, lifespan of, 5, 220 Sheep, erythrocytes, 4, 200-203, 204, 206 in research on ageing, 1, 182 lifespan of, 3, 12, 13, 14, 15, 30; 5, 91-94 placenta in, 2, 126 Sheepshead, lifespan of, 5, 221 Shrews, 3, 28 lifespan of, 5, 97 Siderosis, 3, 107-112, 188, 189 Siluroidei, lifespan and size of, 5, 154 Size, effect on lifespan, 5, 147-159 Skeleton, ageing of, 1, 109-125 Skill, effect on performance, 3, 160-161 variations with age, 1, 209-218 Skin, age changes in, 1, 53 ; 3, 96 ectopic flexure lines of, 1, 12-13 elastic tissue in, 1, 100-102, 105, 106, 107, 108 elastic tissue of, ageing of, 1, 88-108 temperature of, age changes in, 3, 80-84 vascular lesions of, incidence of in old age, 1, 80-87 water absorption by, 4, 100-101 Skipjack, metabolism of, 5, 170 Smelt, lifespan of, 5, 147 lifespan and size of, 5, 150 Snakes, fertility of, 3, 21, 22 Snake-head, lifespan of, 5, 224 Snapper, lifespan of, 5, 221 Social adjustment, 3, 175 Sodium, and adrenal function, 4, 166 deficiency of {see also Hyponatraemia) causing hyponatraemia, 4, 44-45 effect of ACTH and cortisone, 4, 176, 177, 178 effect of protein intake on, 4, 121, 126, 132 Cumulative Subject Index 43 Sodium eflFect on water intake, 4, 37 excess of {see Hyperaatraemia) exchangeable amounts in body, 4, 108 exchange of in cell, 4, 18, 19, 20, 21, 22, 24, 30, 34 excretion of, 4, 62-63 after water loading, 4, 167-169 during exercise, 4, 279 effect of aldosterone, 4, 183, 185, 192-194, 196 eflFect of cortexone, 4, 174, 175, 177 eflfect of Cortisol, 4, 187-188, 189, 192-194 effect of cortisone, 4, 171, 172, 173, 178 eflfect of vasopressin, 4, 170 in congestive heart failure, 4, 277, 288, 289, 299 in elderly and schizophrenics, 1, 221- 222, 223-236 in pancreatic juice, 4, 63, 65, 71 in parotid saliva, 4, 62-63, 65, 66, 69, 71 in respiratory acidosis, 4, 266 in sweat, 4, 62-63, 65, 66, 69, 71, 74, 75, 76, 77 in tears, 4, 63, 65, 71, 75, 76 in body of rat, 4, 120, 121, 126, 132 in erythrocytes, 4, 200, 201-202, 203, 206, 207, 208 in foetal urine, 4, 217 in plasma, 4, 65 in saliva, during menstrual cycle, 4, 83- 88 in submaxillary gland, 4, 71-72 loss of, in adrenal hyperplasia, 4, 79-80 ranges of intake, 4, 142 retention of, in congestive heart failure, 4,288 in pregnancy, 4, 88-89 transfer of, in placenta, 2, 1 1 1 Sodium/potassium ratios, during menstrual cycle, 4, 83-88 eflfect of aldosterone on, 4, 184-185, 186-187, 192-194, 196-197 eflfect of Cortisol on, 4, 190-192, 193- 194, 196 in saliva, 4, 94 Sodium pump, 4, 203, 207 Sole, lifespan and size of, 5, 1 50 Sorex araneus, lifespan of, 5, 97 Souslik, lifespan of, 5, 97 Soya flour, eflfect on bees, 5, 235 Spadeflsh, lifespan of, 5, 221 Sparrows, lifespan of, 5, 99 Spermatids, 2, 87-96 Spermatocytes, division of, 2, 87 Spermatogenesis, 2, 86-91 age changes in, 2, 97 condensation of karyoplasm in, 2, 90- 91 formation of aerosome and head cap, 2, 88-89 Spermatozoa, de%elopment of, 2, 87-91 low-temperature storage of, 1, 164, 171 Sphenisciformes, mortality rate of, 5, 101 Spidernaevi, 1, 81-82 Spirocercosis, 3, 33, 102 Sprat, lifespan