.""Tv i" j.. i % =* <^^is* / 193O MBL/WHOI LIBRARY memory of ru cr u. -j; o m CD MONOGRAPHS ON EXPERIMENTAL BIOLOGY EDITED BY JACQUES LOEB, Rockefeller Institute T. H. MORGAN, Columbia University W. J. V. OSTERHOUT, Harvard University THE BIOLOGY OF DEATH BY RAYMOND PEARL THE JOHNS HOPKINS UNIVERSITY MONOGRAPHS ON EXPERIMENTAL BIOLOGY THE BIOLOGY OF DEATH Being a Series of Lectures Delivered at the Lowell Institute in Boston in December 1920 BY RAYMOND PEARL THE JOHNS HOPKINS UNIVERSITY PHILADELPHIA AND LONDON J. B. LIPPINCOTT COMPANY COPYRIGHT, 1922, BY J. B. LIPPINCOTT COMPANY PRINTED BY J. B. LIPPINCOTT COMPANY AT THE WASHINGTON SQUARE PRESS PHILADELPHIA, U. S, A, TO MY WISEST COUNSELLOR M. D. P. EDITOR'S ANNOUNCEMENT THE rapidly increasing specialization makes it im- possible for one author to cover satisfactorily the whole field of modern Biology. This situation, which exists in all the sciences, has induced English authors to issue series of monographs in Biochemistry, Physiology, and Physics. A number of American biologists have decided to provide the same opportunity for the study of Experimental Biology. Biology, which not long ago was purely descriptive and speculative, has begun to adopt the methods of the exact sciences, recognizing that for permanent progress not only experiments are required but that the experi- ments should be of a quantitative character. It will be the purpose of this series of monographs to emphasize and further as much as possible this development of Biology. Experimental Biology and General Physiology are one and the same science, by method as well as by contents, since both aim at explaining life from the physico-chemical constitution of living matter. The series of monographs on Experimental Biology will therefore include the field of traditional General Physiology. JACQUES LOEB, T. H. MORGAN, W. J. V. OSTERHOUT. AUTHOR'S PREFACE IN preparing the material of a series of lectures, given at the Lowell Institute in Boston in December 1920, for book publication, I have deemed it on the whole best to adhere rather closely to the original lecture mode of pre- sentation with all its informality. Except for the fact that the matter is here set forth in somewhat greater detail than was possible under the rigid time limitations of the Lowell Institute, and that the breaking into chap- ters is slightly different, the whole is substantially as it was presented in Boston. What I tried to do in these lectures was to bring together under a unified viewpoint some of the more im- portant contributions which have been made to our know- ledge of natural death, from three widely scattered sources: namely general biology, experimental biology, and statistical and actuarial science. It will be obvious to anyone who knows the literature from these fields regarding natural death and the duration of life that in such an amount of space as is here used, no one could hope to cover a field so wide with anything approaching completeness. To do so would require a series of volumes in place of one small one. But this has in no wise been my object ; I have instead hoped that the very incomplete- ness itself of this work, necessitated by my limitations of space and knowledge, might stimulate the reader to penetrate for himself further into the literature of this fascinating and important field of biology. To help him to start upon this excursion a brief bibliography is appended. It by no means completely covers the field, but may perhaps serve as an introduction. 9 10 AUTHOR'S PREFACE I am indebted to a number of authors and publishers for permission to use illustrations and wish here to ex- press my great appreciation of this courtesy. The indi- vidual sources for these borrowed figures are in every case indicated in the legends. To Dr. J. McKeen Cattell I am especially grateful for allowing me the use of the blocks from the magazine publication of this material in the Scientific Monthly; to Dr. Alexis Carrel for permis- sion to use unpublished photographs of his tissue cultures ; and, finally, to Professor T. H. Morgan for critically reading the manuscript and making many helpful suggestions. E. P. BALTIMORE, April 19, 1922. CONTENTS CHAPTBR PAGE I. THE PROBLEM 17 II. CONDITIONS OP CELLULAR IMMORTALITY 51 III. THE CHANCES OP DEATH 79 IV. THE CAUSES OP DEATH 102 j V. EMBRYOLOGY AND HUMAN MORTALITY 138 VI. THE INHERITANCE OP DURATION OP LIFE IN MAN 150 VII. EXPERIMENTAL STUDIES ON THE DURATION OP LIFE 186 VIII. NATURAL DEATH, PUBLIC HEALTH, AND THE POPULATION PROBLEM 223 BIBLIOGRAPHY 259 INDEX . 269 ILLUSTRATIONS FIG. PAGE 1. Photograph of John Shell, claimed to be 131 years old, but actually about 100, with his wife and putative son (From Nascher) 26 2. Showing the changes in nerve cells due to age (From Donaldson after Hodge) 29 3. Paramecium, viewed from the oral surface (From Jennings) 31 4. Diagram showing the process of reproduction by fission in the uni- cellular organism Paramecium 32 5. Conjugation in Paramecium 32 6. Planaria dorotocephala (From Child) 34 7. Beginning of process of agamic reproduction by fission in Planaria (From Child) 35 8. Progress of agamic reproduction in Stenostomum (From Child) .... 36 9. Section across the posterior part of an embryo dog-fish (Acanthias) of 3.5 mm. (From Minot after Woods) 38 10. First and second division in egg of Cyclops (From Child) 39 11. Diagram to show mode of descent (Modified from Jennings) 41 12. Artificially parthenogenetic frogs (Loeb) 52 13. Piece of tissue from frog embryo cultivated in lymph (From Harrison) 58 14. Group of nerve fibers which have grown from an isolated piece of neural tube of a chick embryo (From Harrison after Burrows) .... 59 15. Human connective tissue cells fixed and stained with Giemsa stain (After Losee and Ebeling) 60 16. Pennaria (From Wilson) 62 17. Culture of old strain of connective tissue (Ebeling) 63 18. Life table diagram 81 19. Comparing the expectation of life in the 17th century with that of the present time 84 20. Comparing the expectation of life in the 18th century with that of the present time 86 21. Comparing the expectation of life of ancient Egyptians with that of present day Americans 88 22. Comparing the expectation of life of ancient Romans with that of present day Americans 90 13 14 ILLUSTRATIONS FIG. PAGB 23. Comparing the expectation of life of the population of the Roman provinces Hispania and Lusitania with that of present day Americans 91 24. Comparing the expectation of life of the population of the Roman provinces in Africa with that of present day Americans 92 25. Showing Pearson's results in fitting the dx line of the life table with 5 skew frequency curves 95 26. Showing the relative importance of the different organ systems in human mortality 108 27. Diagram showing the specific death rate at each age for deaths from all causes taken together 116 28. The specific death rate at each age from breakdown of the circulatory system, blood and blood forming organs 118 29. The specific death rate at each age from breakdown of the respira- tory system 120 30. Specific death rates at each age from breakdown of the primary and secondary sex organs 125 31. Specific death rates at each age from breakdown of the kidneys and related excretory organs 127 32. Specific death rates at each age from breakdown of the skeletal and muscular systems 128 33. Specific rates of death at each age from breakdown of the alimentary tract and associated organs of metabolism 129 34. Specific death rates at each age from breakdown of the nervous sys- tem and sense organs 130 35. Specific death rates at each age chargeable against the skin 132 36. Specific death rates at each age from breakdown of the endocrinal system 133 37. Specific death rates from all other causes of death not covered in the preceding categories 134 38. Percentages of biologically classifiable human mortality resulting from breakdown of organs developing from the different germ layers. . . 140 39. Specific death rates in males according to the germ layer from which the organs developed 144 40. Specific death rates for females 146 41. Survival curves of members of the Hyde family (Plotted from Bell's data) 153 ILLUSTRATIONS 15 42. Influence of father's age at death upon longevity of offspring (After Bell) ..................................................... 156 43. Influence of mother's age at death upon longevity of offspring (After Bell) ..................................................... 157 44. Influence of age at death of parents upon the percentage of offspring dying under 5 years (After Ploetz) ............................ 178 45. Snow's results on selective death rate in man .................... 182 46. Male and female fruit fly (Drosophila melanogaster) (From Morgan) . 187 47. Life lines for Drosophila melanogaster ............................ 188 48. Life lines for different inbred lines of descent in Drosophila ......... 192 49. Life lines showing the result of Mendelian experiments on the dura- tion of life in Drosophila .................................... 195 50. Distribution of poverty in Paris U911-13). (After Hersch) ........ 202 51. Death rates in Paris (1911-13) from all causes. (After Hersch) ..... 203 52. Trend of death rates for four causes of death against which public health activities have been particularly directed ................ 230 53. Trend of death rates from four causes of death upon which no direct attempt at control has been made ............................ 232 54. Trend of combined death rate from the four causes shown in figure 52 as compared with the four causes shown in figure 53 .......... 233 55. Course of the weighted average death rate for the countries in the A and B groups, from typhoid fever ............................. 236 56. Like figure 55, but for diphtheria and croup ..................... 237 57. Record of malaria control by antimosquito measures, Crossett, Ark., 1916-1918. (From Rose) .................................... 241 58. Disappearance of yellow fever from Guayaquil, Ecuador, as a result of control measures. (By permission of International Health Board) 242 59. Showing the change in percentage which deaths were of births in each of the years 1912 to 1919 ............................... 246 60. Theoretical curve of population growth ......................... 249 61. Curve of growth of the population of the United States ............ 250 62. Curve of growth of the population of France ..................... 251 63. Curve of growth of the population of Serbia ...................... 253 64. Growth of a Drosophila population kept under controlled experi- mental conditions. . 254 THE BIOLOGY OF DEATH CHAPTER I THE PROBLEM PROBABLY no subject so deeply interests human beings as that of the duration of human life. Presumably just because the business of living was such a wonderfully interesting and important one from the viewpoint of the individual, man has endeavored, in every way he could think of, to prolong it as much as possible. He has had recourse to both natural and supernatural schemes for attaining this objective. On the mundane plane he has developed the sciences and arts of biology, medicine and hygiene, with the fundamental purpose of learning the underlying principles of vital processes, so that it might ultimately be possible to stretch the length of each indivi- dual's life on earth to the greatest attainable degree. Recognizing pragmatically, however, that at best the limi- tations in this direction were distinctly narrow, when conceived in any historical sense, he has with singularly wide-spread unanimity, deemed it wise to seek another means of satisfying his desires. Man's body plainly and palpably returns to dust, after the briefest of intervals, measured in terms of cosmic evolution. But, patent as this fact is it has not precluded the postulation of an infin- ite continuation of that impalpable portion of man's be- ing which is called the soul. With the field thus open we 2 17 18 THE BIOLOGY OF DEATH see some sort of notion of immortality incorporated in an integral part of almost all folk philosophies of which any record exists. Now, perhaps unfortunately, perhaps fortunately, it has up to the present time proved impossible absolutely to demonstrate, for reasons which will presently appear, by any scientifically valid method of experimentation or reasoning, that any real portion of that totality of being which is an individual living man persists after he dies. Equally, for the same reasons, science cannot absolutely demonstrate that such persistence does not occur. The latter fact has had two important consequences. In the first place, it has permitted many millions of people to derive a real comfort of soul in sorrow, and a fairly abiding tran- quility of mind in general from the belief that immortality is a reality. Even the most cynical of scoffers can find lit- tle fault with such a result, the world and human nature being constituted as they are. The other consequence of science's present inability to lay bare, in final and irre- fragable terms, the truth about the course, if any, of events subsequent to death is more serious. It opens the way for recurring mental epidemics of that intimate mix- ture of hyper-credulity, hyper-knavery, and mysticism, which used to be called spiritualism, but now usually pre- fers more seductive titles. We are at the moment in the midst of perhaps the most violent and destructive epi- demic of this sort which has ever occurred. Its evil lies in the fact that in exact proportion to its virulence it des- troys the confidence of the collective mind of humanity in the enduring efficacy of the only thing which the history of mankind has demonstrated to contribute to the real advancement of his intellectual, physical, spiritual and moral well being, namely that orderly progression of ascertained knowledge which we now call science. THE PROBLEM 19 The reason why science finds itself helpless to pre- vent spiritualism's insidious sapping of the intellectual fiber of the race is because it is asked to prove a negative, upon the basis of unreal data. How difficult such a task is is obvious as it is proverbial. Until science has demon- strated that there is not a continuation of individual supernatural existence after natural death, the spiritual- ist can, and will, come forward with supposed demonstra- tions that there is such a continuation. But the most characteristic feature of science is its actuality, its reality, its naturality. Pearson has pointed out, in characteristi- cally clear and vigorous language, the reason why, in the minds of uninformed persons, science appears helpless in this situation. He says : Scientific ignorance may either arise from an insufficient classification of facts, or be due to the unreality of the facts with which science has been y v called upon to deal. Let us take, for example, fields of thought which were very prominent in medieval times, such as alchemy, astrology, witch- craft. In the fifteenth century nobody doubted the "facts" of astrology and witchcraft. Men were ignorant as to how the stars exerted their influence for good or ill; they did not know the exact mechanical process by which ail the milk in a village was turned blue by a witch. But for them it was nevertheless a fact that the stars did influence human lives, and a fact that the witch had the power of turning the milk blue. Have we solved the problems of astrology and witchcraft today? Do we now know how the stars influence human lives, or how witches turn milk blue? Not in the least. We have learnt to look upon the facts themselves as unreal, as vain imaginings of the untrained human mind; we have learnt that they could not be described scientifically because they involved notions which were in themselves contradictory and absurd. With alchemy the case was somewhat different. Here a false classification of real facts was combined with inconsistent sequences — that is, sequences not deduced by a rational method. So soon as science entered the field of alchemy with a true classification and a true method, alchemy was con- verted into chemistry and became an important branch of human knowl- edge. Now it will, I think, be found that the fields of inquiry, where science has not yet penetrated and where the scientist still confesses 20 THE BIOLOGY OF DEATH ignorance, are very like alchemy, astrology, and witchcraft of the Middle Ages. Either they involve facts which are in themselves unreal — con- ceptions which are self-contradictory and absurd, and therefore incapable of analysis by the scientific or any other method — or, on the other hand, our ignorance arises from an inadequate classification and a neglect of scientific method. This is the actual state of the case with those mental and spiritual phenomena which are said to lie outside the proper scope of science, or which appear to be disregarded by scientific men. No better example can be taken than the range of phenomena which are entitled Spiritualism. Here science is asked to analyse a series of facts which are to a great extent unreal, which arise from the vain imaginings of untrained minds and from atavistic tendencies to superstition. So far as the facts are of this character, no account can be given of them, because, like the witch's supernatural capacity, their unreality will be found at bottom to make them self-contradictory. Combined, however, with the unreal series of facts are probably others, connected with hypnotic and other conditions, which are real and only incomprehensible because there is as yet scarcely any intelligent classification or true application of scientific method. The former class of facts will, like astrology, never be reduced to law, but will one day be recognized as absurd; the other, like alchemy, may grow step by step into an important branch of science. Whenever, therefore, we are tempted to desert the scientific method of seeking truth, whenever the silence of science suggests that some other gateway must be sought to knowledge, let us inquire first whether the elements of the problem, of whose solution we are ignorant, may not after all, like the facts of witch- craft, arise from a superstition, and be self-contradictory and incompre- hensible because they are unreal. Let us recapitulate briefly our discussion to this point. Mankind has endeavored to prolong the individual life by natural and by supernatural means. This latter plan falls outside the present purview of the scientific method. The former is, in last analysis, responsible for a consid- erable part, at least, of the development of the science of biology, pure and applied, and the arts which found their operations upon it. Biology can and has contributed much to our knowledge of natural death and the causes which determine the duration of life. It is the purpose of this book to review some of the more important aspects THE PROBLEM 21 of this phase of biological science, and endeavor to set forth in an orderly and consistent manner the present state of knowledge of the subject. The problem of natural death has two aspects, one general, the other special. These may be stated in this way: 1. Why do living things die? What is the meaning of death in the general philosophy of biology? 2. Why do living things die ivhen they do? What factors determine the duration of life in general and in particular, and what is the relative influence of each of these factors in producing the observed result? Both of these problems have been the subject of much speculation and discussion. There has accumulated, especially in recent years, a considerable amount of new experimental and statistical data bearing upon them. I hope to be able in what follows to show that this new material, together with that which has for a long time been a part of the common store of biological knowledge, makes possible a clearer and more logically consistent picture than we have had of the meaning of death and the determination of longevity. Let us first examine in brief review the broad generalizations about death which have grown up in the course of the development of biology, and which may now be regarded as agreed to by practi- cally all biologists. BIOLOGICAL, GENEKALIZATIONS ABOUT NATUBAL DEATH The significant general facts which are known about natural death are these : (A). There is an enormous variation in the duration of life, both intra and inter-racially. Table I, which is adapted from various authorities, is to be read with the 22 THE BIOLOGY OF DEATH understanding that the figures are estimates, frequently based upon somewhat general and inexact evidence, and record extreme, though it is believed authentic instances. While the figures, on the accounts which have been men- tioned, are subject to large probable errors, the table does give a sufficiently reliable general picture of the truth to indicate the enormous differences which exist among different forms of animal life in respect of longevity. TABLE 1 Longevity of Animals Animal Approximate limits of maximum duration of life in different species Under 100 hours to ? Insects Under 100 hours to 17 years ? to 267 vpfl.rs Lower invertebrates Fish Amphibia Reptiles Birds Mammals ? to 267 years ? to 36 years ? to 175 years 9 years to 1 18 years 1/^2 years to over 100 years We see from this table that life may endure in differ- ent forms from only the briefest period, measured in hours as in the case of Ephemeridae, to somewhere in the hundreds of years. The extremely long durations are of course to be looked upon with caution and reserva- tion, but if we accept only extreme cases of known dura- tion of life in man, the range of variation in this characteristic of living things is sufficiently wide. It is probable that man, in exceptional instances, is nearly the longest lived of all mammals. The common idea that whales and elephants attain great longevity appears to be not well founded. The absolutely authentic instances of human survival beyond a century are, con- trary to the prevalent view and customary statistics, extremely rare. The most painstaking and accurate THE PROBLEM 23 investigation of the frequency of occurrence of centen- arians which has ever been made is that of T. E. Young. Because of the considerable intrinsic interest of the matter, and the popular misconceptions which generally prevail about it, it will be worth while to take a little time to examine Young's methods and results. He points out in the beginning that the evidence of great age which is usually accepted by census officials, by registrars of death, by newspaper reporters, and by the general public, is, generally speaking, of no validity or trustworthiness whatever. Statements of the person concerned, or of that person's relatives or friends, as to extreme longevity, can almost invariably be shown by even a little investiga- tion to be extremely unreliable. To be acceptable as scientific evidence any statement of great age must be supported by unimpeachable documentary proof of at least the following points: a. The date of birth, or of baptism. b. The date of death. c. The identity of the person dying at a supposed very advanced age with the person for whom the birth or baptismal record, upon which the claim of great age is based, was made out. d. In the case particularly of married women the date of marriage, the person to whom married, and any other data which will help to establish proof of identity. In presumptive cases of great longevity, which on other grounds are worthy of serious consideration, it is usually in respect of item c — the proof of identity — that the evidence is weakest. Every student of genealogical data knows how easy it is for the following sort of thing to happen. John Smith was born in the latter half of the eighteenth century. His baptism was duly and pro- perly registered. He unfortunately died at the age of 24 BIOLOGY OF DEATH say 15. By an oversight his death was not registered. In the same year that he died another male child was born to the same parents, and given the name of John Smith, in commemoration perhaps of his deceased brother. This second John Smith was never baptized. He at- tained the age of 85 years, and then because of the appear- ance of extreme senility which he presented* his stated age increased by leaps and bounds. A study of the baptismal records of the town disclosed the apparent fact that he was just 100 years old. The case goes out to the public as an unusually well authenticated case of centenarianism, when of course it is nothing of the sort. Young applies vigorously the criteria above enumer- ated first, to the historically recorded cases of great long- evity such as Thomas Parr, et id genus omne, and rejects them all; and second to the total mortality experience of all the Life Assurance and Annuity Societies of Great Britain and the annuity experience of the National Debt Office. The number of persons included in the experience was close upon a million. He found in this material, and from other outside evidence, exactly 30 persons who lived 100 or more years. In Table 2 the detailed results of his inquiry are shown in condensed form. It will be noted from this table that the most extreme case of longevity which Young was able to authenticate was about a month and a half short of 111 years. Of the 30 centenarians recorded 21 were women and 9 were men. The superiority of women in expectation of life is strikingly apparent at the very high age of 100 years. We shall later see that this is merely a particularly noteworthy instance of a phenomenon which is common to a great- portion of the life span. THE PROBLEM 25 The contrast between these proved findings of Young, exceedingly modest both in respect of numbers, and ex- tremity of longevity, and the loose data on centenarianism TABLE 2 Authentic Instances of Centenarianism (from Young] Sex Social status (single or married) Age at death (or living) Years Months Days 9 M 110 • • 321 9 M 108 • • 144 9 M 105 8 • • • 9 S 104 9 16 9 M 103 9 28 9 ? 103 • • 269 9 M 103 3 7 13 O) o » cr CONDITIONS OF CELLULAR IMMORTALITY 53 sion and further development. The latter processes may be regarded as inhibiting or modifying the mortal pro- cess. Loeb and Lewis' undertook experiments, based upon this view, to see whether it would be possible by chemical treatment of the egg to prolong its life. Since in general specific life phenomena are perhaps, on the chemical side, chiefly catalytic phenomena, it was held to be reasonable that if some substance could be brought to act on the egg, which would inhibit such phenomena without permanently altering the constitution of the living material, the life of the cell should be considerably prolonged. The first agent chosen for trial was potassium cyanide, KCN. It was known that this substance weakened or inhibited entirely a number of enzymatic processes in living material, without materially or permanently alter- ing its structure. It was found that, normally, the unfertilized egg of the sea-urchin would live in sea-water at room temperature, and maintain itself in condition for successful fertiliza- tion and development, up to a period of about twenty-three hours. After that time the eggs began to weaken. Either they could not be successfully fertilized, or if they were fertilized, development only went on for a short time. After 32 hours, the eggs could not, as a rule, be fertilized at all. The experiment was then tried of adding to the sea-water, in which the unfertilized eggs were kept, small amounts of KCN in a graded series, and then exam- ining the results of fertilizations undertaken after a stay of the unfertilized eggs of 75 hours in the solution. It will be noted that this period of 75 hours is more than three times the normal duration of life of the cell in normal sea-water. The results of this experiment are shown in summary form in Table 4. 54 BIOLOGY OF DEATH TABLE 4 Experiments of Loeb and Lewis on the Prolongation of Life of the Sea-urchin Egg by KCN Concentration of KCN Pure sea-water n/64000 KCN n/16000 KCN n/8000 KCN n/4000 KCN n/2000 KCN n/1000 KCN n/750 KCN n/400 n/300 n/250 n/200 n/100 KCN KCN KCN KCN KCN Result of fertilization after a 75 hours' stay in the solution No egg segments No egg segments No egg segments Very few eggs show a beginning of seg- mentation Very few eggs show a beginning of seg- mentation Few eggs go through the early stages of segmentation Many eggs segment and develop into swim- ming larvae Many eggs segment and develop into swim- ming larvae A few eggs develop into swimming larvae No egg segments No egg segments No egg segments No egg segments From this table it is seen that in concentrations of KCN from n/750 to n/1000 the eggs developed perfectly into swimming larvae. In other words, by the addition of this very small amount of KCN, the life period has been prolonged to three times what it would normally be under the same environmental conditions. Concen- trations of KCN weaker than n/1000 were incapable of producing this result, or at best, if development started, the process came very quickly to an end. In stronger concentrations than n/400 the eggs were evidently poi- soned, and no development occurred. Other experiments of Loeb's show that the lethal effects of various toxic agents upon the egg cell may be inhibited or postponed for a relatively long time, by CONDITIONS OF CELLULAR IMMORTALITY 55 suitable chemical treatment, such as lack of oxygen, KCN, or chloral hydrate. A typical experiment of this kind made upon the sea-urchin, Strongylocentrotus purpuratus may be quoted: Eggs were fertilized with sperm and put eleven minutes later into three flasks, each of which contained 100 c. o. of sea-water + 16 c. c. 2-y2 m CaCl2. One flask was in contact with air, while the other two flasks were connected with a hydrogen generator. The air was driven out from these two flasks before the beginning of the experiment. The eggs were transferred from one of these flasks after four hours and fourteen minutes, from the second flask after five hours and twenty-nine minutes, into normal (aerated) sea-water. The eggs that had been in the hypertonic sea-water exposed to air were transferred simultaneously with the others into separate dishes with aerated normal sea-water. The result was most striking. Those eggs that had been in the hypertonic sea-water with air were all completely disintegrated by " black cytolysis." Ten per cent, of the eggs had been transformed into "shadows" (white cytolysis). It goes without saying that all the eggs that had been in the aerated hyper- tonic sea-water five and a half hours were also dead. The eggs that had been in the same solution in the absence of oxygen appeared all normal when they were taken out of the solution, and three hours later — the temperature was only 15°C. — they were all, without exception in a per- fectly normal two- or four-cell stage. The further development was also in most cases normal. They swam as larvae at the surface of the vessel and went on the third day (at the right time) into a perfectly normal pluteus stage, after which their observation was discontinued. Of the eggs that had been five and a half hours in the hypertonic sea-water deprived of oxygen, about 90 per cent, segmented. Let us consider one more illustration from Loeb's work in this field. Normally, in the forms with which he chiefly worked, sea-urchin, starfish, and certain mol- luscs, an absolutely essential condition for the continua- tion of life of the germ-cells after they are discharged from the body is that two cells, the ovum and the sper- matozoon, shall unite in normal fertilization. Put in another way, parthenogenesis does not normally occur in these forms. Fertilization is an essential condition for the continuation of life and development. But Loeb's 56 BIOLOGY OF DEATH painstaking and brilliant researches, extending over a number of years, show that when we say that fertilization is an essential condition for the continued life of the germ-cells outside the body, our language tends to ob- scure the most important fact, which is simply that for the continuation of life in these cells only certain internal physico-chemical conditions and adjustments' must be realized. It makes no essential difference to the result whether these conditions are realized through the intervention of the sperm, as in normal fertiliza- tion, or by purely artificial chemical methods initiated, controlled and directed at every step by human agency. We can, in other words, regard all cases of suc- cessful artificial parthenogenesis as fundamentally a con- tribution to the physiology of natural death, and a demon- stration of its essentially mechanistic basis. The condi- tions of continued existence are physical and chemical, and controllable as such. The methods finally worked out as optimum afford a complete demonstration of the thesis we have just stated. Thus, for example, the unfertilized egg of the sea-urchin, Strongylocentrotus purpuratus, will continue in life and develop perfectly normally if it is subjected to the following treatment: The eggs are first placed in sea-water to which a definite amount of weak solution of butyric acid has been added (50 cc. of sea-water + 2.8 c.c. n/10 butyric acid). In this solution at 15° C. the eggs are allowed to remain from l1/^ to 3 or 4 minutes. They are then transferred to normal sea- water, in which they remain from 15 to 20 minutes. They are then transferred for 30 to 60 minutes at 15° C. to sea-water which has had its osmotic pressure raised by the addition of some salts (50 c.c. of sea-water+8 c.c. of 2y2 m NaCl, or 2y2 m NaCl+KCL+CaCl2 in the proportion in which these CONDITIONS OF CELLULAR IMMORTALITY 57 salts exist in sea-water). After the stay of from 30 to 60 minutes in this solution, the eggs are transferred back to normal sea-water, the transferring being in batches at intervals of 3 to 5 minutes between each batch transferred. It is then found that those eggs which have been just the right length of time in the hypertonic sea-water develop into perfectly normal sea-urchin larvae. In other words, we have here a definite and known physico-chemical pro- cess completely replacing what was, before this work, universally regarded as a peculiarly vital process of extraordinary complexity,, "probably beyond power of human control. These three examples from Loeb's work on the sub- ject of prolongation of life in the egg cell will suffice for our present purposes. The lesson which they teach is plain, and is one which has, as will be readily perceived, a most important bearing upon the general concept of life and death outlined in the preceding chapter. The experiments demonstrate that the conditions essential to continued life of the germ-cells outside the body are phy- sico-chemical conditions, and that when these cells die it is because the normal physico-chemical machinery for the continuation of life has either broken down, or has not been given the proper activating chemical conditions. Lack of space alone prevents going in detail into an- other extremely interesting and important development of this subject, due to Dr. Frank R. Lillie of the Univer- sity of Chicago. He has, in recent years made a thorough analysis of the biological factors operating when the egg of the sea-urchin is normally fertilized by a spermato- zoon. The conception of the process of fertilization to which Lillie comes is ' ' that a substance borne by the egg (fertilizin) exerts two kinds of actions: (1) an agglutin- 58 BIOLOGY OF DEATH ating action on the spermatozoon and (2) an activating action on the egg. In other words, the spermatozoon is conceived, bv means of a substance which it bears and / %r which enters into union with the f ertilizin of the egg, to release the activity of this substance within the egg.' From the standpoint of the present discussion it is ob- vious that Lillie's results so far present nothing which in any way disturbs the conclusion we have reached as to the essentially physico-chemical nature of the processes which condition the continuation of life and development of the egg. TISSUE CULTURE IN VITEO Let us turn now to another question. Are the germ- cells the only cells of the metazoan body which possess the characteristic of potential immortality? There is now an abundance of evidence that such is not the case, but that, on the contrary, there are a number of cells and tissues of the body, which, under appropriate conditions, may continue living indefinitely, except for the purely accidental intervention of lethal circumstances. Every child knows that all the tissues do not die at the same time. It is proverbial that the tail of the snake, whose head and body have been battered and crushed until even the small boy is willing to admit that the job of killing is complete, ' ' will not die until the sun goes down. ' ' Galvani 'si famous experiment with the frog's legs only succeeded because some parts survive after the death of the organism as a whole. As Harrison points out " Almost the whole of our knowledge of muscle-nerve physiology, and much of that of the action of the heart, is based upon experiments with surviving organs ; and in surgery, where we have to do with changes involved in the repair of injured parts, -, ] • FIG. 13. — Piece of tissue from frog embryo cultivated in lymph, two days old. The dark portion shows original bit of tissue. Lighter portions are new growth (From Harrison.) V FIG. 14. — Group of nerve fibers which have grown from an isolated piece of neural tube of a chick embryo. (From Harrison after Burrows.) CONDITIONS OF CELLULAR IMMORTALITY 59 including processes of growth and differentiation, the power of survival of tissues and organs and their trans- plantability to strange regions, even to other individuals, has long formed the basis of practical procedures. " The first successful cultures of somatic cells and tis- sues outside the body were those of Leo Loeb, described in 1897. His first method consisted in cultivating the tissues in appropriate media in test tubes. Later he used also another method, which involved the transplantation of the solid medium and the tissue into the body of an- other animal. What has been regarded as a defect of both these methods is that they do not permit the contin- ued observation of the cells of the growing cultured tissue. To Harrison is due the development of a method which does permit such study. In 1907 he announced the dis- covery that if pieces of the developing nervous system of a frog embryo were removed from the body with fine needles, under strictly aseptic precautions, placed on a sterile cover slip in a drop of frog lymph, and the cover slip then inverted over a hollow glass slide, that the tis- sues would remain alive for many days, grow and exhibit remarkable transformations. By tin's technique it was possible to study the changes with a high power micro- scope and photograph them. Figure 13 is a general view of one of these tissue cul- tures two days old. It shows a piece of nervous tissue from the frog embryo, with cells growing out from it into the lymph. The lighter portions are the new cells. In his remarkable monograph Harrison shows nerve cells developing fibers at first thickened, but presently becoming of normal character and size. At the ends are pseudopodial processes, by which the growing fiber at- taches itself to the cover slip or other solid bodies. Fig- 60 BIOLOGY OF DEATH lire 14 shows a particularly beautiful nerve fiber prepar- ation made by Burrows. The fibers grew from a preparation of the embryonic nervous system of the chick. There can be no doubt, as these figures so clearly show, of the life of these cells outside the body, or of the normality of their develop- mental and growth processes. Under the guidance of Harrison, another worker, Burrows, improved the technique of the cultivation of tissues outside the body, first by using plasma from the blood instead of lymph and later in various other ways. He devised an apparatus for affording the tissue culture a continuous supply of fresh nutrient medium. There is in this apparatus a large culture chamber which takes the place of the plain hanging drop in an hermetically sealed cell. On the top of this culture chamber there is a wick, which carries the culture fluid from a supplying chamber and discharges it into a receiving chamber. The tissue is planted among the fibers of the wick, which are pulled apart where it crosses the top of the chamber. The whole system is kept sterile and so arranged that the growing tissue can be kept under observation with high powers of the microscope. The nutrient medium may be modified at will, and the effects of known sub- stances upon the cellular activities of every sort may be studied. Burrows began his investigations in this field on the tissues of the embryo chick. With the success of these cultures was established the fact that the tissues of a warm blooded animal were as capable of life, develop- ment, and growth outside the body as were those of cold- blooded animals, such as the frog. Burrows succeeded in cultivating outside the body, cells of the central nervous ' i . .'• ' •'I,' ',' ,.->".'•>.•.'••. . ' . -; W\ ' . .. ••. A- ' .' $£*& ' • " ;'iw . ^^Mi^^&^^^S-- a^A--*f-»!?&fc--CT*^T''-r--^r>-- * .«- -• . • * .-* - > fef s^^cr". •>*!-*- -xX-r^^ •••„ ;&&*>&£ ' ~ "• rtr**^-.--^.^*^!7^-^ ~r-_-~^-.T^3*w>-^'- vj '. . -" - -^ ;^ ,/'?v, /--'/. f-v \ ^ •• r . • FIG. 15. — Human connective tissue cells fixed and stained with Gienisa stain. The culture was made by extirpating the central portion of culture 285 in its 16th passage, washing the remaining portion of the culture with Ringer solution without removing it from the cover-glass, and drop- ping on Jresh plasm and extract. The preparation shows the extent of growth obtained in 4S hours from peripheral cells remaining after extirpation of the fragment. (After Losee and Ebeling.) CONDITIONS OF CELLULAR IMMORTALITY 61 system, the heart, and mesenchymatous tissue of the chick embryo. At the same time Carrel was carrying on studies in this same direction at the Rockefeller In- stitute. In his laboratory were made the first successful cultures in vitro of the adult tissues of mammals. He developed a method of culture on a plate which permitted the growing of large quantities of material. He found that almost all the adult and embryonic tissues of dog, cat, chicken, rat, guinea pig, and man could be cultivated in vitro. Figure 15 shqws a culture of human tissue, made at the Rockefeller Institute. I am indebted to Doctor Carrel and Doctor Ebeling for permission to pre- sent this photograph here. According to the nature of the tissues cultivated, con- nective or epithelial cells were generated, which grew out into the plasma medium in continuous layers or radiating chains. Not only could normal tissues be cultivated but also the cells of pathological growths (cancer1 cells). It has been repeatedly demonstrated that normal cell division takes place in these tissues cultivated outside the body. The complex process of cell division, which is technically called mitosis, has been rightly regarded as one of the most characteristic, because complicated and unique, phenomena of normal life processes. Yet this process occurs with perfect normality in cells cultivated outside the body. Tissues from various organs of the body have been successfully cultivated,, including the kidney, the spleen, the thyroid gland, etc. Burrows was even able to demonstrate that the isolated heart muscle cells of the chick embryo can divide as well as differen- tiate, and beat rhythmically in the culture medium. Perhaps even more remarkable than the occurrence of such physiological activity as that of the heart muscle 62 BIOLOGY OF DEATH cells in vitro is the fact that in certain lower forms of life a small bit of tissue or even a single cell, may develop in culture into a whole organism, demonstrating that the capacity of morphogenesis is retained in these isolated somatic cells. H. V. Wilson has shown that in coelenter- ates and sponges complete new individuals may develop in vitro from isolated cells taken from adult animals. By squeezing small bits of these animals through bolting cloth he was able to separate small groups of cells or even single cells. In culture these would grow into small masses of cells which would then differentiate slowly into the normal form of the complete organism. Figure 16 shows an example of this taken from Wilson's work. It was early demonstrated by Carrel and Burrows that the life of the tissues in vitro, which varied in differ- ent experiments from 5 to 20 days, could be prolonged by a process of successive transfers of the culture to an indefinite period. Cells which were nearing the end of their life and growth in one culture need only be trans- ferred to a new culture medium to keep on growing and multiplying. Dr. and Mrs. Warren H. Lewis made the important discovery that tissues of the chick embryo could be cultivated outside the body in purely inorganic solutions, such as sodium chloride, Binger's solution, Locke's solution, etc. No growth in these inorganic cul- tures took place without sodium chloride. Growth was prolonged and increased by adding calcium and potas- sium. If maltose or dextrose, or protein cleavage pro- ducts were added proliferation of the cells increased. By the method of transfer to fresh nutrient media, Carrel has been able to keep cultures of tissue from the heart of the chick embryo alive for a long period of years. In a letter, recently received, he says: "The FIG. 16. — Pennaria. Restitution mass six days old, completely metamorphosed, with developed hydranths. Op. perisarc ot original mass; x, perisarc of outgrowth adherent to glass. (From Wilson.) FIG. 17. — Culture of old strain of connective tissue. 1614 passage. 