and size of, 5, 148 Squirrel fish, lifespan of, 5, 220 Stallions, causes of death in, 5, 68 lifespan of, 5, 65-67 Starvation, eflfect of age, 4, 266 Steady state system, 2, 216, 228 Stentor, 3, 46 Sterility, in cows, 5, 63, 64 Sterna hirundo, lifespan of, 5, 97-99 Steroid metabolism, changes due to age, 4, 90-92 Steroids, administration of in old age, 1, 134-136 excretion of, in urine, 1, 220 Sticklebacks, 3, 26 eflfect of temperature on, 5, 168 growth of, 5, 156, 157, 158 lifespan of, 5, 152, 220, 223 size of, 5, 152 Stimuli, reactions to at diflferent ages, 3, 51-58 Stomach, cancer of in old people, 1, 21, 24 Strength, in old age, 1, 137-138 Stress, 5, 88 eflfect on kidney, 4, 260 recurrent, 1, 12, 13, 27 Strigiformes, mortahty rate of, 5, 101 Strix aluco, lifespan of, 5, 99 Sturgeon, growth of, 5, 156, 157, 158, 163 lifespan of, 5, 147, 152, 163, 220, 222, 226 oldest age of, 5, 191 reproduction in, 5, 171 size of, 5, 152 survival curves of, 5, 143, 144 Sublingual gland, electrolytes in, 4, 69-70 Submaxillary gland, electrolytes in, 4, 69- 70 Succinic dehvdrogenase, in placenta, 2, 107, 108 Sugar, eflfect on bees, 5, 238 excretion of, in elderly and schizo- phrenics, 1, 231 Sunfish, eflfect of environment on, 5, 168 lifespan of, 5, 223 protein metaboUsm of, 5, 182-183 Sweat, 4, 100 chlorides in, 4, 64, 74 in mercury poisoning, 4, 99 potassium excretion in, 4, 63, 64, 65, 74, 76,77 sodium excretion in, 4, 62-63, 65, 66, 69, 71, 74, 75, 76, 77 urea in, 4, 67-69 Sweat glands, apocrine, (see Apocrine sweat glands) Swifts, lifespan of, 5, 100, 103, 104 Syngnathiformes, lifespan and size of, 5, 154 44 Cumulative Subject Index Tarpon, lifespan of, 5, 220 Tears, chloride in, 4, 64, 71 sodium excretion in, 4, 63, 65, 71, 75, 76 urea in, 4, 67-69 Temperature, adaptation to change, 3, 63, 64 effect on bees, 5, 245, 246 effect on cell, 5, 281 effect on dissolution of red cells, 2, 218- 229 effect on lifespan, 3, 24-25; 5, 167-168 in Drosophila, 5, 271-279, 283, 284 effect on protein metabolism, 5, 185- 186 effect on senescence in plants, 2, 210 Tench, lifespan of, 5, 223 Tenebrio molitor, effect of parental age on lifespan, 5, 262 Tentorium cerebelli, diffusion of solutes through, 1, 74-75 Terns, lifespan of, 5, 97-99, 104 Testis, ageing in, 2, 97 cytomorphosis of cells in, 2, 86-99 effect on development of reproductive tract, 2, 7-17, 19-20 interstitial cells of, 2, 91-96 ketosteroids in, 2, 98, 99 Leydig cell in, 2, 91, 92, 93, 94, 95, 97, 98,99 of deer, 2, 179 pituitary control by, 2, 20-21, 28 transplantation of, 3, 123 Testosterone, 1, 128, 129 administration of in old age, 1, 134- 136 effect on deer antlers, 2, 179-180, 186 effect on foetal growth, 2, 173 Tests, mental, 1, 137 Tests of ability and function, 1, 137 Tetrodontiformes, lifespan of, 5, 221 Thalamus, in old people, 1, 35-36, 49 Thirst, effect of, 4, 143, 144 failure of, 4, 41-43 Thirst centre, 4, 37 Threonine, in elastin, 1, 96 Thrombosis, in centenarians, 1, 16 Thunniformes, lifespan and growth of, 5, 163 lifespan and size of, 5, 1 54 Thymallus signifer, 3, 