8 years and 8 months old, lacking 2 days. 48 hours' growth. x20. (Ebeling). CONDITIONS OF CELLULAR IMMORTALITY 63 strain of connective tissue obtained from a piece of chick heart is still alive, and will be nine years old the seven- teenth of January, 1921." Figure 17 is a photograph showing the present condition of this culture. It should be understood that this long continued culture has gone on at body temperature in an incubator, and not by keep- ing the culture at a low temperature and merely slowing down the vital processes. This is indeed a remarkable result. It completes the demonstration of the potential immortality of somatic cells, when removed from the body to conditions which permit of their continued existence. Somatic cells have lived and are still living outside the body for a far longer time than the normal duration of life of the species from which they came. I think the present extent of Carrel's cultures in time fully disposes of Harrison's criticism to the effect that we are "not justified in referring to the cells as potentially immortal or even in speaking of the prolongation of life by artificial means, at least not until we are able to keep the cellular elements alive in cultures for a period exceeding the duration of life of the organism from which they are taken. There is at present no reason to suppose this cannot be done, but it simply has not been done as yet." I have had many years ' experience with the domestic fowl, and have par- ticularly studied its normal duration of life, and discus- sed the matter with competent observers of poultry. I am quite sure that for most breeds of domestic poultry the normal average expectation of life at birth is not substantially more than two years. For the longest lived races we know this normal average expectation of life cannot be over four years. I have never been able to keep a Barred Plymouth Rock alive more than seven 64 BIOLOGY OF DEATH years. There are on record instances of fowls living to as many as 20 years of age. But these are wholly excep- tional instances, unquestionably far rarer than the occur- rence of centenarians among human beings. There can be no question that the nine years of life of Carrel's culture has removed whatever validity may have origin- ally inhered in Harrison's point. And further the cul- ture is just as vigorous in its grqwth today as it ever was, and gives every indication of being able to go on indefinitely, for 20 or 40, or any desired number of years. The potential immortality of somatic cells has been logically just as fully demonstrated in another way as it has by these tissue cultures. Some nineteen years ago, Leo Loeb first announced the important discovery that potential immortality of somatic cells could be demon- strated through tumor transplantations. His latest sum- mary of this work may well be quoted here : "We must remember that common, transplantable tumors are the direct descendants of ordinary tissue cells, such as we normally find in the individuals of the particular species which we use. The tumors may be derived from a variety of normal tissues and, in general, the transfor- mation from normal cells into tumor cells takes place under the influence of a long continued action of various factors enhancing growth. Tumor cells are, therefore, merely somatic cells which have gained an increased growth energy and at the same time somehow gained, in some cases, the power to escape the destructive consequences of homoiotoxins. This ability of cer- tain tumors to grow in other individuals of the same species has enabled us to prove, through apparently endless propagation of these tumor cells in other individuals, that ordinary somatic cells possess potential im- mortality in the same sense in which protozoa and germ cells possess immortality. Thus tumor transplantation made possible the establishment of a fact of great biological interest, which, because of the homoiosensitive- ness of normal tissues, could not be shown in the latter. "We wish, however, especially to emphasize the fact that our experi- ments did not merely prove the immortality of tumor cells, but of the ordinary tissue cells as well, the large majority or all of which can be transformed into tumor cells. At an early stage of our investigations CONDITIONS OF CELLULAR IMMORTALITY 65 we drew, therefore, on the basis of these experiments, the conclusion that ordinary tissue cells are potentially immortal; notwithstanding the fact that, especially under Weismann's influence, the opposite view had been generally accepted, and as it seems to us, with full justification, inasmuch as no facts were known at that time which suggested the immortality of somatic cells. It was the apparently endless transplantation of tumor cells which proved the contrary view. "To recapitulate what we stated above : tumors are merely transformed tissue cells. All or the large majority of adult tissues are potential tumor cells. Tumor cells have been shown experimentally to be potentially im- mortal, therefore tissue cells are potentially immortal. " This wider conclusion I expressed nineteen years ago. Quite recently, the immortality of certain connective tissue cells has been demonstrated by Carrel through in vitro culture of these cells. Under those conditions the tissue cells escape the mechanisms of attack to which the homoiotoxins expose the ordinary tissue cells in other individuals of the same species. Under these conditions the reactions of the host tissue ag;ainst homoiotoxins which would have taken place in vivo, are eliminated. We must, however, keep in mind that this method of proving the immortality of somatic cells applies only to one particular, very favorable kind of cells; and it is very doubtful, if, by cultivation in vitro, the same proof could be equally well supplied in the case of other tissues. On the basis of tumor transplanta- tions, on the contrary, we were able to show that a considerable variety, perhaps the large majority of all tissue cells possess potential immortality." To Loeb unquestionably belongs the credit for first perceiving that death was not a necessary inherent con- sequence of life in the somatic cell, and demonstrating by actual experiments that somatic cells could, under cer- tain conditions, go on living indefinitely. Before turning to the next phase of our discussion, let us summarize the ground we have covered up to this point. We have seen that by appropriate control of conditions, it is possible to prolong the life of cells and tissues far beyond the limits of longevity to which they would attain if they remained in the multicellular body from which they came. This is true of a wide variety of cells and tissues differentiated in various ways. In- deed, the range of facts which have been ascertained 5 66 BIOLOGY OF DEATH by experimental work in this field, probably warrants the conclusion that this potential longevity inheres in most of the different kinds of cells of the metazoan body, except those which are extremely differentiated for par- ticular functions. To bring this potential immortality to actuality requires, of course, special conditions in each particular case. Many of these special conditions have already been discovered for particular tissues and particular animals. Doubtless, in the future many more will be worked out. We have furthermore seen that in certain cases the physico-chemical nature of the condi- tions necessary to insure the continuance of life has been definitely worked out and is well understood. Again this warrants the expectation that, with more extended and penetrating investigations in a field of research which is really just at its beginning, we shall understand the physics and chemistry of prolongation of life of cells and tissues in a great many cases where now we know nothing about it. One further point and we shall have done with this phase of our discussion. The experimental culture of cells and tissues in vitro has now covered practically all the essential tissue elements of the metazoan body, even including the most highly differentiated of those tissues. Nerve cells, muscle cells, heart muscle cells, spleen cells, connective tissue cells, epithelial cells from various loca- tions in the body, kidney cells, and others have all been successfully cultivated in vitro. We may fairly say, I be- lieve, that the potential immortality of all essential cel- lular elements of the body either has been fully demonstrated, or else has been carried far enough to make the probability very great that properly conducted experiments would demonstrate the continuance of the CONDITIONS OF CELLULAR IMMORTALITY 67 life of these cells in culture to any definite extent. It is not to be expected, of course, that such tissues as hair, or nails, would be capable of independent life, but these are essentially unimportant tissues in the animal econ- omy as compared with those of the heart, the nervous system, the kidneys, etc. What I am leading to is the broad generalization, perhaps not completely demon- strated yet, but having regard to Leo Loeb's work, so near it as to make little risk inhere in predicting the final outcome, that all the essential tissues of the meta- zoan body are potentially immortal. The reason that they are not actually immortal, and that multicellular animals do not live forever, is that in the differentiation and specialization of function of cells and tissues in the body as a whole, any individual part does not find the conditions necessary for its continued existence. In the body any part is dependent for the necessities of its existence, as for example nutritive material, upon other parts, or put in another way, upon the organization of the body as a whole. It is the differentiation and spe- cialization of function of the mutually dependent aggre- gate of cells and tissues which constitute the metazoan body ivhich brings about death, and not any inherent or inevitable mortal process in the individual cells them- selves. An examination of different lines of evidence has led us to two general conclusions, viz : a. That the individual cells and tissues of the body, in and by themselves, are potentially immortal. b. That death of the metazoan body occurs, funda- mentally, because of the way in which the cells and tis- sues are organized into a mutually dependent system. Is there any further and direct evidence to be had 68 BIOLOGY OF DEATH upon the second of these conclusions? So far our evi- dence in its favor has been indirect and inferential, though cogent so far as it goes. In this connection, a paper of Friedenthal's is of considerable interest. He shows that there is a marked correspondence between the longevity of various species of animals and a constant of organization which he calls the ' ' cephalisation factor. ' ' This cephalisation factor in pure form, in his sense, is given by the equation. „ , ,. ,. e Brain weight Cephalisation factor = „ . 2 — -t Total mass of body protoplasm. Now " total mass of body protoplasm," as distinct from supporting structures, such as bone etc., is obviously difficult to determine directly. But Friedenthal is well convinced that, to a first approximation, the cephalisa- tion factor may be written in this way: -, , ,. ,. . Brain weight Cephalisation factor = -r^— -, ?__ 2A (Body weight) ' Computed upon the latter basis he sets up tables of the relation between cephalisation factor and longevity for mammals and for birds. It is not necessary to repro- duce here the long tables, but to show the general point, the following table for five selected species of mammals will suffice: TABLE 5 Relation between the cephalisation factor and longevity (Friedenthal) Species Cephalisation index Duration of life Mouse 0.045 6 years Rabbit .066 8 years Marmoset (Callithrix) .216 12 years Deer .35 15 years Man 2.7 100 years There appears in this short selected table a defect CONDITIONS OF CELLULAR IMMORTALITY 69 which is even more apparent in his long ones, namely, that the figures for duration of life are distinctly round numbers. There is no evidence, for example, that the normal life span of the mouse is 6 years. All who have statistically studied the matter agree upon a much smal- ler figure than this. But, leaving this point aside, it is apparent that there is a parallelism of striking sort be- tween the cephalisation factor and duration of life. In other words, it appears that the manner in which higher vertebrates, at least, are put together in respect of the proportionality of brain and body is markedly associated with the duration of life. It would be a matter of great interest to see whether this correlation between relative brain-weight and the expectation of life holds intra- racially as well as it does inter-racially. The bearing of these results of FriedenthaPs upon our results as to the distribution of mortality upon a germ-layer basis, to be discussed in Chapter V infra, is obvious. Another possible illustration of the general point now under discussion may be found in some recent work of Robertson and Ray. These authors, in a recent paper, have analyzed the growth curves of relatively long-lived mice as compared with the curves shown by relatively short-lived individuals. In the experiment both groups were subjected to the same kind of experimental treat- ment of various sorts, and the care with which the experi- ments were conducted in respect of control of the environmental factors renders the results highly inter- esting and valuable. The long-lived animals form a group which grows more rapidly in early life, and at the same time is less variable than the short lived group. The short-lived animals often grow much more rapidly in later life than the long-lived, but this accretion of tissue 70 BIOLOGY OF DEATH was found to be relatively unstable.. They further found that the long-lived animals represent a relatively stable group, highly resistant to external disturbing factors, and showing a more or less marked but not invariable tendency to early overgrowth and relative paucity of tissue accretion in late life. The short-lived animals are on the contrary relatively unstable, sensitive to external disturbing factors, and, as a rule, but not invariably, dis- play relatively deficient early growth and a tendency to rapid accretion of tissue in later life. In interpreting these results, Robertson and Ray be- lieve that the differences are based upon the fact that in early or embryonic life the outstanding characteristic of the tissues is a high proportion of cellular elements, whereas in old age there is a marked increase in connective tissues. They further point out that connective tissue elements are ultimately dependent upon cellular tissues for their support, and that the connective tissues are expensive to maintain. They believe that the reason that the substance tethelin (cf. Chap. VII infra) prolongs life is because it accelerates the metabolism of the cellular elements to the detriment of the connective tissue ele- ments. Longevity on this view is determined not by the absolute mass of living substance, but by the relative proportions of parenchymatous to sclerous tissues. SENESCENCE The facts presented in this and the preceding chapter clearly make it necessary to review with some care the current conception of senescence. Senescence, or grow- ing old, is commonly considered to be the necessary prel- ude to "natural," as distinguished from accidental death. CONDITIONS OF CELLULAR IMMORTALITY 71 But is the evidence really sound and complete that such is the fact ? A careful and unprejudiced examination will inevi- tably suggest to the open mind, I think, that much of the existing literature on senescence is really of no funda- mental importance, because it has unwittingly reversed the true sequential order of the causal nexus. If cells of nearly every sort are capable, under appropriate con- ditions, of living indefinitely in undiminished vigor, and cytological normality, there is little ground for postu- lating that the observed senescent changes in these cells while in the body, such as those described by Minot and others, are expressive of specific and inherent mortal processes going on in the cells ; or that these cellular pro- cesses are the cause of senescence, as Minot has concluded. That there is such a phenomenon as senescence is, of course, certain. It is observable both in Protozoa and in Metazoa. The real question, however, is a twofold one, viz: (a) is senescence in either Protozoa or Metazoa an inevitable consequence of the strain or the individual having lived; and (b) is senescence a necessary asso- ciate and forerunner of natural death? Let us briefly reconsider the facts. In Protozoa a slowing down of the division rate in culture has been frequently observed; and it has been held, first, that this is a phenomenon essentially homologous to senes- cence in the metazoan; and second, that if nuclear reorganization, by the way either of endomixis or of conjugation, did not occur that the strain would die out. Indeed, Jennings, in discussing the matter in his last book says: "Thus it appears that in these organisms nature has employed the method of keeping on hand a reserve stock of a material essential to 72 BIOLOGY OF DEATH life; by replacing at intervals the worn out material with this reserve, the animals are kept in a state of perpetual vigor; not, as individuals, growing old or dying a natural death. Nevertheless, a wearing out pro- cess, such as might be called getting old, does occur in the structures employed in the active functions of life, and these must be replaced after a time of service. So far as the conditions in these organisms are typical, deterioration and death do appear to be a consequence of full and active life; life carries within itself the seeds of death. It is not mating with another individual that avoids this end; but replacement of the worn material by a reserve The great mass of cells subject to death in the higher animals dwindles in the infusorian to the macronucleus ; this alone represents a corpse. But the dissolution of this corpse occurs within the living body. It much resembles such a process as the wasting away and destruction of minute parts of our own bodies, which we know is taking place at all times, and which does not interrupt the life of the individual." It is doubtful if this position is warranted. Since Jennings wrote the statement quoted, some new and pertinent data have appeared in regard to amicronu- cleate infusoria. Woodruff and his co-workers have shown that such races may occur rather commonly. Thus Woodruff, in 1921, says: "During the past year, the isolation for certain experiments of 14 "wild" lines representing 6 species of hypotrichous ciliates revealed 7 lines (4 species) with micronuclei and 7 lines (2 species) without morphological micronuclei. Ten of the lines were all isolated from a "wild" mass culture of the same species Urostyla grandis, found in a laboratory aquarium. Six of these lines were amicronucleate. All of the lines of all of the species have bred true with respect to the character in question, and one amicronucleate line at present is at the 102d generation. Similarly a culture of Paramecium caudatum, which the present writer supplied a year ago to a course in protozoology for the study of the nucleus, failed to reveal a micronucleus, although in other races the micronucleus was readily demonstrated." Now, since it is the micronucleus which furnishes for the process of endomixis the "reserve stock of a material essential to life" which Jennings discusses, it is plain that the existence of amicronucleate races of Protozoa CONDITIONS OF CELLULAR IMMORTALITY 73 at once puts a new face upon the whole matter. Dawson has studied in continued culture one of these amicronu- cleate races of Oxytricha Jiymenostoma Stokes. His con- clusion is as follows : "The existence of a form which not only apparently may live indefi- nitely without conjugation, autogamy, or endomixis (assuming the possi- bility of the latter phenomenon in an hypotrichous form), but also apparently does not possess the ability to undergo any of these phenomena, brings to light an entirely new possibility in the life history of ciliates. It has been proved quite conclusively, (Woodruff, '14), that in forms which ordinarily conjugate, the continued prevention of this process brings about no loss of viability if a favorable environment be provided. How- ever, in the organism under consideration there is apparently no possi- bility not only of conjugation or endomixis, but also of autogamy; and thus we have from another source crucial evidence that none of these phenomena is an indispensable factor in the life-history of this hypo- trichous form." In the light of these clean cut and definite results one is more disposed than was formerly the case to accept at their face value the results of Enriques with Glaucoma pyriformis, and those of Hartmann with Eudorina elegant, in which reproduction went on indef- initely with undiminished vigor and no evidence of any process comparable to endomixis. Altogether, it seems to me that the weight of the evi- dence now is that in the Protozoa, senescence (or death), is not a necessary or inevitable consequence of life. Given the appropriate and necessary conditions of envi- ronment, true immortality — the absence of both senes- cence and natural death, each defined in the most critical manner — is in fact the reality for a number of forms. Turning to the metazoan side of the case, the evidence regarding senescence is equally cogent. In the first place, in the longest continued in vitro tissue cultures known (those of Carrel) there is, as already stated, no appear- 74 BIOLOGY OF DEATH ance of senescence in the cells. But it may be objected that an element of uncertainty is injected into the case, by the fact that, as Carrel and Ebeling have lately dis- cussed in some detail, it has been necessary in carrying along this long-continued culture to add regularly to the culture medium a small amount of "embryonic juice/' One might urge that, but for the * ' embryonic juice, ' ' cellu- lar senescence and death would have appeared. But suppose this to be granted fully. It does not mean that senescence is a necessary and inevitable consequence of life, but only that to realize a potential immortality the cells must have an appropriate environment, one element of which is presumably some chemical combination which, so far, one has supplied only through "embry- onic juice.' An entirely different sort of evidence and one of great significance is found in the facts of clonal propaga- tion of plants, well known to horticulturists. An individ- ual apple tree grows old, and eventually dies, as a tree. But at all periods of its life, including all stages of senescence up to the terminal one, death, it produces shoots each spring. If one of these shoots is grafted to another root, it will, in the passage of time, make first a young tree, then a middle aged tree, and finally an old, senescent tree; which, in turn, will make new shoots, which may, in turn, be grafted to new roots, and so on ad infinitum. It is not even absolutely necessary that the shoot be grafted to a new root; though, of course, this is the manner in which the great majority of our orchards are, in fact, propagated, and have been since the beginning of horticultural history. Anyone who is familiar with the woods of New England, not too far from settlements, has seen apple trees in the woods where a CONDITIONS OF CELLULAR IMMORTALITY 75 shoot, whose continuity with the base of its parent tree has never been broken, makes a new tree after the old one has died — indeed in some cases the shoot has helped the mortiferous process by the vigorous crowding of youth. In this whole picture how fares any idea of the necessity or inevitableness of cellular (somatic) senescence? Such an idea plainly has no place in the realities of the con- tinued existence of apple trees. From these facts it is a logically cogent induction to infer that when cells show the characteristic senescent changes which were discussed in the preceding chapter, it is because they are reflecting in their morphology and physiology a consequence of their mutually dependent association in the body as a whole, and not any necessary progressive process inherent in themselves. In other words, may we not tentatively, in the light of our present knowledge, regard senescence as a phenomenon appear- ing in the multicellular body as a whole, as a result of the fact that it is a differentiated and coiiferentiated (to employ a useful term lately introduced by Ritter) mor- phologic and dynamic organization. This phenomenon is reflected morphologically in the component cells. But it does not primarily originate in any particular cell because of the fact that that cell is old in time, or because that cell in and of itself has been alive ; nor does it occur in the cells when they are removed from the mutually dependent relationship of the organized body as a whole and given appropriate physico-chemical condi- tions. In short, senescence appears not to be a primary attribute of the physiological economy of cells as such. If this conception of the phenomenon of senescence is correct in its main features, it suggests the essential futility of attempting to investigate its causes by purely 76 BIOLOGY OF DEATH cytological methods. On the other hand, by clearing away the unessential elements, it indicates where research into the problem of causation of senescence may be profitable. An extremely interesting contribution to the problem of senescence has been made by Carrel and Ebeling in their most recent paper, in which they show that the rate of multiplication of fibroblasts in vitro, and the duration of life of such cultures, is inversely proportional to the age of the animal from which the serum for the culture medium is taken. These results are of such considerable interest that it will be well to quote in full the summary of them given by the authors: "Pure cultures of fibroblasts displayed marked differences in their activity in the plasma of young, middle aged, and old chickens. The rate of cell multiplication varied in inverse ratio to the age of the animal from which the plasma was taken. There was a definite relation between the age of the animal and the amount of new tissue produced in its plasma in a given time. The chart obtained by plotting the rate of cell prolifera- tion in ordinates, and the age of the animal in abscissae, showed that the rate of growth decreased more quickly than the age increased. The de- crease in the rate of growth was 50 per cent, during the first 3 years of life, while in the following 6 years it was only 30 per cent. When the duration of the life of the cultures in the four plasmas was compared, a curve was obtained which showed about the same characteristics. The duration of life of the fibroblasts in vitro varied in inverse ratio to the age of the animal, and decreased more quickly than the age increased. "As the differences in the amount of new tissue produced in the plasma of young, middle aged, and old chickens were large, the growth of a pure culture of fibroblasts could be employed as a reagent for detect- ing certain changes occurring in the plasma under the influence of age. " A comparative study of the growth of fibroblasts in media containing no serum, and serum under low and high concentrations was made, in order to ascertain whether the decreasing rate of cell multiplication was due to the loss of an accelerating factor, or to the increase of an inhibiting one. In high and low concentrations of the serum of young animals, no difference in the rate of multiplication of fibroblasts was observed. This showed that the serum of an actively growing animal did not contain any accel- THE CHANCES OF DEATH 77 erating agent. The same experiments were repeated with the serum of a 3 year old and a 9 year old chicken. The medium made of a high concentration of serum had a markedly depressing effect on the growth, and this effect was greater in the serum of the older animal. "The results of the experiments showed in a very definite manner that certain changes occurring in the serum during the course of life can be detected by modifications in the rate of growth of pure cultures of fibro- blasts and that these changes are characterized by the increase of an inhibiting factor, and not by the loss of an accelerating one. It appeared, therefore, that the substances which greatly accelerate the multiplication of fibroblasts and are found in the tissues do not exist in the blood serum, or are constantly shielded by more active inhibiting factors. The curve which expresses the variations of the inhibiting factor in function of the age was compared with that showing the variations of the rate of healing of a wound according to the age of the subject. For wounds of equal size, the index of cicatrization, which expresses the rate of healing, varies in inverse ratio to the age. The different values of the index of cicatrization of a wound 40 sq cm. in area, taken from measurements made by du Noiiy, were plotted in ordinates, and the age of the subject in abscissae. The curve showed a decrease in the activity of cicatrization, which resembled the decrease in the rate of growth of fibroblasts in function of the age of the lanimal. This suggested the existence of a relation between the factors determining both phenomena." These results suggest that there is produced in some cases by the body or some of its parts, a substance which inhibits the power of cells to multiply or to remain alive. How general such a phenomenon is in occurrence does not yet appear, but, apparently, it must be absent in the case of clonal reproduction in plants already dis- cussed, and in the analogous case of agamic reproduction in lower Metazoa (cf. planarians). It seems possible that the results of Carrel and Ebeling might be open to a slightly different interpretation than that which they give, which hypothecates a specific inhibiting substance in the serum, increasing in either amount or specific potency with age. It seems to me that all of their facts could be interpreted with equal cogency on the supposi- tion that the serum from an old animal is itself senes- 78 BIOLOGY OF DEATH cent as a whole ; that is, has undergone a physico-chemi- cal alteration (as compared with that of a young ani- mal), which is comparable to the morphological and physiological changes which are observable in senescent cells. It may further quite reasonably be supposed that "senescent' serum, because of these physico-chemical alterations, does not furnish so favorable a nutrient me- dium for in vitro cultures as does "young' ' serum. Such a view avoids the necessity of postulating a specific "senescent' substance, the existence of which would be exceedingly difficult to prove. But in any case, whatever explanation is suggested for Carrel and Ebeling's brilliant results, it does not seem to me that the results themselves, which alone are the realities pertinent in the premises, either offer any obstacle to or, indeed, alter the interpretation of senes- cence which I have suggested above. For, what the re- sults really demonstrate is, essentially, that the serum of old animals is a less favorable component of the nutrient medium of cells in vitro than is the serum of young ani- mals. This fact is a contribution to our knowledge of the phenomena and attributes of senescence of first-class importance; but it does not per se, as it appears to me, permit of any new generalization as to the etiology of senescence. CHAPTER III THE CHANCES OF DEATH THE LIFE TABLE UP to this point in our discussion of death and lon- gevity we have, for the most part, dealt with general and qualitative matters, and have not made any particular examination as to the quantitative aspects of the prob- lem of longevity. To this phase attention may now be directed. For one organism, and one organism only, do we know much about the quantitative aspects of longevity. I refer, of course, to man, and the abundant records which exist as to the duration of his life under various condi- tions and circumstances. In 1532 there began in London the first definitely known compilation of weekly " Bills of Mortality." Seven years later, the official registra- tion of baptisms, marriages and deaths was begun in France, and shortly after the opening of the seventeenth century similar registration was begun in Sweden. In 1662 was published the first edition of a remarkable book, a book which marks the beginning of the subject which we now know as ' ' vital statistics. ' ' I refer to ' ' Natural and Political Observations Mentioned in the Following Index, and made upon the Bills of Mortality ' ' by Captain John Graunt, Citizen of London. From that day to this, in an ever widening portion of the inhabited globe we have had more or less continuous published records about the duration of life of man. The amount of such material which has accumulated is enormous. We are only at the 79 80 BIOLOGY OF DEATH beginning, however, of its proper mathematical and bio- logical analysis. If biologists had been furnished with data of anything like the same quantity and quality for any other organism than man it is probable that a vastly greater amount of attention would have been devoted to them than ever has been given to vital statistics, so-called, and there would have been as a result many fundamental advances in biological knowledge now lacking, because material of this sort so generally seems to the profes- sional biologist to be something about which he is in no way concerned. Let us examine some of the general facts about the normal duration of life in man. We may put the matter in this way : Suppose we started out at a given instant of time with a hundred thousand infants, equally distributed as to sex, and all born at the same instant of time. How many of these individuals would die in each succeeding year, and what would be the general picture of the changes in this cohort with the passage of time ? The facts on this point for the Eegistration Area of the United States in 1910 are exhibited in Figure 18, which is based on Glover's United States Life Tables. In this table are seen two curved lines, one marked I x and the other dx. The lx line indicates the number of individuals, out of the original 100,000 starting together at birth, who survived at the beginning of each year of the life span, indicated along the bottom of the diagram. The dx line shows the number dying within each year of the life span. In other words, if we subtract the num- ber dying within each year from the number surviving at the beginning of that year we shall get the series of figures plotted as the lx line. We note that in the very first year of life the original hundred thousand lose over THE CHANCES OF DEATH 81 one-tenth of their number, there being only 88,538 sur- viving at the beginning of the second year of life. In the next year 2,446 drop out, and in the year following that 1,062. Then the line of survivors drops off more slowly between the period of youth and early adult life. At 40 years of age, almost exactly 30,000 of the original 100,000 have passed away, and from that point on the I x line descends with ever increasing rapidity, until about UfC TABUi • S9IO or urn FIG. 18. — Life table diagram. For explanation see text. age 80, when it once more begins to drop more slowly, and the last few survivors pass out gradually, a few each year until something over the century mark is reached, when the last one of the 100,000 who started across the bridge of life together will have ended his journey. This diagram is a graphic representation of that im- portant type of document known as a life or mortality table. It puts the facts of mortality and longevity in their best form for comparative purposes. The first such table actually to be computed in anything like the modern fashion was made by the astronomer, Dr. E. Halley, and 6 82 BIOLOGY OF DEATH was published in 1693, although thirty years before that time Pascal and Fermat (c/. Levasseur) had laid down certain mathematical rules for the calculation of the probabilities of human life. Since Halley's time a great number of such tables have been calculated. Dawson fills a stout octavo volume with a collection of the more important of such tables, computed for different coun- tries and different groups of the population. Now they have become such a commonplace that elementary classes in vital statistics are required to compute them (see for example Dublin's New Haven life table). CHANGES IN EXPECTATION OF LIFE I wish to pass in graphic review some of these life tables in order to call attention in vivid form to an impor- tant fact about the duration of human life. In order to bring out the point with which we are here concerned it will be necessary to make use of another function of the mortality table than either the / or dx lines which are shown in Figure 18. I wish to discuss expectation of life at each age. The expectation of life at any age is defined in actuarial science as the mean or average number of years of survival of persons alive at the stated age. It is got by dividing the total survivor-years of after life by the number surviving at the stated age. Or, if we let e.x denote what is called the curtate expectation of life lx + lx+l + lx+2 + -Mz-fn ex = £ To a first approximation, sufficiently accurate for our present purposes, the total expectation of life, called ex , may be obtained from the curtate expectation by the simple relation «£ = ** + 1/2 THE CHANCES OF DEATH TABLE 6 Changes in expectation of life from the seventeenth century to the present time 83 Average length of life remaining Average length of life remaining to each one alive at beginning to each one alive at beginning of age interval of age interval Age Age Breslau, Carlisle, Breslau, Carlisle, 17th 18th U.S. 1910 17th 18th U.S.1910 century century century century 0- 1 33.50 38.72 51.49 50- 51 16.81 21.11 20.98 1- 2 38.10 44.67 57.11 51- 52 16.36 20.39 20.28 2- 3 39.78 47.55 57.72 52- 53 15.92 19.68 19.58 3- 4 40.75 49.81 57.44 53- 54 15.48 18.97 18.89 4- 5 41.25 50.76 56.89 54- 55 14.99 18.27 18.21 5- 6 41.55 51.24 56.21 55- 56 14.51 17.58 17.55 6- 7 41.62 51.16 55.47 56- 57 14.02 16.89 16.90 7- 8 41.16 50.79 54.69 57- 58 13.54 16.21 16.26 8- 9 40.95 50.24 53.87 58- 59 13.06 15.55 15.64 9-10 40.50 49.57 53.02 59- 60 12.57 14.92 15.03 10-11 39.99 48.82 52.15 60- 61 12.09 14.34 14.42 11-12 39.43 48.04 51.26 61- 62 11.62 13.82 13.83 12-13 38.79 47.27 50.37 62- 63 11.14 13.31 13.26 13-14 38.16 46.50 49.49 63- 64 10.67 12.81 12.69 14-15 37.51 45.74 48.60 64- 65 10.20 12.30 12.14 15-16 36.86 44.99 47.73 65- 66 9.73 11.79 11.60 16-17 36.22 44.27 46.86 66- 67 9.27 11.27 11.08 17-18 35.57 43.57 46.01 67- 68 8.81 10.75 10.57 18-19 34.92 42.87 45.17 68- 69 8.36 10.23 10.07 19-20 34.26 42.16 44.34 69- 70 7.91 9.70 9.58 20-21 33.61 41.46 43.53 70- 71 7.53 9.17 9.11 21-22 32.95 40.75 42.73 71- 72 7.17 8.65 8.66 22-23 32.34 40.03 41.94 72- 73 6.85 8.16 8.22 23-24 31.67 39.31 41.16 73- 74 6.56 7.72 7.79 24-25 31.00 38.58 40.38 74- 75 6.25 7.33 7.38 25-26 30.38 37.86 39.60 75- 76 5.99 7.00 6.99 26-27 29.76 37.13 38.81 76- 77 5.79 6.69 6.61 27-28 29.14 36.40 38.03 77- 78 5.71 6.40 6.25 28-29 28.51 35.68 37.25 78- 79 5.66 6.11 5.90 29-30 27.93 34.99 36.48 79- 80 5.67 5.80 5.56 30-31 27.35 34.34 35.70 80- 81 5.74 5.51 5.25 31-32 26.76 33.68 34.93 81- 82 5.86 5.20 4.96 32-33 26.18 33.02 34.17 82- 83 6.02 4.93 4.70 33-34 25.59 32.36 33.41 83- 84 5.85 4.65 4.45 34-35 25.05 31.68 32.66 84- 85 4.39 4.22 35-36 24.51 31.00 31.90 85- 86 4.12 4.00 36-37 23.97 30.32 31.16 86- 87 3.90 3.79 37-38 23.43 29.63 30.42 87- 88 3.71 3.58 38-39 22.88 28.95 29.68 88- 89 3.59 3.39 39-40 22.33 28.27 28.94 89- 90 3.47 3.20 40-41 21.78 27.61 28.20 90- 91 3.28 3.03 41-42 21.23 26.97 27.46 91- 92 3.26 2.87 42-43 20.73 26.33 26.73 92- 93 3.37 2.73 43-44 20.23 25.71 25.99 93- 94 3.48 2.59 44-45 19.72 25.08 25.26 94- 95 3.53 2.47 45-46 19.22 24.45 24.54 95- 96 3.53 2.35 46-47 18.72 23.81 23.82 96- 97 3.46 2.24 47-48 18.21 23.16 23.10 97- 98 3.28 2.14 48-49 17.71 22.50 22.39 98- 99 3.07 2.04 49-50 17.25 21.81 21.69 99-100 2.77 1.95 In each of the series of diagrams which follow there is plotted the approximate value of the expectation of 84 BIOLOGY OF DEATH life for some group of people at some period in the more or less remote past, and for comparison the expectation of life, either from Glover's table, for the population of the United States Eegistration Area in 1910 — the expec- tation of life of our people now, in short — or equivalent figures for a modern English or French population. Because of the considerable interest of the matter, and the fact that the data are not easily available to 60 SS SO 45 4O 35 JO 25 HALLCY'S BRLSLAU 1687- 1691 LIFL TABLE. 7 p: <0 &> K Q * 10 ZOZ53O3540453O&56O65 or TO •K eo as 90 as /co FIG. 19. — Comparing the expectation of life in the 17th century with that of the present time. biologists, Table 6 is inserted, giving the expectations of life from which certain of the diagrams have been plotted. Figure 19 gives the results from Halley's table, based upon the mortality experience in the city of Breslau, in Silesia, during the years 1687 to 1691. This gives us a rough, but in its general sweep sufficiently accurate picture of the forces of mortality towards the end of the seventeenth century From this diagram it appears that at birth the expectation of life of an individual born in Breslau in the seventeenth century was much lower than THE CHANCES OF DEATH 85 that of an individual born in the United States in 1910. The difference amounts to approximately 18 years! Probably the actual difference was not so great as this, as these early life tables are known to be inaccurate at the ends of the lifespan, particularly at the beginning. At 10 years of age, the difference in expectation of life had been reduced to just over 12 years; at age 20, to a little less than 10 years ; at age 30 to 7-% years ; at age 50 to just over 4 years; at age 70 to l-1/^ years. At age 80 the lines have crossed, but owing to the inade- quate methods, of graduation used by this pioneer actuary, together with the paucity and probably somewhat inac- curate character of his material, no stress is to be laid upon the crossing of the lines, or upon the superior expectation of life at the high ages in the seventeenth century material. What the diagram shows is that the expectation of life at early ages was vastly inferior in the seventeenth century to what it is now, while at advanced ages the chances of living were substantially the same. Let us defer the further discussion of the meaning and explanation of this curious fact until we have examined some further data, Figure 20 compares the expectation of life in England at the middle of the eighteenth century, or about a cen- tury later than the last, with present conditions in the United States. Again we see that the expectation at birth was greatly inferior then to what it is now, but the difference is not so great as it was a century earlier, amounting to but 12-% years instead of the 18 we found before. Further it is seen that, just as before, the expec- tations come closer together with advancing age. By the time age 45 — middle life — is reached the expectation of life was substantially the same in the eighteenth cen- 86 BIOLOGY OF DEATH tury as it is now. At age 47 the eighteenth century line crosses that for the twentieth century, and with a few trifling exceptions, notably in the years from 56 to 62, the expectation of life for all higher ages was greater then than it is now. We see in the eighteenth century the same kind of result as was indicated in the seven- teenth, only differing in degree. to 55 SO 45 MILNE'S CARLISLE 1760 - 1737 IJfT TABLE X. ^ JO ao IS to £ B ?\ IO YEARS CT LIFE FIG. 20 — Comparing the expectation of life in the 18th century with that of the present time. It should be noted that all data as to mortality in the seventeenth and eighteenth centuries lack the degree of accuracy which one desires for purely scientific purposes. By erring generally on the safe side these old mortality tables did well enough for insurance purposes. But quite different results as to the detailed values of life table constants in these early periods are to be found in the literature. For example, Richards constructed some life tables from New England genealogical records, and compared them with Wigglesworth's table, and also with tho3e of modern times. His general conclusion, for the THE CHANCES OF DEATH 87 New England population, is: "that during the last half- century longevity* in Massachusetts, and probably in New England, has increased, that from 1793 to 1850 the increase is less certain and from the seventeenth to the eighteenth century what data we have point rather to a decrease than to anything else." This result may mean any one of a number of things. It may mean merely inadequate and inaccurate data on which the seventeenth century tables were calculated. It may mean a result of less stringent selection in the makeup of the population with the passage of time. In any case it applies only to a small and rather homogeneous group of people. The changes in expectation of life from the middle of the seventeenth century to the present time where the records are most extensive and reliable appear to fur- nish a record of a real evolutionary progression. In this respect at least man has definitely and distinctively changed, as a race, in a period of three and a half cen- turies. This is, of course, a matter of extraordinary interest, and at once stimulates the desire to go still farther back in history and see what the expectation of life then was. Fortunately, through the labors of Karl Pearson, and his associate, W. E. Macdonell, it is pos- sible to do this, if not with precise accuracy, at least to a rough first approximation. Pearson has analyzed the records as to age at death which were found upon mummy cases studied by Professor W. Spiegelberg. These mummies belonged to a period between 1,900 and 2,000 years ago, when Egypt was under Roman dominion. The data were extremely meagre, but from Pearson's analysis of them it has been possible to * Richards somewhat loosely uses this term when he means "expectation of life." BIOLOGY OF DEATH construct the diagram which, is shown in Figure 21. Each circle marks a point where it was possible definitely to calculate an expectation of life. The curve running through the circles is a rough graphic smoothing of the scattered observed data, Unfortunately, there were no records of deaths in early infancy. Either there were no baby mummies, or if there were they have disappeared. 