26 effect of environment on, 5, 168 Thyroid, functions, 3, 52-55, 58 radio-iodine uptake in, 3, 52-58 relationship with pituitary gland, 2, 21 transplantation of, 1, 157-158 Thyrotrophic hormone, produced by pituitary gland, 2, 21-22 Time, appreciation of, age differences in, 1,217-218 Tissue, biophysical changes during freez- ing, 1, 171, 172 oxygen requirements of, 1, 76, 77 preservation of, in study of ageing, 1, 162-172 Tissue changes, in old age, 1, 55, 78 in old people, 1, 27 Tissue transplantation techniques, in the ageing of reproductive organs, 1, 141- 161 Tits, lifespan of, 5, 100 Toadfish, lifespan of, 5, 221 Tobacco leaves, protein in, 2, 203 Tortoise, ageing in, 1, 28 Totoaba, lifespan and size of, 5, 154 Toxaemia of pregnancy, 4, 88, 89-90 Transplantation, in inbred strains, 3, 116, 120, 122-123, 126, 129 Trigger fish, lifespan of, 5, 221 Triton, lifespan of, 3, 9 Trout, effect of diet on, 5, 169, 253, 254 effect of reproduction on, 5, 179 fecundity in, 5, 192, 197 food conversion in, 5, 186 growth rates of, 5, 182 lifespan of, 1, 29; 5, 150, 152, 222, 265 size of, 5, 150, 152 mortality rates of, 5, 145 Trypsin, 3, 101 Tryptophan, in elastin, 1, 91, 96 Tuberculosis, in cows, 5, 63, 64 Tuna, fecundity in, 5, 193 lifespan and size of, 5, 1 54 Turtles, in research on ageing, 1, 177 Twins, data from, in genetics of ageing, 3, 131-148 lifespan of, 3, 140, 141 Tyrosine, in elastin, 1, 96 Undernutrition, effect on children's weight, 1, 195 Ungulates, Ufespan of, 5, 91-95 Uterus, effect of corpus luteum on, 2, 78- 79 isolated, 1, 167 low-temperature storage of, 1, 167 radio-potassium uptake of in pregnancy, 2, 148-160 Urea, excretion of, 4, 40-41, 73 in pancreatic juice, 4, 68-69 in parotid saliva, 4, 67-69, 75 in plasma, 4, 67 in sweat, 4, 67-69 in tears, 4, 67-69 Uric acid, excretion of, in elderly and schizophrenics, 1, 221-222, 223-236 Urinary disease, incidence of, in old peo- ple, 1, 18, 19, 20 Urine, acids excreted in, 4, 210, 215-217, 220-222 ammonium salts in, 4, 209-210 calcium excretion in, 1, 120, 121, 122, 123 cychc changes in, 1, 236-237 inbabies, 4, 210-211 in foetus, 4, 217 magnesium excretion in, 4, 305, 309 output, effect of aldosterone on, 4, 182, 192-194, 196 Cumulative Subject Index 45 Urine output (continued) effect of Cortisol, 4, 187, 188, 192-194, 196 pH of, 4, 209-210, 211-212, 215, 221, 222 potassium in during menstrual cycle, 4, 84-88 sodium in, during menstrual cycle, 4, 84-88 steroid excretion in, 1, 220 Urogenital sinus, development of, 2, 20 Valine, in elastin, 1, 91, 94, 95, 96 Variability, 3, 71 Vascular lesions of skin, 1, 80-87 Vasopressin, effect on electrolytes, 4, 167, 170 effect on water diuresis, 4, 12, 13, 169, 170, 195 Venous lakes, 1,85 Venous pressure, in congestive heart failure, 4, 272, 274, 275 Venous stars, 1, 82, 86 Ventilation, measurement of, 1, 62, 63, 66 Vertebrates, lifespan of, 3, 9, 10, 11-17 Virus infection, in leaves, 2, 21 1 Vital capacity, variations with age, 1, 58, 59,66 Vitamin A, effect on kidney function, 4, 247 Vitamin B, deficiency in animals, 