10 15 20 25 30 35 40 45 50 55 6O 65 70 75 30 OS 90 30 FIG. 21. — Comparing the expectation of life of Ancient Egyptians with that of present day Americans. Plotted from Pearson's and Glover's data. For comparison, the expectation of life from Glover's 1910 United States life table is inserted. It will be seen at once that the general sweep of the line is of the same sort that we have already observed in the case of the seventeenth century table. In the early years of life the expectation was far below that of the present time, but somewhere between ages 65 and 70 the Egyptian line crosses the modern American line, and from that period on the individuals living in Egypt at about the time of the birth of Christ could apparently look forward to a longer remaining duration of life, on the aver- THE CHANCES OF DEATH 89 age, than can the American of the present day. Pearson's comment on this fact is worth quoting. He says: "In the course of those centuries man must have grown re- markably fitter to, his environment, or else he must have fitted his environment immeasurably better to himself. No civilized community of to-day could show such a curve as the civilized Romano-Egyptians of 2,000 years ago exhibit. We have here either a strong argument for the survival of the physically fitter man or for the survival of the civilly fitter society. Either man is constitution- ally fitter to survive to-day, or he is mentally fitter, i.e., better able to organize his civic surroundings. Both con- clusions point perfectly definitely to an evolutionary progress. . . . That the expectation of life for a Romano-Egyptian over 68 was greater than for a modern English man or woman is what we might expect, for with the mortality of youth and of middle age enormously emphasized only the very strongest would survive to this age. Out of 100 English alive at 10 years of age 39 survive to be 68; out of 100 Romano-Egyptians not 9 survived. Looking at these two curves we realize at a glance either the great physical progress of man, which enables him far more effectually to withstand a hostile environment, or the great social and sanitary progress he has made which enables him to modify the environ- ment. In either case we can definitely assert that 2,000 years has made him a much ' fitter' being. In this com- parison it must be remembered that we are not placing a civilized race against a barbaric tribe, but comparing a modern civilization with one of the highest types of ancient civilization. ' ' Macdonell was able to continue this investigation on much more extensive material extracted from the Corpus 90 BIOLOGY OF DEATH Inscriptionum Latinarum of the Berlin Academy, which gives records as to age of death for many thousand Roman citizens dying, for the most part, within the first three or four centuries of the Christian era. His mate- rial may, therefore, be taken to represent the conditions a few centuries later than those of Pearson's Romano- Egyptian population. Macdonell was able to calculate UNITED STATLS 0 5 10 15 10 Z5 30 35 40 45 50 55 OO 65 70 75 6O 66 90 95 100 YEARS OF AGL FIG. 22. — Comparing the expectation of life of Ancient Romans with that of present day Americans. Plotted from Macdonell's and Glover's data. three tables of expectation of life — the first for Roman citizens living in the city of Rome itself; second for those living in the provinces of Hispania and Lusitania ; and third, for those living in Africa. The results are plotted against the United States 1910 data, as before, in Figures 22, 23, and 24. Figure 22 relates to inhabitants of the city of Rome itself. The deaths from which the expectations are calculated run into the thousands, and fortunately one is able to separate males and females. As in Pearson's case, which we have just examined, modern American THE CHANCES OF DEATH 91 data are entered for comparison. It will be noted at once that just as in the Romano-Egyptian population the expectation of life of inhabitants of ancient Rome was, in the early years of life, apparently immensely inferior to that of the modern population. From about the age of 60 on, however, the expectation of life appears to have been better then than now. Curiously enough, the expectation 10 15 2O 3O 4O 50 YEARS OT AC-L IOO FIG. 23 — Comparing the expectation of life of the population of the Roman provinces Hispania and Lusitania with that of present day Americans. Plotted from Macdonell's and Glover's data. of life of females was poorer at practically all ages of life than that of the males which exactly reverses the modern state of affairs. Macdonell believes this difference to be real and to indicate that there were special influences adversely affecting the health of females in the Roman Empire, which no longer operate in the modern world. Up to something like age 25 the expectation of life of dwellers in the city of Rome was extremely bad, worse than in the Romano-Egyptian population which Pearson studied, or in the populations of other parts of the Roman Empire as we shall see in the following diagram. Macdonell thinks 92 BIOLOGY OF DEATH that this difference is real and due to circumstances pecu- liar to Eome. The general features of the diagram for the popu- lation of Hispania and Lusitania (Figure 23) are similar to those that we have seen, with the difference that the expectation of life up to age 20 or 25 is not as bad as in the city of Eome itself. Again the females show a lower expectation practically throughout life than do the males. 5 K> 15 20 25 X 35 4O 45 5O FIG. 24 — Comparing the expectation of life of the population of the Roman provinces in Africa with that of present day Americans. Plotted from MacdonelTs and Glover's data. The lines cross the modern American lines at about age 60 and from that point on these colonial Komans appar- ently had a better expectation of life than the modern American has. The Romano-African population diagram appears to start at nearly the same point at birth as does the modern American, and in general the differences up to age 35 are not substantially more marked from modern condi- tions than they are in the seventeenth century Breslau table. The striking thing, however, is that at about age THE CHANCES OF DEATH 93 40 the lines cross, and from then on the expectation of life was definitely superior in the early years of the Christian era to what it is now. It should be said that the curious zigzagging of the lines in all of these Koman tables of Macdonell is due to the tendency, which ancient Eomans apparently had in common with present day American negroes, towards heavy grouping on the even multiples of 5 in the state- ment of their ages. Summarizing the whole matter we see that during a period of approximately 2,000 years man's expectation of life at birth and subsequent early ages has apparently been steadily improving, while at the same time his expec- tation of life at advanced ages has been steadily worsening. The former phenomenon may probably be attributed essen- tially to ever increasing knowledge of how best to cope with the lethal forces of nature.* Progressively better sanitation, in the broadest sense, down through the centur- ies has saved for a time the lives of ever more and more babies and young people who formerly could not with- stand the unfavorable conditions they met, and died in consequence rather promptly. But just because this pro- cess tends to preserve the weaklings, who were speedily eliminated under the rigorous action of unmitigated nat- * No absolute reliance can, of course, be put upon Macdonell's or Pearson's curves. Besides laboring under the serious actuarial difficulty of being expectations calculated from a knowledge of deaths alone, the randomness of the sampling, even on that basis, is extremely doubtful. The only real evidence that these Roman curves represent a rough pic- ture of the truth as to expectation of life in those days, arises from the consideration that they show a difference from present-day expectations which is of the same kind as that which is found between populations of one and two centuries ago and the present, and of a greater amount, as would be expected from the longer time interval, and from what we know has occurred in the material development of civilization in the meantime. 94 BIOLOGY OF DEATH ural selection, there appear now in the higher age groups of the population many weaker individuals than formerly ever got there. Consequently the average expectation of life at ages beyond say 60 to 70 is not nearly so good now as it was under the more rigorous regime of ancient times. Then, any individual who attained age 70 was the surviving resultant of a bitterly destructive process of selection. To run successfully the gauntlet of early and middle life, he necessarily had to have an extraor- dinarily vigorous and resistant constitution. Having come through successfully to 70 years of age it is no mat- ter of wonder that his prospects were for a longer old age than his descendants of the same age to-day can look forward to. Biologically, these expectation of life curves give us the first introduction to a principle which we shall find as we go on to be of the very foremost impor- tance in fixing the span of human longevity, namely that inherited constitution fundamentally and primarily de- termines how long an individual will live. ANALYSIS OF THE LIFE TABLE I shall not develop this point further now, but instead will turn back to consider briefly certain features of the dx line of a life table. Figure 18 shows that this line, which gives the number of deaths occurring at each age, has the form of a very much stretched letter S resting on its back. Some years ago, Pearson undertook the analysis of this complex curve, and drew certain inter- esting conclusions as to the fundamental biological causes lying behind its curious sinuosity. His results are shown in Figure 25. He regarded the dx line of the life table as a compound curve, and by suitable mathematical analysis broke it up THE CHANCES OF DEATH 95 into five component frequency curves. The data which he used were furnished by the df line of Ogle's life table, based on the experience of 1871 to 1880 in England. This line gives the deaths per annum of one thousand persons born in the same year. The first component which he sepa- rated was the old age mortality. This is shown by the dotted curve having its modal point between 70 and 75 years, at the point lettered Ol on the base of the diagram. PEARSON'S GRADUATION OF IOQ or un; FIG. 25. — Showing Pearson's results in fitting the dx line of the life table with 5 skew frequency curves. Plotted from the data of Pearson's original memoir on "Skew Variation" iu Phil. Trans. Roy. Soc. This component, according to Pearson's graduation, accounted for 484.1 deaths out of the total of 1,000, or nearly one-half of the whole. Its range extends from under 20 years of age to the upper limit of life, at approx- imately 106 years. The second component includes the deaths of middle life. This is the smooth curve having its modal point between 40 and 45 years at the point on the base marked 02. Its range extends from about 5 years of age to about 65. It accounts for 175.2 deaths out of the total of 1,000. It is a long, much spread out 96 BIOLOGY OF DEATH curve, exhibiting great variability. The third compo- nent is made up by the deaths of youth. This accounts for 50.8 deaths out of the total of a thousand, and its range extends from about the time of birth to nearly 45 years. Its mid-point is between 20 and 25 years, and it exhibits less variability than either the middle life or the old age curves. The fourth component, the modal point of which is at the point on the base of the diagram marked 04 covers the childhood mortality. It accounts for 46-4 deaths out of the total of 1,000. Its range and variability are obviously less than those of any of the other three components so far considered. The last, excessively skew component, is that which describes the mortality of in- fancy. It is given by a J shaped curve accounting for 245.7 deaths after birth, and an antenatal mortality of 605. In order to get any fit at all for this portion of the mortality curve it is necessary to assume that the deaths in utero and those of the first months after birth are a homogeneous connected group. Summing all these components together it is seen that the resulting smooth curve very closely fits the series of small circles which are the original observations. From the standpoint merely of curve fitting no better result than this could be hoped for. But about its bio- logical significance the case is not quite so clear, as we shall presently see. Pearson himself thinks of these five components of the mortality curve as typifying five Deaths, shooting with different weapons, at different speeds and with dif- fering degrees of precision at the procession of human beings crossing the Bridge of Life. The first Death is, according to Pearson, a marksman of deadly aim, con- centrated fire, and unremitting destructiveness. He kills THE CHANCES OF DEATH 97 before birth as well as after and may be conceived as beating down young lives with the bones of their ances- tors. The second marksman who aims at childhood has an extremely concentrated fire, which may be typified by the machine gun. Only because of the concentration of this fire are we able to pass through it without appal- ling loss. The third marksman Death, who shoots at youth has not a very deadly or accurate weapon, perhaps a bow and arrow. The fire of the fourth marksman is slow, scattered and not very destructive, such as might result from, an old fashioned blunderbuss. The last Death plies a rifle. None escapes his shots. He aims at old age but sometimes hits youth. His unremitting activity makes his toll large. We may let Pearson sum the whole matter up in his own words : ' i Our investigations on the mortality statis- tics have thus led us to some very definite conclusions with regard to the chances of death, Instead of seven we have five ages of man, corresponding to the periods of infancy, of childhood, of youth, of maturity or middle age, and of senility or old age. In the case of each of these periods we see a perfectly regular chance distri- bution, centering at a given age, and tailing off on either side according to a perfectly clear mathematical law. . . "Artistically, we no longer think of Death as striking chaotically; we regard his aim as perfectly regular in the mass, if unpredictable in the individual instance. It is no longer the Dance of Death which pictures for us Death carrying off indiscriminately the old and young, the rich and the poor, the toiler and the idler, the babe and its grandsire. We see something quite different, the cohort of a thousand tiny mites starting across the Bridge of Life, and growing in stature as they advance, 7 98 BIOLOGY OF DEATH till at the far end of the bridge we see only the gray- beard and the 'lean and slippered pantaloon.7 As they pass along the causeway the throng is more and more thinned ; five Deaths are posted at different stages of the route longside the bridge, and with different skewness of aim and different weapons of precision they fire at the human target till none remains to reach the end of the causeway — the limit of life/' This whole, somewhat fanciful, conception of Pear- son's needs a little critical examination. What actually he has done is to get a good empirical fit of the dx line by the use of equations involving all told some 17 con- stants. Because the combined curve fits well, and funda- mentally for no other reason, he implicitly concludes that the fact that the fit is got by the use of five compo- nents means biologically that the dx line is a compound curve, and indicates a five-fold biological heterogeneity in the material. But it is a very hazardous proceeding to draw biological conclusions of this type from the mere fact that a theoretical mathematical function or functions fits well a series of observational data. I fully discussed this point several years ago and pointed out: "The kind of evidence under discussion can at best have but inferential significance; it can never be of de- monstrative worth. It is based on a process of reasoning which assumes a fundamental or necessary relationship to exist between two sets of phenomena because the same curve describes the quantitative relations of both sets. A little consideration indicates that this method of rea- soning certainly cannot be of general application, even though we assume it to be correct in particular cases. The difficulty arises from the fact that the mathematical functions commonly used with adequate results in physi- THE CHANCES OF DEATH 99 cal, chemical, biological, and mathematical investigations are comparatively few in number. The literature) of science shows nothing clearer than that the same type of curve frequently serves to describe with complete accuracy the quantitative relations of widely different natural phenomena. As a consequence, any proposition to conclude that two sets of phenomena are causally or in any other way fundamentally related solely because they are described by the same type of curve is of a very doubtful validity/' Henderson has put Pearson's five components together in a single equation, as follows : 7 7525 0.2215 (x— 71.5) — [.05524 Or— 41.5) ]2 — [.09092 (Z — 22.5)]2 -f 2.6 e + 8.5 (x — 2\ *3271 & — .3271 (x — 3) + 415.6 (, + . 75) -^-.75(* + Henderson says regarding this method of Pearson's for analyzing the life table: ". . . it is difficult to lay a firm foundation for it, because no analysis of the deaths into natural divisions by causes or otherwise has yet been made such that the totals in the various groups would conform to those frequency curves." The italics in this quotation are the present writer's for the purpose of em- phasizing the crucial point of the whole matter. Now it is altogether probable that one could get just as good a fit to the observed dx line as is obtained by Pearson's five components by using a 17 constant equa- tion of the type y = a -f bx + cxz + dx3 + ex* +/# 4- gx< -f ........ -f ^l6 100 BIOLOGY OF DEATH and in that event one would be quite as fully justified (or really unjustified) in concluding that the dx line was a homogeneous curve as Pearson is in concluding from his five-component fit that it is compound. Indeed Witt- stein's formula involving but four constants n n — (M — x) i — (mx) qx = a ~^"^,a gives a substantially good fit over the whole range of life. It is, of course, apparent that the formula as here given is in terms of another function, qx , of the life table, rather than the dx which we have hitherto been discussing. But no difference is in fact involved, qx values may be imme- diately converted into dx values by a simple arithmeti- cal transformation. But in neither Pearson's, Wittstein's, nor any other case is the curve-fitting evidence, by and of itself, in any sense a demonstration of the biological homogeneity or heterogeneity of the material. Of far greater impor- tance, and indeed conclusive significance, is the fact, to be brought out in a later chapter, that in material experi- mentally known to be biologically homogeneous, a popu- lation made up of full brothers and sisters out of a brother x sister mating and kept throughout life in a uniform environment identical for all individuals, one gets a dx line in all its essential features, save for the absence of excessive infant mortality arising from perfectly clear biological causes, identical with the human dx line. It has long been apparent to the thoughtful biologist that there was not the slightest biological reason to suppose that the peculiar sinuosity of the human dx line owed its origin to any fundamental heterogeneity in the material, or differentiation in respect of the forces of mortality. THE CHANCES OF DEATH 101 Now we have experimental proof, to be discussed in a later chapter, that with complete homogeneity of the material, both genetic and environmental, one gets just the same kind of d line as in normal human material. 3? We must then, I think, come to the conclusion that bril- liant and picturesque as is Pearson's conception of the five Deaths, actually there is no slightest reason to sup- pose that it represents any 'biological reality, save in the one respect that his curve fitting demonstrates, as any other equally successful would, that deaths do not occur chaotically in respect of age, but instead in a regular manner capable of representation by a mathematical function of age. An interesting and suggestive analysis of the dx line, resting upon a sounder biological basis than Pearson's, has lately been given by Arne Fisher. He breaks the curve up into 8 or 9 components, based upon the compar- atively stable values of the death ratios for different groups of diseases characteristic of different ages. The resulting total curve fits the facts from age 10 on, very well, and makes possible the calculation of a complete life table from a knowledge of deaths only. CHAPTER IV THE CAUSES OF DEATH IT has been suggested in earlier chapters that natural death of the metazoan body may come about fundamen- tally because of the differentiation and consequent mu- tual dependence of structure and function of that body. It is a complex aggregate of cells and tissues, all mutually dependent upon each other and in a delicate state of adjustment and balance. If one organ for any accidental reason, whether internal or external, fails to function normally it upsets this delicate balance, and if normal functioning of the part is not promptly restored, death of the whole organism eventually results. Furthermore, it is apparent that death does not strike in a haphazard or random manner, but instead in a most orderly way. There are certain periods of life — notably youth — where only an insignificant fraction of those ex- posed to risk ever die. At other ages, as, for example, extreme old age and early infancy, death strikes with appalling precision and frequency. Further we recall with Seneca that nascimus uno modo multis morimur. Truly there are many ways of dying. The fact is obvious enough. But what is the biological meaning of this mul- tiplicity of pathways to the river Styx? There is but one pathway into the world. Why so many to go out? To the consideration of some phases of this problem attention is directed in this chapter. By international agreement among statisticians the causes of human mortality are, for statistical purposes, 102 THE CAUSES OF DEATH 103 rather rigidly defined and separated into something over 180 distinct units. It should be clearly understood that this convention is distinctly and essentially statistical in its nature. In recording the statistics of death the regis- trar is confronted with the absolute necessity of putting every demise into some category or other in respect of its causation. However complex biologically may have been the train of events leading up to a particular end, the statistician must record the terminal ' ' cause of death' as some particular thing. The International Classifica- tion of the Causes of Death is a code which is the result of many years' experience and thought. Great as are its defects in certain particulars, it nevertheless has cer- tain marked advantages, the most conspicuous of which is that by its use the vital statistics of different countries of the world are put upon a uniform basis. The several separate causes of death are grouped in the International Classification into fourteen general classes. These are: I. General diseases. II. Diseases of the nervous system and of the organs of special sense. III. Diseases of the circulatory system. IV. Diseases of the respiratory system. V. Diseases of the digestive system. VI. Non-venereal diseases of the genito-urinary system and annexa. VII. The puerperal state. VIII. Diseases of the skin and of the cellular tissue. IX. Diseases of the bones and organs of locomotion. X. Malformation. XI. Early infancy. XII. Old age. XIII. External causes. XIV. Ill-defined diseases. Perhaps the most outstanding feature which strikes one about the International List is that it is not primarily 104 BIOLOGY OF DEATH a biological classification. Its first group, for example, called " General Diseases,' which caused in 1916 in the Registration Area of the United States approximately one-fourth of all the deaths, is a most curious biological and clinical melange. It includes such diverse entities as measles and malaria, tetanus and tuberculosis, cancer and gonococcus infection, alcoholism and goiter, and many other unlike causes of death. For the purposes of the statistical registrar it perhaps has useful points to make this "General Diseases' grouping, but it clearly corresponds to nothing natural in the biological world. Again in such parts of the scheme as do have some biological foundation the basis is different in different rubrics. Some have an organological basis, while others have a causational. For purposes of biological analysis, I developed some time ago an entirely different classification of the causes of death, on what appears ,to be a reasonably consistent basis.