3, 35 Vitamin B12, importance of, 1, 197, 198, 199 Vitamin D, in calcium absorption, 1, 110, 111-114, 124 Vitamins, effect on bees, 5, 238 in pollen, 5, 237, 238, 239 Vultures, arteriosclerosis in, 5, 110 Walleye, fecundity/length relationship in, 5, 197 Wasps, 5, 232, 255 Water, cellular aspects of in body, 4, 15-35 content, control of, 4, 10-11 variations with age, 1, 207 deficiency of, causing hypernatraemia, 4, 38-44 effects of, 4, 140, 160, 163 in children, 4, 160 symptoms, 4, 39 deprivati o n of , effect on hyponatraemia, 4, 51-52 diuresis, at various ages, 4, 6-10 effect of adrenahne, 4, 9, 14 effectofage, 4, 238-240 effect of pitressin, 4, 7-8, 11 in congestive heart failure, 4, 272- 273, 275 effect of load in rats, 4, 167-168 effect of vasopressin on loss of, 4, 169, 170 Water excess of, effects due to, 4, 46-47, 140, 144, 145-146, 150, 151 effect on diuresis, 4, 6, 8 effect on hyponatraemia, 4, 52 effect on urine output, 4, 4 in children, 4, 159 excretion, during exercise, 4, 279 response to adrenal steroids, 4, 180- 194, 196-198 exchanges of in body, 4, 3 extracellular, changes in, 3, 191, 192 in adults, 4, 106-110 in children, 4, 103-106 variations with age, 4, 31, 110-112, 114, 115 in body, 4, 102-115 effect of age on, 4, 110-112, 114, 115, 180 effect of growth, 4, 103-106 measurements of, 4, 102-103 in muscles, 4, 113, 163-164 in parenteral fluid therapy, 4, 146-148 in rat body, 4, 119, 122, 123 intake of, control of, 4, 9-10 intracellular, and cardiac function, 3, 88 effects of age on, 4, 110-112, 114, 115 in adults, 4, 106-110 in children, 4, 103-106 lossof, 4, 39-41, 195 during labour, 4, 90 effect of ACTH and cortisone, 4, 176,. 177, 178 effect of cortexone, 4, 174, 175 effect of cortisone, 4, 171, 172, 176,. 178 following adrenalectomy, 4, 172 metabolism, hormonal aspects of, 4, 78-98 in congestive heart failure, 4, 271— 300 in infants, 4, 96-98, 154-164 in malnutrition, 4, 156-157 in pregnancy, 4, 88-90 movement of in cell, 4, 19, 20, 22, 25, 27-29, 34 physiological regulation of, 4, 3-14 ranges of intake, 4, 142 regulation of, 4, 37-38 by kidney, 4, 229-249 retention of, eff'ects of oestrogen, 4^ 79 in congestive heart failure, 4, 288 in pregnancy, 4, 88-89 in premenstrual period, 4, 150, 151 tolerance to excess, 4, 150, 151 Water load, effects of, 4, 170 Wax glands in bees, 5, 233, 239 Weakfish, lifespan of, 5, 221 Weight, effect on lifespan, 1, 203 Whitebait, lifespan and size of, 5, 150 mortahty rate of, 5, 146 46 Cumulative Subject Index Whitefish, growth of, 5, 156, 157, 158 lifespan of, 5, 1 50, 222 size of, 5, 150 survival curves of, 5, 143, 144 Whiting, lifespan and size of, 5, 154 Wolffian ducts, effect on Miillerian ducts, 2,16 normal development of, 2, 5-6, 7-8, 10, 19 Wolfhound, lifespan of, 3, 12, 13 Wolves, lifespan of, 3, 12, 14 Wrinkling, 1, 12-13, 23, 27 X-irradiation, effect on follicular atresia, 2, 67, 68 effect on oocytes, 2, 34, 46, 49-51 Yellow tail, lifespan of, 5, 221 Yolk-sac, mitochondria in, 2, 101 Printed by Spottiswoode, Ballantyne <& Co. Ltd., London and Colchester V