* The underlying idea of this new classification was to group all causes of death under the heads of the several organ systems of the body, the functional break- down of which is the immediate or predominant cause of the cessation of life. All except a few of the statistically recognized causes of death in the International Classifi- cation can be assigned places in such a biologically * It should be clearly understood that I am not advocating a new classification of the causes of death for statistical use. I should oppose vigorously any attempt to substitute a new classification (mine or any other) for the International List now in use. Uniformity in statistical classification is essential to usable, practical vital statistics. Such uni- formity has now become well established through the International Classi- fication. It would be most undesirable to make any radical changes in the Classification now. I have made a rearrangement of the causes of death, for the purposes of a specific biological problem, and no other. I am not "proposing a new classification of vital statistics" for official or any other use except the one to which it is here put. THE CAUSES OF DEATH 105 grouped list. It has a sound logical foundation in the fact that, biologically considered, death results because some organ system, or group of organ systems, fails to continue its functions. The headings finally decided upon in the new classi- fication were as follows: I. Circulatory system, blood and blood-forming organs. II. Respiratory system. III. Primary and secondary sex organs. IV. Kidneys and related excretory organs. V. Skeletal and muscular systems. VI. Alimentary tract and associated organs concerned in metabolism. VII. Nervous system and sense organs. VIII. Skin. IX. Endocrinal system. X. All other causes of death. The underlying idea of this rearrangement of the causes of death is to put all tho^e lethal entities together which bring about death because of the functional organic breakdown of the same general organ system. The cause of this functional breakdown may be anything whatever in the range of pathology. It may be due to bacterial infection ; it may be due to trophic disturbances ; it may be due toj mechanical disturbances which prevent the continuation of normal function; or to any cause what- soever. In other words the basis of the classification is not that of pathological causation, but it is rather that of organological breakdown. We are now looking at the question of death from the standpoint of the biologist, who concerns himself not with what causes a cessation of function, but rather with what part of the organism ceases to function, and therefore causes death. In a series of papers already published I have given a detailed account of this classification, and the reasoning on which particular causes of death are placed in it where 106 BIOLOGY OF DEATH they are. Space is lacking here to go into the details, and I must consequently ask the reader either to take it on faith for the time being that the classification is at least a fairly reasonable one, or to take the trouble to go over it in detail in the original publication.* GENEKAL RESULTS OF BIOLOGICALLY CLASSIFIED DEATH BATES Here I should like to present first some general statis- tical results of this classification. The data which we shall first discuss are in the form of death rates, from various causes, per hundred thousand living at all ages, arranged by organ systems primarily concerned in death from specified diseases. The statistics came from three widely separated localities and times, viz., (a) from the Registration Area of the United States; (b) from England and Wales; and (c) from the City of Sao Paulo, Brazil. The summarized results are shown in Table 7, and in graphic form in Figure 26. The rates are arranged in descending order of magni- tude for the United States Registration Area, with the exception of those of group X, all other causes of death. We note in passing that this biologically unclassifiable group includes roughly 10 to 15 per cent of the total mortality. It may be well to digress a moment to con- sider why these deaths cannot be put into our general scheme. Table 8 exhibits the rates included in class X. This residue comprises in general three categories (a) accidental and homicidal deaths; (b) senility; and * Cf. particularly Pearl, R. " On the embryological basis of human mortality." (Proc. Natl. Acad. Sci. Vol. 5, pp. 593-598, 1919) and "Cer- tain evolutionary aspects of human mortality rates." (Amer. Natl. Vol. LIV. pp. 5-44, 1920). The following section as well as Chapter V are largely based upon the second of the two papers. THE CAUSES OF DEATH 107 (c) deaths from a variety of causes which are statisti- cally lumped together and cannot be disentangled. Ac- cidental and homicidal deaths find no place in a biologi- TABLE 7 Showing the Relative Importance of Different Organ Systems in Human Mortality Group No. Organ System Death Rates per 100,000 Registration Area, U. S. A. England and Wales 1914 Sao Paulo 1917 1906-10 1901-05 II VI I VII IV III V VIII IX Respiratory system 395.7 334.9 209.8 175.6 107.2 88.1 12.6 10.1 1.5 460.5 340.4 196.8 192.9 107.4 77.4 13.7 13.3 1.2 420.2 274.1 208.6 151.9 19.4 95.4 18.2 12.0 1.9 417.5 613.8 254.8 124.3 83.4 103.2 6.8 7.9 1.1 Alimentary tract and associated organs Circulatory system, blood Nervous system and sense organs . Kidneys and related excretory organs Primary and secondary sex organs . Skeletal and muscular system .... Skin Endocrinal system Total death rate classifiable on a biological basis 1,335.5 1,403.6 1,201.7 1,612.8 X All other causes of death 171.3 211.8 141.4 109.8 cal classification of mortality. A man organically sound in every respect may be instantly killed by being struck by a railroad train or an automobile. The best possible case that could be made out for a biological factor in such deaths would be that contributory carelessness or negli- gence, which is a factor in some portion of accidental deaths, bespeaks a small but definite organic mental in- feriority or weakness, and that, -therefore, accidental deaths should be charged against the nervous system. This, however, is obviously not sound. For, in the first place, in many accidents there is no factor of contributory RESPIRATORY SYSTEM ALIMENTARY TRACT AND ASSOCIATED ORGANS CIRCULATORY SYSTEM. BLOOD NERVOUS SYSTEM AND SENSE ORGANS KIDNEYS AND RELATED EXCRETORY , ORGANS PRIMARY AND SECONDARY SEX ORGANS SKELETAL ANi MUSCULAR SYSTEM 395.7 4 2.0. a \ 417.5 6/3.3 SKIN ENDOCRINAL SYSTEM 209.6 208.6 ] Z54.& 103. Z U$. REG AREA 19O6-IO ENGLAND AND WALES 1914- SAO PAULO 1917 THE CAUSES OF DEATH 109 negligence in fact, and, in the second place, in those cases where such negligence can fairly be alleged its degree or significance is undeterminable and in many cases surely slight. Senility as a cause of death is not further classifiable TABLE 8 All Other Causes No. "Cause of Death" as per International Classification Registration Area, U. S. A. England and Wales 1914 Sao Paulo 1917 11906-10 1901-05 187, 188 & 189 154 45 152* 34 46 55 153 19 All external causes (except suicide) Ill-defined diseases 91.9 29.4 29.0 12.9 3.4 2.1 1.0 1.0 0.3 0.3 87.8 47.8 41.0 16.1 2.6 2.0 1.5 0.5 12.3 0.2 26.1 7.3 81.5 16.6 5.1 1.6 0.5 1.5 0.6 0.6 36.4 36.3 11.1 17.9 3.3 0.2 0.9 3.5 0 0.2 Senility Cancer of other organs or of organs not specified Other causes peculiar to early in- fancy . Tuberculosis of other organs Other tumors (female genital or- gans excepted) Other general diseases Lack of care Other epidemic diseases Totals . 171.3 211.8 141.4 109.8 * In part. on an organological basis. A death really due to old age, in the sense of Metchnikoff, represents, from the point of view of the present discussion, a breaking down or wearing out of all the organ systems of the body con- temporaneously. In a strict sense this probably never, or at best extremely rarely, happens. But physicians and registrars of mortality still return a certain number of deaths as due to ' ( senility. ' ' Under the circumstances 110 BIOLOGY OF DEATH it is not possible to go behind such returns biologically. The second line of Table 8, "Ill-defined diseases, " furnishes a striking commentary on the relative efficiency of the medical profession in the United States and Eng- land in respect of the reporting of the causes of death. Only about one-fourth as many deaths appear in the English vital statistics as due to ill-defined and unknown causes as in the United States figures. Returning now to the consideration of the general results set forth in Table 7 and Figure 26, a number of interesting points about human mortality are apparent. In the United States, during the decade covered, more deaths resulted from the breakdown of the respiratory system than from the failure of any other organ system of the body. The same thing is true of England and Wales. In Sao Paulo the alimentary tract takes first position, with the respiratory system a rather close second. The tremendous death rate in Sao Paulo charge- able to the alimentary tract is chiefly due to the relatively enormous number of deaths of infants under two from diarrhoea and enteritis. Nothing approaching such a rate for this category as Sao Paulo shows is known in this country or England. In all three localities studied the respiratory and the alimentary tract together account for rather more than half of all the deaths biologically classifiable. These are the two organ systems which, while physically internal, come in contact directly at their surfaces with environ- mental entities (water, food, air) with all their bacterial contamination. The only other organ system directly exposed to the environment is the skin. The alimentary canal and the lungs are, of course, in effect invaginated surfaces of the body. The mucous membranes which line them are far less resistant to environmental stresses, THE CAUSES OF DEATH 111 both physical and chemical, than is the skin with its pro- tecting layers of stratified and cornified epithelium. The organs concerned with the blood and its circula- tion— the heart, arteries and veins, etc. — stand third in importance in the mortality list. Biologically the blood, through its immunological mechanism, constitutes the second line of defense which the body has against noxious invaders. The first line is the resistance of the outer cells of the skin and the lining epithelium of alimentary tract, lungs, and sexual and excretory organs. When invading organisms pass or break down these first two lines of defense, the battlers then with the home guard, the cells of the organ system itself, which, like the industrial workers of a commonwealth, keep the body going as a whole functioning mechanism. Naturally it would be ex- pected that the casualties would be far heavier in the first two defense lines (respiratory and alimentary systems and the blood and circulation) than in the home guard. Death rates, when biologically classified, bear out this expectation. In the United States the kidneys and related excre- tory organs are responsible for more deaths than the sex organs. This relation is reversed in England and Wales, and in Sao Paulo. This difference is mainly due in both countries to premature birth. The higher premature birth rate for these two localities as compared with the United States might conceivably be explained in any one of several ways. It might mean better obstetrics here than in the other localities, or it might mean that the women of this country, as a class, are somewhat superior physiologically in the matter of reproduction, when they do reproduce, or it might be in some manner connected with differences in birth rates. 112 BIOLOGY OF DEATH The last three organ systems, skeletal and muscular system, skin and endocrinal organs, are responsible for so few deaths relatively as not to be of serious moment. There is one general consequence of these results upon which I should like to dwell a moment longer. • In a broad sense the efforts of public health and hygiene have been directed against the affections comprised in the first two items in the chart, those of the respiratory system and the alimentary tract. The figures for the two five-year periods in the United States, 1901-05 and 1906-10, indi- cate roughly the rate of progress such measures are making, looking at the matter from a broad biological standpoint. In reference to the respiratory system there was a decline of fourteen per cent, in the death rate be- tween the two periods. This is substantial. It is prac- tically all accounted for in phthisis, lobar pneumonia and bronchitis. For the alimentary tract the case was not so good — indeed far worse. Between the two periods the death rate from this cause group fell only 1.8 per cent. All the gain made in typhoid fever was a great deal more than offset by diarrhoea and enteritis (under two), congenital debility and cancer. Child welfare, both prenatal and postnatal, seems by long odds the most hopeful direction in which public health activities can expect, at the present time, substantially to reduce the general death rate. This is a matter funda- mentally of education. SPECIFIC DEATH KATES BIOLOGICALLY CLASSIFIED Up to this point in our discussions we have been deal- ing with crude death rates, uncorrected for the age and sex distributions of the populations concerned. It is, of course, a well known fact that differences in age and THE CAUSES OF DEATH 113 sex constitution of populations may make considerable differences in crude death rates, in cases where no real differences in the true force of mortality exist. What is essential for the further prosecution of the analysis of the causes of death is to get specific death rates for the several causes. By an age and sex specific death rate is meant the rate got by dividing the number of persons, of particular specified age and sex, dying from a particu- lar cause, by the total number of persons living in the same population of the same age and sex. In other words, we need to get as the divisor of the rate fraction the number of persons who can be regarded as truly ex- posed to risk. This exposed-to-risk portion of the popu- lation is never correctly stated in a crude death rate. For example, a person now 75 years old cannot be re- garded as exposed to risk of death at age 45. He was once exposed to that risk but passed it safely. Yet in a crude death rate he is counted with those of age 45. Age and sex specific death rates have hitherto been available for the American people, in any general or com- prehensive form, only from the extensive memoir by Dublin, Kopf and Van Buren, based upon the mortality experience of the Metropolitan Life Insurance Company with its industrial policy holders. In a broad way, it may be said that the data on which the following discus- sion is based, derived from the general population of the Registration Area, are essentially in accord with those of Dublin on a more restricted group. Owing to limita- tions of space, it is not possible to present all the detailed rates here. With the aid of Dr. William H. Davis, director of vital statistics in the Census Bureau, who very kindly provided me with the necessary unpublished data, it has 8 114 BIOLOGY OF DEATH been possible to calculate the specific death rates for each of the 189 causes of death of the International List, for each sex separately, and for each age in 5 year groups, for the United States Registration Area, exclusive of North Carolina, in 1910. These results have been put together in the biological scheme of classification and may be presented briefly in the form of diagrams. The summary table from which these curves are plot- ted is given as Table 9. Let us first consider deaths from all causes taken together, in order to recall to mind the general form of a death rate curve. It will be noted, at once, that the rates are plotted along the vertical axis on what strikes one at first as a peculiar scale. The scale is logarithmic. The horizontal lines are spaced in proportion to the logarithms of the numbers at their left, instead of in pro- portion to the numbers themselves. The advantages of this method of plotting in the present case are two-fold. First, it is possible to get a much wider range of values on the diagram ; and second the logarithmic scale permits direct and accurate estimation of the rate of change of a variable. A straight line forming an angle with the hor- izontal on a logarithmic scale means that the variable is increasing or decreasing, as the case may be, at a con- stant rate of change. Figure 27 gives the specific death rates for the com- bined total of all causes. The curve in general has the form of a V, with one limb much extended and pulled over to the right. Examining it more in detail, we note that in the first year of life, the specific death rate, or, as we may roughly call it, the force of mortality, bears heav- ier on female infants than on the males. Out of a thou- sand exposed to risk, 124 male babies die in that year, THE CAUSES OF DEATH 115 00 O 00 os 00 00 00 00 p b co 00 05 o t-l N5 Ol co 00 os co oo P O to Cn 00 bs CO to os -~J OS OS •£• CO tO i— i—OS en 01 co to i-* t-i OS "^4 rf^ CO tO *-^ t— i CO »-i p|£-rf».pipOOpOi*.|f».COCOCOtOtO|-'- t-i p p 00 i—OiCsas^-CStOCOCS-JCTiGCCOCnOOOOOi-iCSCOCi OS Cn rf>> bo i— >• >— i p oo p oo po t-i • *•• CO tO 1-1 oo ^ o GO OS to to >-i *-i i— i tof».COtOi-'H4p-i OS tf^OS H- tO tO tO i-1 H* 1-1 00 tOrf^i— i-^]COOOOCntf».COtOi-'>-t Or COrf^ co co co 10 to I-- t— COCO^IOCOOOCn*-tOtOi-ii-' o •O H. OS i-i co co to H-I 1-1 p-i Cn i-iMi-iOOCO>-iCO^JCn4>. COOCCOOCni-iOiOO co to 1-1 >-i OS to to >-i j-i bs b bs co '»a Cn to >-* i-iOOOOOOOOOOOtO OO'M^COCOtOtOi-'OO^'vI Cn 00 i i t OOOOOOOOOOOOOOO OO K-itOtOtOtOtOtOi-'i— OOOOOO O>— •0 M ^. CO :-U t-0 O 1 *-'OOOO>-'O>-'OOOOOOOOOOOO OOtOOiOCOOOOO-J-vl-vlCnrfkCOtOi—OOOO I-* -^ 00 CO O O Ci i— i co Oi CO Oi tO 1-1 i_» COCOtOpJ-ipSrfkCOtOtOtOJ-it-'i-ij-ij-i — — bb>-ioinrf>-bi-icnt-ibbobob5bsbTobsbscocn •O •o -o ^ tO I— H- O Cn tO >-i CO to i— ' i— *•! OS i-i OO CO CO tO 1— >-i OtOCnOOSt^tOtO^-'i-' O •o o o c •o a o •o O HI o •c O o c •o O o O o •a O o •a O 3 c •o O I O o c •a X O p s i o to 2- a «5 r^. ^ >- f ^ O y-*- — ^ ^ «> . &s &3 & *s. Cc •« ^ a ^ s • ^ 1 1 &** H c^ o r § CC i* . s i « A ^ "-S CQ ^c f I ^ °- ?^ I- ^> I ^ " <^ ^.. » » §• §£ o a. t •e si ce 8 116 BIOLOGY OF DEATH and 143 female. This is the only year of life in which the total force of mortality is heavier among females than males. From that time on to the end of the span of life, I,UOU ^A 1* A & 100 \ TOTALS A / \ t £' o ^ 7 5 1 / // _y U // 7 Q /' -i 1 ^ rx^ m \ > Xf y ' \ 1 j& ^1 y ^ ^ fv ft ^ ,>• li \ ^ ^~ \ ^ s^ ^*" ^"^ s> s s ^X ^^*** w V f>**1 ^x i O -5 IO 15 2O 3O 35 4O 45 5O 55 GO 65 7O 76 6O AGE. 9O 1OO FIG. 32. — Diagram showing specific death rates at each age from breakdown of the skeletal and muscular systems (Group V). late twenties. The whole curve shows a very gradual change in the rates. The next diagram, Figure 33, shows one of the most THE CAUSES OF DEATH 129 significant organ groups in the force of its specific mor- tality. Breakdown and failure to function properly of the primary organs of metabolism — the organs which too s oo O 5 10 15 20 2-5 30 35 4O 45 5O 55 CO 65 AGL FIQ. 33. — Diagram showing the specific rates of death at each age from breakdown of the alimentary tract and associated organs of metabolism (Group VI). transform the fuel of the human machine into vital energy — occur with relatively heavy frequency at all periods of life. These curves are among the few which show an 9 130 BIOLOGY OF DEATH absolutely higher specific force of mortality in infancy than in extreme old age. There is practically no signif- icant difference between the male and female curve at too 10 01 OOl NERVOUS SYSTEM AND SENSE ORGANS 4 10 15 2O JO 35 -40 45 £O £5 6O 65 7O 75 6O 85 9O 95 IOO AGL FIG. 34. — Diagram showing the specific death rates at each age from breakdown of the nervous system and sense organs (Group VII). any portion of life. During early adult life the female curve lies below the male, but by only a small amount. Out of every thousand infants under one, about sixty THE CAUSES OF DEATH 131 die in the first year of life from breakdown of the ali- mentary tract and its associated organs. After the low point, which falls in the relatively early period of 7 to 12 years of age, there is a rapid rise for about ten years in the specific rates of mortality, followed by a slowing off in the rate of increase for the next ten or fifteen years, after which point the curve ascends at a practically uni- form rate until the end of the span of life. Figure 34 shows the trend of the specific mortality from breakdown of the nervous system and sense organs. This organ group, on the whole, functions very well, giv- ing a relatively low rate of mortality until towards the end of middle life. Then the specific rates get fairly large. The low point in this curve is, as in most of the others, at about the time of puberty. From then on to the end of the life span the specific rates increase at a practically uniform rate. The female curve everywhere lies below the male curve except at the extreme upper end of the life span. Before that time, and particularly between the ages of 20 and 50, the business of living evidently either imposes no such heavy demand on the nervous system of the female as it does on that of the male, or else the nervous system of the female is organi- cally sounder than that of the male. The former sug- gestion seems the more probable. That breakdown and failure to function properly, of the skin as an organ system, is a relatively insignificant factor in human mortality, is demonstrated by Figure 35. From a specific death rate of about 1 per thousand in the first year of life it drops abruptly, practically to zero, in early childhood. At about the time of puberty it be- gins to rise again, and ascends at a steady rate during all the remainder of life. The final high point reached 132 BIOLOGY OF DEATH is absolutely low, however, amounting to a specific death rate among those exposed to risk of only a little more than 4 per thousand at the extreme end of life. The female too OOI AGL FIG. 35. — Diagram showing the specific death rates at each age chargeable against the skin (Group VIII). curve lies well below the male curve practically through- out its course. Deaths from failure to function properly of the organs THE CAUSES OF DEATH 133 of the endocrinal system, including the thyroid gland, suprarenal glands, etc., do not become significant until middle life in the case of the male, as shown in Figure 36, 100 10 I O.I O.Ol ENDOCRINAL SYSTEM l\ L/ IO 15 ZO £5 3O 35 4O 45 6O 55 6O 65 7O AGE. QO 65 9O 95 IOQ FIG. 36. — Diagram showing the specific death rates at each age from breakdown of the endocrinal system (Group IX).Jj although in the female the curve begins to rise from pu- berty on. The specific rates at all ages, of course, are extremely small, practically never rising to more than 134 BIOLOGY OF DEATH 1/10 of one person per thousand exposed to risk. The well-known fact that these glandular organs, whose se- cretions are so important for the normal conditions of IOO to O.I ALL OTHER CAUSES OF DEATH /5 £O 2*5 3O 35 4O SO 55 6O 65 7O 75 AGE. 85 9O 95 IOO FIQ. 37. — Diagram showing the specific death rates from all other causes of death not covered in the preceding categories (Group X). life, are much more unstable and liable to breakdown in the female than in the male, is strikingly shown by this diagram. THE CAUSES OF DEATH 135 Finally, we have the diagram for our omnium gatherum group, the "All other causes of death," in Figure 37. Here we see that, because of accidental and violent deaths, the male specific mortality curve lies far above the female, from youth until old age has set in, about age 75. From that point on to the end of the span of life both curves ascend rapidly together, as a result of the deaths recorded as resulting from senility. Eventually it is to be expected that no deaths will be registered as result- ing from senility. We shall have them all put more nearly where they belong. These diagrams of specific forces of mortality give altogether a remarkably clear and definite picture of how death occurs among men. We see that failure of certain organ systems, such as the lungs, the heart, the kidneys, to maintain their structural and functional integrity, has an overwhelmingly great effect in determining the total rate of mortality as compared with some of the other organ systems. One cannot but be impressed, too, with the essential orderliness of the phenomena we have ex- amined. The probability of any particular organ system breaking down and causing death is mathematically def- inite at each age, and changes in a strikingly orderly manner as age changes, as is shown in Table 11. Thus we find that in the first year of life it is the alimentary tract and its associated organs which most frequently break down and cause death. From age 1 to age 60 the specific force of mortality from breakdown of the respiratory system is higher (with a few insignificant exceptions in the females) usually by a considerable amount, than that associated with any other organ system of the body. From 60 to 90 years of age the circulatory 136 BIOLOGY OF DEATH system takes the front rank, with a higher specific mor- tality rate than any other organ system. TABLE 11 The most fatal organ systems at different ages MALES FEMALES Per cent, of all i Per cent, of all biologically classifiable deaths due to Organ system concerned in largest Age Group Organ system concerned in largest biologically classifiable deaths due to breakdown of specified organ proportion of fatalities proportion of fatalities breakdown of specified organ system system 68.8 Alimentary tract 0- 1 Alimentary tract 40.6 50.1 Respiratory 1— 4 Respiratory 51.3 41.2 Respiratory 5— 9 Respiratory 42.5 27.1 Respiratory 10—14 Respiratory 33.3 43.6 Respiratory 15—19 Respiratory 43.8 52.6 Respiratory 20—24 Respiratory 46.0 49.7 Respiratory 25—29 Respiratory 44.2 45.6 Respiratory 30—34 Respiratory 39.5 39.9 Respiratory 35—39 Respiratory 33.2 33.3 Respiratory 40—44 Respiratory 27.5 28.0 Respiratory 45—49 Respiratory 22.1 23.6 Respiratory 50—54 Alimentary tract 21.6 25.0 Circulatory 55—59 Alimentary tract 22.6 28.4 Circulatory 60—64 Circulatory 24.4 30.9 Circulatory 65—69 Circulatory 25.6 32.5 Circulatory 70—74 Circulatory 28.0 32.9 Circulatory 75—79 Circulatory 28.4 33.3 Circulatory 80—84 Circulatory 30.4 85—89 Circulatory 30.8 If our lungs were as organically good relatively as our hearts, having regard in each case for the work the organ is called upon to do and the conditions under which it must do it, we should live a considerable number of years longer on the average than we do now. One cannot but feel that the working out of a rational and scientifi- cally grounded system of personal hygiene of the respir- THE CAUSES OF DEATH 137 atory organs, on the broadest basis, to include all such matters as ventilation of buildings, etc., and the putting of such a personal hygiene into general use through education, would pay about as large dividends as could be hoped for from any investment in public health secu- rities. I am aware that much has already been done in this direction, but in order to reap any such dividends as I am thinking of, a vast amount must be added to our present knowledge of the physiology, pathology, epidemi- ology, and every other aspect of the functions and struc- tures of respiration. CHAPTER V EMBRYOLOGY AND HUMAN MORTALITY the preceding chapter attention was confined strictly to the organological incidence of death. It is possible to, push the matter of human mortality still farther back. In the embryological development of the vertebrate body, there are laid down at an early stage, in fact immediately following the process of gastrulation, three morphologically definite primitive tissue elements, called respectively the ectoderm, the mesoderm and the endoderm. These are termed the germ-layers, and em- bryological science has, for a great many forms, succeeded in a broad way in tracing back to the primitive germ layer from which it originally started its development, substantially every one of the adult organs and organ systems of the body. Itj makes no difference to the validity or significance of the discussion which we are about to enter upon, in what degree of esteem or contempt in biological philosophy the germ layer theory or doctrine, which oc- cupied so large a place in morphological speculation 50 years ago, may be held. We are here concenied only with the well-established broad descriptive fact, that in general all adult organ systems may be traced back over the path of their embryological development to the germ layer, or combination of germ layers, from which they origin- ally started. Having arranged, so far as possible, all causes of death on an organological basis, it occurred to me to go one 138 EMBRYOLOGY AND HUMAN MORTALITY 139 step further back and combine them under the headings of the primary germ layers from which the several organs developed embryologically. To do this was a task of considerable difficulty. It raised intricate, and in some TABLE 12 Showing the relative influence of the primary germ layers in human mortality (Items 64 and 65 charged to ectoderm) Locality Death rate per 100,000 due to functional breakdown of organs embryologically developing from Ecto- derm Per cent. Meso- derm Per cent. Endo- derm Per cent. United States Registration Area, 1906-10 191.1 210.6 177.1 134.9 14.3 15.0 14.4 8.4 425.2 407.1 374.0 468.0 31.8 29.0 30.3 29.0 719.6 786.2 681.5 1009.9 53.9 56.0 55.3 62.6 United States Registration Area, 1901-05 England and Wales, 1914. Sao Paulo. 1917 . TABLE 13 Showing the relative influence of the primary germ layers in human mortality (Items 64 and 65 charged to mesoderm) Death rate per 100,000 due to functional breakdown of organs embryologically developing from juuuaiiLy Ecto- derm Per cent. Meso- derm Per cent. Endo- derm Per cent. United States Registration Area, 1906-10 116 9 8 7 499 4 37 4 719 6 53 9 United States Registration Area, 1901-05 137 3 9 8 480 4 34 2 786 2 56 0 England and Wales, 1914. .. Sao Paulo. 1917 . 107.9 101.3 6.7 6.3 443.2 501.6 36.0 31.1 681.5 1009.9 55.3 62.6 cases still unsettled, questions of embryology. Further- more, the original statistical rubrics under which the data are compiled by registrars of vital statistics were never planned with such an object as this in mind. Still the thing seemed worth trying because of the biological interest which would attach to the result, even though it were some- 140 BIOLOGY OF DEATH what crude and, in respect of minor and insignificant details, open to criticism. It is not possible here to go into details as to how the causes of death were combined in - - 53.9 US REGISTRATION AREA 1306 -JO ENGLAND AND WALES 1914- SAO PAULO J9/7 ENDOD5RM MESODERM ECTODEffM FIG. 38. — Diagram showing the percentages of biologically classifiable human mortality resulting from breakdown of organs developing from the different germ layers. Upper bar of pair gives upper limit of mortality chargeable to ectoderm: lower bar gives lower limit of mortality chargeable to ectoderm. making up the final tables. For these details one must refer to the original papers. Tables 12 and 13, and Figure 38, give the results for the crude mortality of the U. S. Registration Area, Eng- land and Wales, and Sao Paulo, Brazil. EMBRYOLOGY AND HUMAN MORTALITY 141 The figures show that in man, the highest product of organic evolution, about 57 per cent, of all the biologically classifiable deaths result from a breakdown and failure further to function of organs arising from the endoderm in their embryological development, while but from 8 per cent, to 13 per cent, can be regarded as a result of breakdown of organ systems arising from the ectoderm. The remaining 30 to 35 per cent, of the mortality results from failure of mesodermic organs. The two values stated for ectoderm and mesoderm, shown by the two bars in the diagram, differ by virtue of the fact that two important causes of death, cerebral hemorrhage and apoplexy, and softening of the brain, are put in the one case with the ectoderm and in the other case with the mesoderm. The pathological arguments for the one disposition as against the other of these two diseases are interesting, but lack of space prevents their exposition here. I have chosen rather to present the facts in both ways. Taking a general view of comparative anatomy and embryology it is evident that in the evolutionary history through which man and the higher vertebrates have passed it is the ectoderm which has been most widely differ- entiated from its primitive condition, to the validity of which statement the central nervous system furnishes the most potent evidence. The endoderm has been least pro- gressively changed structurally and functionally in the process of evolution, while the mesoderm occupies, on the whole, an intermediate position in this respect. Degree of differentiation of organs in evolution im- plies degree of adaptation to environment. From the pre- sent point of view we see that the germ layer, the endo- derm, which has evolved or become differentiated least in 142 BIOLOGY OF DEATH the process of evolution is least able to meet successfully the vicissitudes of the environment. The ectoderm has changed most in the course of evolution. Of this the cen- tral nervous system of man is the best proof. There have also been formed in the process of differentiation, protective mechanisms, the skull and vertebral column, which very well keep the delicate and highly organized central nervous system away from direct contact with the environment. The skin also exhibits many differen- tiations of a highly adaptive nature to resist environmen- tal difficulties. It is then not surprising that the organ systems developed from the ectoderm break down and lead to death less frequently than any other. The fig- ures make it clear that man's greatest enemy is his own endoderm. Evolutionary speaking, it is a very old- fashioned and out-of-date ancestral relic, which causes him an infinity of trouble. Practically all public health ac- tivities are directed towards overcoming the difficulties which arise because man carries about this antediluvian sort of endoderm. We endeavor to modify the environ- ment, and soften its asperities down to the point where our own inefficient endodermal mechanism can cope with them, by such methods as preventing bacterial contam- ination of water, food and the like, warming the air we breathe, etc. But our ectoderm requires no such exten- sive amelioration of the environment. There are at most only a very few, if any, germs which can gain entrance to the body through the normal, healthy unbroken skin. We do, to be sure, wear clothes. But it is at least a debat- able question whether, upon many parts of the earth's surface, we, should not be better off without them from the point of view of health. These data indicate further in another manner how EMBRYOLOGY AND HUMAN MORTALITY 143 important are the fundamental embryological factors in determining the mortality of man. Of the three local- ities compared, England and the United States may be fairly regarded as much more advanced in matters of public health and sanitation than Sao Paulo. This fact is reflected with perfect precision and justice in the re- lative proportion of the death rates from endoderm and ectoderm. In the United States and England about 55 per cent, of the classifiable deaths are chargeable to endo- derm and about 9 to 14.5 per cent, to ectoderm. In Sao Paulo 62.6 per cent, fall with the endoderm, and but 6.3 to 8.4 per cent, with the ectoderm. Since public health measures can and do affect practically only the death rate chargeable to endoderm, this result, which is actually obtained, is precisely that which would be expected. A question which naturally occurs is as to what the age distribution of breakdown of ectodermic, mesoder- mic, or endodermic organs may be. Are the endodermic organs, for example, relatively more liable to breakdown in early life, and less so later, as general observation would lead one to conclude? To answer this and similar questions which come to mind it is necessary to distribute the specific rates of Table 9 upon an embryological basis. In Figure 39 the result of doing this is shown for males. We note that prior to age 60 the curve for the breakdown of organs of endodermic origin lies at the top of the diagram; next below it comes the curve for thq breakdown of organs of mesodermic -origin; and finally at the bottom the curve for the breakdown of or- gans of ectodermic origin. All three of the curves have in general the form of a specific death rate curve. The rates for all three germ layers are relatively high in in- 144 BIOLOGY OF DEATH fancy and drop at a practically constant rate to a low point in early youth. In infancy the heaviest mortality in males is due to the breakdown of organs of endodermic rlfOOO rz too O § 10 E O.I 0 5 1O 15 20 25 3O 35 40 45 £0 AGE 6O 65 7O '/5 8O &5 9O 95 IOO FIG. 39. — Showing specific death rates in males according to the germ layer from which the organs developed. origin. This part of the death rate accounts for some- thing like 10 times as many deaths as either mesoderm or ectoderm at this period of life. From about age 12 on in EMBRYOLOGY AND HUMAN MORTALITY 145 the case of organs of ectodermic origin, and from about age 22 on in cases of mesodermic origin, the death rate curves rise at a practically constant rate to extreme old age. The ectodermic and mesodermic curves during this portion of the life span are nearly parallel, diverging only slightly from each other with advancing age. The curve for the death rate resulting from breakdown of organs of endodermic origin has an entirely different course. It rises sharply for ten years after the low point in early youth, and then makes a rather sharp bend at about age 22, and passes off to the end of the life span, at a reduced rate of change. In consequence of this it crosses the mesodermic line at age 60. From that point on to the end of life deaths from breakdown of organs of mesodermic origin stand first in importance. Figure 40 shows the same set of facts for the female, and at once a number of striking differences between the conditions in the two sexes appear. In the first place, the breakdown of mesodermic organs is practically of equal importance in determining the mortality of infants with the breakdown of endodermic organs, in the case of the female. This fact, of course, arises because of the heavy mortality of infancy due to failure of the female reproductive organs, a matter which has already been discussed. The curve for breakdown of the ectodermic organs follows substantially the same kind of course in the female as it does in the male. The mesoderm and endoderm lines cross nearly 20 years earlier in the case of females than in the males. This circumstance arises from the fact that throughout life the mesodermic organs play a relatively more important role in the determina- tion of mortality in the female than they do in the male. What reward in the way of useful generalization may 10 146 BIOLOGY OF DEATH be claimed from the details reviewed in this and the pre- ceding chapter? I hope that these facts will have served in some measure to complete and round out in clearer l.OOQ 100 5 FEMALL ss s 9 5 A y* r^ 4 // j ^** I- ^ /g / / i i i \ 4 ^ / 7 /t/ i 9 & f^j y^ i f 1 \ s£ f i \\ ' ^ S^.' / 1 i s£s f i i **1 .— ' ^ 9** / \l t i /] ,'" S •' A * 1 \ \ 1 / t 4 0^ I X ' \ I/ J •N ^ * ^s J* % i^r \ s \ / r s s* ^ O.I f \»- 'x 0 C/ /V W £.U £O Ol/ ^K> IV ^ AGE FIG. 40. — Showing specific death rates for females, classified in the same manner as in Fig. 39. outlines one part of the picture of the general biology of death. It has been shown in what has preceded that nat- ural death is not a necessarv or inherent attribute or EMBRYOLOGY AND HUMAN MORTALITY 147 consequence of life. Many cells are potentially immor- tal and the potentiality is actually realized if appropriate conditions are provided. Protozoa are immortal. Germ cells are immortal. Various somatic cells, and even tis- sues have been proved to be potentially immortal by demonstrating in a variety of ways that under appro- priate conditions they continue to live indefinitely. This is the lesson taught us on the one hand by successive transplantations of tumor cells, which are only modified somatic cells, and on the other hand by successful cul- ture of many sorts of somatic cells in vitro. Analytical consideration of the matter shows very clearly that the somata of multicellular organisms die because of the differentiations and specializations of structure and function which they exhibit in their make-up. Certain cells are differentiated to carry on certain specialized functions. In this specialization they forego their power of independent and indefinitely con- tinued existence. The cells lining the lungs, for example, must depend in the body upon the unfailing normal ac- tivity of the cells of the alimentary tract and the blood in order that they, the epithelial cells of the lungs, may get proper nutrition. If in such an interlocking and mu- tually dependent system any one part through accident or in any way whatever gets deviated from its normal functioning, the balance of the whole system is upset. If the departure of any part from its normal functional course is great enough to be beyond correction promptly through the normal regulatory powers of the organism, death of the whole will surely ensue. What I have tried to show in this and the preceding chapter is a quantitative picture of how the different organ systems get out of balance, and wreck the whole 148 BIOLOGY OF DEATH machine. The broad orderliness and lawfulness of the whole business of human mortality is impressive. We have seen that different organ systems have well-defined times of breakdown. Or, put in another way, we see that in the human organism, just as in the automobile, the serviceability of the different parts varies greatly. The heart outwears the lungs, the brain outwears both. But we have further, I believe, got an inkling of the funda- mental reason why these things are so. It is broadly speaking, because evolution is a purely mechanistic pro- cess instead of being an intelligent one. All the parts are not perfected by evolution to even an approximately equal degree. It is conceivable that an omnipotent person could have made a much better machine, as a whole, than the human body which evolution has produced, assuming, of course, that he had first learned the trick of making self-regulating and self-reproducing machines, such as living machines are. He would presumably have made an endoderm with as good resisting and wearing qualities as the mesoderm or ectoderm. Evolution by the hap- hazard process of trial and error which we call natural selection, makes each part only just good enough to get by. In the very nature of the process itself it cannot possibly do anything any more constructive than this. The workmanship of evolution, from a mechanical point of view, is extraordinarily like that of the average automobile repair man. If evolution happens to be fur- nished by variation with fine materials, as in the case of the nervous system, it has no objection to using them, but it is equally ready to use the shoddiest of endoderm provided it will hold together just long enough to get the machine by the reproductive period. It furthermore seems to me that the results presented EMBRYOLOGY AND HUMAN MORTALITY 149 in this chapter add one more link to the already strong chain of evidence which indicates the highly important part played by innate constitutional biological factors as contrasted with environmental factors in the deter- mination of the observed rates of human mortality. Here we have grouped human mortality into broad classes which rest upon a strictly biological basis. When this is done it is found that the proportionate subdivision of the mortality among the several causes — in short the death ratios in the sense of Fisher — is strikingly similar in such widely dissimilar environments as the United States, England and Southern Brazil. CHAPTER VI THE INHERITANCE OF DURATION OF LIFE IN MAN WE have seen that in the case of man, where alone quantitative data are available, the breakdown of partic- ular organ systems, and consequent death of the whole, occurs in a highly orderly manner in respect of time or age. Each organ system has a characteristic time curve for its breakdown, differing from the curve of any other system. The problem which now confronts us is to find out what lies back of these characteristic time curves and determines their form. In view of the biological facts about death which we have learned, what determines that John Smith shall die at 58, while Henry Jones lives to the obviously more respectable age of 85? We have seen that there is every reason to believe that all the essential cells of both their bodies are inherently capable under proper conditions of living indefinitely. It fur- ther appears probable that it is the differentiated and specialized structure of their bodies which prevents the realization of these favorable conditions. But all this helps us no,t at all to understand why in fact one lives nearly 30 years longer than the other. It may help to visualize this problem of the determina- tion of longevity to consider an illustrative analogy. Men behave in respect of their duration of life not unlike a lot of eight-day clocks cared for by an unsystematic person, who does not wind them all to an equal degree and is not careful about guarding them from accident. Some he winds up fully, and they run their full eight days. 150 THE INHERITANCE OF DURATION 151 Others he winds only halfway, and they stop after four days. Again the clock which has been wound up for the full eight days may fall off the shelf and be brought to a stop at the third day. Or someone may throw some sand in the works when the caretaker is off his guard. So, similarly, some men behave as though they had been wound up for a full 90-year run, while others are but partially wound up and stop at 40 or 65, or some other point. Or, again, the man wound up for 80 years may, like the clock, be brought up much short of that by an accidental invasion of microbes, playing the role of the sand in the works of the clock. It is of no avail for either the clock or the man to say that the elements of the mechanism are in whole or in major part capable of fur- ther service. The essential problem is : what determines the goodness of the original winding? And what rela- tive part do external things play in bringing the running to an end before the time wThich the original winding was good for? It is with this problem of the winding up and running of the human mechanism that the present chap- ter will deal. There are two general classes of factors which may be involved here. These are, on the one hand, heredity and, on the other hand, environment, using the latter term in the broadest sense. Inasmuch as we can be reason- ably sure on a priori grounds that longevity, like most other biological phenomena, is influenced by both hered- ity and environment the problem practically reduces itself to the measuring of the relative importance of each of these two factor groups in determining the results we see. But before we start the discussion of exact measurements in this field let us first examine some of the general evi- 152 BIOLOGY OF DEATH dence that heredity plays any part at all in the deter- mination of longevity. THE HYDE FAMILY The first material which we shall discuss is that pro- vided by the distinguished eugenist, Dr. Alexander Graham Bell, in his study of the Hyde family. Every genealogist is familiar with the ' * Genealogy of the Hyde Family/' by Reuben H. Walworth. It is one of the fin- est examples in existence of careful and painstaking genealogical research. Upon the data included in this book, Bell has made a most interesting and penetrating analysis of the factors influencing longevity. At first thought one might conclude that highly biased results would probably flow from the consideration of only one family. Bell meets this point very well, however, in the following words : A little consideration will show that the descendants did not constitute a single family at all, and indeed had very little of the Hyde blood in them. Even the children of William Hyde owed only half of their blood to him, and one-half to his wife. The grandchildren owed only one-quarter of their blood to William Hyde, and three-quarters to other people, etc. The descendants of the seventh generation, and there are hundreds of them, owed only one sixty-fourth of their blood to William Hyde, and sixty-three sixty-fourths to the new blood introduced through successive generations of marriages with persons not of the Hyde blood at all. It will thus be seen that the thousands of descendants noted in the Hyde Genealogy constitute rather a sample of the general population of the country than a sample of a particular family in which family traits might be expected to make their appearance. The substantial normality of the material is shown in Figure 41, which gives the lx line, that is, the number of survivors at each age, of the 1,606 males and 1,352 females for whom data were available. The solid line is the male lx line and the dotted line the female lx . It is at once apparent that the curves have the same general THE INHERITANCE OF DURATION 153 sweep in their passage over the span of life as has the general population life curve discussed in the preceding chapter. The descent is a little steeper in early adult life. The female curve differs in two respects from the normal general population curves. In the first place, roo fl 90 \ &O | ^ ^ 60 ^1 Nj I \ 1 SURVIVING § 1 ^ \^ \ * \ 1 I V L \ JS fc i 0. •»/-> \ \ \ \ zo j-^K i ^K o rV>- B 5 10 15 ZO 25 30 35 4O 45 SO 55 6O 65 70 75 SO 85 SO 95 AGE: FIG. 41. — Showing survival curves of members of the Hyde family (Plotted from Bell's data). beginning at age 15 and continuing to age 90, the female curve lies below that for the males, whereas normally for the general population it lies above it. This denotes a shorter average duration of life in the females than in the males, the actual figures being 35.8 years for the males and 33.4 years for the females. Bell attributes the dif- ference to the strain of child-bearing by the females in 154 BIOLOGY OF DEATH this rather highly fertile group of people, belonging in the main to a period when restrictions upon size of family were less common and less extensive than now. In the second place, the female lx curve is actually convex to the base throughout a considerable portion of middle life whereas, normally, this portion of the curve presents a concave face to the base. Apart from these deviations, which are of no partic- ular significance for the use which Bell makes of the data, the Hyde material is essentially normal and simi- lar to what one would expect to find in a random sample of the general population. In this material there were 2.287 cases in which the ages at death of the persons and the ages at death of their fathers were known. It occurred to Bell to arrange this material in such a way as to show what, if any, relation existed between age at death of the parent and that of the offspring. He arranged the parents into four groups, according to theiage at which they died, and the offspring into five groups upon the same basis. In the case of the parents the groups were : First, those dying under 40 ; second, between 40 and 60 ; third, between 60 and 80 ; and fourth, at age 80 and over. The groups for the offspring were the same, except that the first was divided into two parts, namely, those dying under 20 and those dying between 20 and 40. The result- ing figures are exhibited in Table 14. The results for father and offspring are shown in Figure 42, based upon the data of Table 14. In each of the 5 polygons, one for each offspring group, the first dot shows the percentage of fathers dying under 40; the second dot the percentage of fathers dying between 40 and 60 ; and so on, the last dot in each curve showing the percentage of fathers dying at age 80 and over. It THE INHERITANCE OF DURATION TABLE 14 155 Analysis of the Hyde family data by person's age at death, showing the number and percentage having (a) fathers and (6) mothers who died at the age periods named. (From Bell} Person's age at death Father1 s age at death Stated -40 40-60 60-80 80 + Stated Under 20 and 40 and 60 and 80 and 2,287 669 538 467 428 185 66 20 18 12 13 3 522 189 140 116 57 20 1,056 299 269 215 196 77 643 161 111 124 162 85 20... under under under over . 40 60 80 Percentages Stated Under 20 and 40 and 60 and 80 and 100.0 100.0 100.0 100.0 100.0 100.0 2.9 3.0 3.4 2.6 3.0 1.6 22.8 28.2 26.0 24.8 13.3 10.8 46.2 44.7 50.0 46.0 45.8 41.6 28.1 24.1 20.6 26.6 37.5 46.0 20 ... under under under over. 40 60 80 Person's age at death Mother 's age at death Stated -40 40-60 60-80 80 + Stated Under 20 and 40 and 60 and 80 and 1,805 511 407 379 360 148 191 88 42 27 26 8 435 129 104 92 80 30 713 199 176 159 129 50 466 95 85 101 125 60 20 . .. under under under over . 40 60 80 Percentages Stated Under 20 and 40 and 60 and 80 and 100.0 100.0 100.0 100.0 100.0 100.0 10.6 17.2 10.3 7.1 7.2 5.4 24.1 25.2 25.6 24.3 22.2 20.3 39.5 39.0 43.2 42.0 35.9 33.8 25.8 18.6 20.9 26.6 34.7 40.5 20 ... under under under over . 40 60 80 156 BIOLOGY OF DEATH is to these last dots that attention should be particularly directed. It will be noted that the dotted line connecting the last dots of each of the 5 polygons in general rises as we pass from the left-hand side of the diagram to the right-hand side. In the case of offspring dying under 20, 24 per cent, of their fathers died at ages over 80. About 669 PERSONS DIED -zo 538 PERSONS DILD 20- 4O 46f PERSONS DIED 4O-CO 426 PERSONS DIED 60-60 105 PERSONS DILD 60-t 30 20 10 40 10 v - 4O 60 8O - 4O 6O 80 - 40 6O 8O - 4O 6O 80 - 4O 6O 8O 40 6O 6O + 40 60 6O + 4O 6O 60 '• + 4O 6O 80 + 40 60 8O + FIG. 42. — Influence of father's age at death upon longevity of offspring. First dot in each diagram shows the percentage having fathers who died at 40; second dot the percent- age having fathers who died from 40-60 ; third dot the percentage having fathers who died from 60-80; fourth dot the percentage having fathers who died 80+ (After Bell). 21 per cent, of the fathers of offspring dying between 20 and 40 lived to be 80 years or over. For the next longer- lived group of offspring, dying between 40 and 60, the percentage of fathers living to 80 or over rose to 27 per cent. In the next higher group, the percentage is nearly 38, and finally the extremely long-lived group of offspring, the. 185 persons who died at ages of 80 and over, had 46 per cent, or nearly one-half of their fathers living to the same great age. In other words, we see in general that the longer-lived a group of offspring is, on the average, the longer-lived are their fathers, on the average; or, put in another way, the higher the percentage of very THE INHERITANCE OF DURATION 157 long-lived fathers which this group will have as com- pared with shorter-lived individuals. Figure 43 shows the same sort of data for mothers and offspring. Here we see the curve of great longevity of parents rising in an even more marked manner than was the case with fathers of offspring. The group of 50 30 20 10 511 PERSONS DIED -^o 407 379 PERSONS DlLD 40-6O 360 PERSONS DlLD 60-80 148 PERSONS D/£D 6O + 50 40 30 10 40 60 80 - 4O 6O 8O - 4O 6O 8O - 4O 4O 6O 8O + 4O 6O 80 + 4O 6O BO + 4O 6O 6O &O - 4O 6O 80 8O + 4O 6O GO + FIG. 43. — Influence of mother's age at death upon longevity of offspring. First dot in each diagram shows the percentage having mothers who died at 40; second dot the per- centage having mothers who died at 40-60; third dot the percentage having mothers who died 60-80; fourth dot the percentage having mothers who died 80+ (After Bell). offspring dying at ages under 20 had only 19 per cent. of their mothers living to 80 and over, whereas the group of offspring who lived to 80 and beyond had 41 per cent, of their mothers attaining the same great age. At the same time we note from the dotted line at the bot- tom of the chart that as the average age at death of the offspring increases, the percentage of mothers dying at early ages, namely, under 40, as given by the first dots, steadily decreases from 17 per cent, at the first group to just over 5 per cent, for the offspring dying at very advanced ages. 158 BIOLOGY OF DEATH These striking results demonstrate at once that there is a definite and close connection between the average longevity of parents and that of their children. Ex- tremely long-lived children have a much higher percent- age of extremely long-lived parents than do shorter lived children. While the diagrams demonstrate the fact of this connection, they do not measure its intensity with as great precision as can be obtained by other methods of dealing with the data. A little farther on we shall take up the consideration of this more precise method of measurement of the hereditary influence in respect of longevity. In the preceding diagrams we have considered each parent separately in connection with the offspring in TABLE 15 Longevity of parents of persons dying at 80 and over. (From Bell) Age at death of parents Number of persons Number of persons lived 80 + Per cent, of persons lived 80 + Stated 1,594 139 8 7 Lived to be 80+ Neither parent 827 44 5 3 One parent (not other) .... 583 57 9 8 Both parents 184 38 20.6 Father (not mother) 337 38 11 3 Mother (not father) 246 19 7 7 regard to longevity. We shall, of course, get precisely the same kind of result if we consider both parents to- gether. For the sake of simplicity, taking only the cases of extreme longevity, namely, persons living to 80 or over — the essential data are given in Table 15. From this table it is seen that where neither parent lived to be 80, only 5.3 per cent, of the offspring lived to be 80 or over, the percentage being based upon 827 THE INHERITANCE OF DURATION 159 cases. Where one parent, but not the other, lived to be 80 or older, 9.8 per cent, of the offspring lived to be 80 or older, the percentage here being based upon 583 cases. Where both parents lived to be 80 or older 20.6 per cent, of the persons lived to the same great age, the percentage be- ing based upon 184 cases. Thus it appears that in this group of people four times as many attained great longev- ity if both their parents lived to an advanced age, as attained this age when neither parent exhibited great longevity. The figures from the Hyde family seem fur- ther to indicate that the tendency of longevity is inherited more strongly through the father than through the mother. Where the father, but not the mother, lived to be 80 or older, 11.3 per cent, of the persons lived to age 80 or more, there being 337 cases of this kind. Where the mother, but not the father lived to be 80 or older, only 7.7 per cent., or nearly 4 per cent, fewer of the persons lived to the advanced age of 80 or more, there being 246 cases of this sort. Too much stress is not, however, to be laid upon this parental difference because the samples after all are quite small. One other point in this table deserves consideration. Out of the 1,594 cases as a whole, less than 9 per cent, of the persons lived to the advanced age of 80 or more. But out of this number there are 767, or 48.1 per cent., nearly one-half of the whole, who had parents who lived to 80 or more years. Another interesting and significant way in which one may see the great influence of the age of the parents at death upon the longevity of the offspring, is indicated in Table 16, where we have the average duration of life of individuals whose fathers and mothers died at the specified ages. 160 BIOLOGY OF DEATH We see that the longest average duration of life, or expectation of life, was of that group which had both mothers and fathers living to age 80 and over. The average duration of life of these persons was 52.7 years. Contrast this with the average duration of life of those whose parents both died under 60 years of age, where TABLE 16 Showing the influence of a considerable degree of longevity in both father and mother upon the expectation of life of the offspring. (After Bell). (In each cell of the table the open figure is the average duration of life of the offspring and the bracketed figure is the number of cases upon which the average is based). Father's age at death Mother's age at death Under 60 60-80 Over 80 Under 60 32. 8 years (128) 33 . 4 years (120) 36 . 3 years (74) 60-80 35.8 (251) 38.0 (328) 45.0 (172) Over 80 42.3 (131) 45.5 (206) 52.7 (184) the figure is 32.8 years. In other words, it added al- most exactly 20 years to the average life of the first group of people to have extremely long-lived parents, instead of parents dying under age 60. In each column of the table the average duration of life advances as we proceed from top to bottom — that is, as the father 's age at death increases — and in each row of the table the aver- age expectation of life of the offspring increases as we pass from left to right — that is, with increasing age of the mother at death. However the matter is taken, a careful selection of one's parents in respect of longevity is the most reliable form of personal life insurance. THE INHERITANCE OF DURATION 161 How great and deep is the significance of the facts shown in Table 16 may best be brought home to the mind by means of a comparison. Suppose this question to be asked: by how great an amount would the average expectation of life at birth (which in a stable population is the same thing as the mean duration of life) be increased if all the reasonably preventable deaths were prevented? If, say 75 per cent, of all the deaths from pulmonary tuber- culosis did not occur; if 40 per cent, of the deaths from Bright 's disease were prevented; and, in general, if all that medicine and hygiene knows today were put into reasonably effective operation, and nobody died except when and from such causes as could in no way be influenced by what medical science, good envi- ronment, etc., have to offer: by how much then would the expectation of life be greater than it now is? We have seen that to have one's parents live to 80 or over increases the expectation of life 20 years, as compared with that of persons whose parents die under 60 years of age. By how much more would the expectation of life be extended if all reasonably preventable deaths were prevented? A thorough and critical answer to this question is afforded by an investigation of Forsyth's:, conducted along the most exact and approved actuarial lines. Some years ago, Professor Irving Fisher sent a list of some 90 diseases to a group of the most prominent medi- cal authorities in this country, and asked them to desig- nate what percentage of the deaths due to each disease they considered preventable. The results of this inquiry were tabulated in an extremely conservative manner, with the result set forth in Table 16a, which is copied from Forsyth's paper (pp. 762-763). 11 162 BIOLOGY OF DEATH TABLE 16 a Showing Fisher's ratios of preventability for the diseases enumerated in the mortality statistics of the United States, together with the relative importance of each disease as indicated by the percentage the number of its deaths bears to the total number of deaths Causes of death 1 Premature birth 2 Congenital malformation of the heart 3 Other congenital malformations 4 Congenital debility 5 Hydrocephalus 6 Venereal diseases 7 Diarrhoea and enteritis 8 Measles 9 Acute bronchitis 10 Bronchopneumonia 11 Whooping cough 12 Croup 13 Meningitis 14 Diseases of larynx — not laryngitis 15 Laryngitis 16 Diphtheria 1 7 Scarlet fever 18 Diseases of lymphatics 19 Tonsillitis 20 Tetanus 21 Tuberculosis — not of lungs 22 Abscess 23 Appendicitis 24 Typhoid fever 25 Puerperal convulsions 26 Puerperal septicaemia 27 Other diseases of childbirth 28 Diseases of tubes 29 Peritonitis 30 Smallpox 31 Tuberculosis of lungs 32 Violence 33 Malarial fever 34 Septicsemia 35 Epilepsy 36 General, ill-defined, and unknown causes (in- cluding "heart failure," "dropsy," and "con- vulsions") , 37 Erysipelas 38 Pneumonia (lobar and unqualified) 39 Acute nephritis 40 Pleurisy 41 Acute yellow atrophy of liver , 42 Obstructions of intestines. . Prominence of disease. Per cent. of all deaths 2.0 .55 .3 2.3 .1 .3 7.74 .8 1.1 2.4 .9 .3 1.6 .07 .06 1.4 .5 .01 .05 .19 .17 .08 .7 2.0 .2 .4 .36 .1 .5 .01 9.9 7.5 .2 .3 .29 9.2 .3 7.0 .6 .27 .02 .6 Ratio of preventability. Per cent. 40 0 0 40 0 70 60 40 30 50 40 75 70 40 40 70 50 20 45 80 75 60 50 85 30 85 50 65 55 75 75 35 80 40 0 30 60 45 30 55 0 25 THE INHERITANCE OF DURATION TABLE 16 a— Continued 163 Causes of death 43 Alcoholism 44 Hemorrhage of lungs 45 Diseases of the thyroid body . . . 46 Ovarian tumor 47 Uterine tumor 48 Rheumatism 49 Gangrene of lungs 50 Anaemia, leukaemia 51 Chronic poisonings 52 Congestion of lungs 53 Ulcer of stomach 54 Carbuncle 55 Pericarditis 56 Cancer of female genital organs 57 Dysentery 58 Gastritis 59 Cholera nostras 60 Cirrhosis of liver 61 General paralysis of insane 62 Hyatid tumors of liver 63 Endocarditis 64 Locomotor ataxia 65 Diseases of veins 66 Cancer of breast 67 Diabetes 68 Biliary calculi 69 Hernia 70 Cancer not specified 71 Tumor 72 Bright's disease 73 Embolism and thrombosis 74 Cancer of intestines 75 Cancer of stomach and liver. 76 Calculi of urinary tract 77 Cancer of mouth 78 Heart disease , 79 Influenza 80 Asthma and emphysema 81 Angina pectoris 82 Apoplexy 83 Cancer of skin 84 Chronic bronchitis 85 Paralysis 86 Softening of brain 87 Diseases of arteries 88 Diseases of bladder 89 Gangrene 90 Old age Prominence of disease. Per cent of all deaths 8 2. .4 .1 .02 .07 .1 .5 .03 .4 .05 .4 .2 .03 .1 .6 .5 .65 .09 .9 .3 .002 .8 .17 .04 .4 .8 .17 .27 .9 .08 i.6 .26 .55 ^ . / .03 .1 .1 .7 .23 .4 .4 .2 .8 .0 .2 .83 .2 .25 0 Ratio of preventability. Per cent. 85 80 10 0 60 10 0 50 70 50 50 50 10 0 80 50 50 60 75 75 25 35 40 0 10 40 70 0 0 40 0 0 0 10 0 25 50 30 25 35 0 30 50 0 10 45 60 0 164 BIOLOGY OF DEATH It will be seen that these ratios of preventability are not all 100 per cent. They are not the wild overstate- ments of the propagandist. But they do represent, if they could be realized, substantial reductions from exist- ing mortality rates. TABLE 16 b Complete expectations of life as based upon the two assumptions that deaths are and are not prevented according to the ratios given in Table 16a Age Deaths Loss in Age Deaths Loss in Not pre- vented Pre- vented Years Days Not pre- vented Pre- vented Years Days 0 49.44 56.03 56.84 56.64 56.15 55.51 54.81 54.06 53.26 52.43 51.57 50.69 49.80 48.91 48.03 47.15 46.31 45.50 44.71 43.93 43.15 42.37 41.60 40.83 40.07 62.11 66.26 66.28 65.67 64.94 64.13 63.27 62.42 61.54 60.63 59.72 58.79 57.86 56.80 56.00 55.07 54.16 53.26 52.36 51.48 50.59 49.70 48.82 47.94 47.06 12 10 9 9 8 8 8 8 8 8 8 8 8 7 7 7 7 7 7 7 7 7 7 7 6 245 84 161 11 288 226 168 131 102 73 55 37 22 321 354 336 310 277 237 201 161 120 80 40 261 25 39.31 38.56 37.82 37.08 36.34 35.61 34.88 34.15 33.42 32.69 31.96 31.23 30.50 29.77 29.03 28.30 27.57 26.85 26.12 25.40 24.68 23.97 23.26 22.56 21.87 46.18 45.31 44.45 43.58 42.72 41.86 41.01 40.15 39.30 38.46 37.61 36.76 35.92 35.08 34.24 33.40 32.57 31.74 30.91 30.09 29.28 28.47 27.67 26.87 26.09 6 6 6 6 6 6 6 6 5 5 5 5 5 5 5 5 5 4 4 4 4 4 4 4 4 318 274 230 183 139 91 47 0 321 281 237 193 153 113 77 37 0 325 288 252 219 183 150 113 80 1 26 2 27 3 28 4 29 5 30 6 31 7 32 8 33 9 34 10 35 11 36 12 .... 37 13 38 14 39 15 40 16 41 17 42 18 43 19 44 20 45 21 46 22 47 23 48 24.. 49.. On the basis of the mortality experience of the Regis- tration Area for 11 years (1900-1910) Forsyth calculated mortality tables on the assumption that the ratios of preventability of Table 16a were actually in full opera- tion. The results, so far as concerns expectation of life, are set forth in Table 16b. THE INHERITANCE OF DURATION 165 Prom the first line of this table it is perceived that the total increase in expectation of life which would result if Fisher's ratios of preventability were fully realized is just under 13 years! How unfavorably this contrasts with the 20 years increase shown by the two TABLE 16 b— Continued Age Deaths Loss in Age Deaths Loss in Not pre- vented Pre- vented Years Days Not pre- vented Pre- vented Years Days 50. ... 21.17 20.47 19.78 19.09 18.40 17.74 17.08 16.45 15.83 15.23 14.63 14.05 13.48 12.92 12.36 11.82 11.29 10.77 10.26 9.77 9.29 25.30 24.52 23.74 22.97 22.21 21.46 20.72 20.00 19.30 18.61 17.93 17.27 16.61 15.96 15.32 14.69 14.07 13.47 12.87 12.29 11.71 4 4 3 3 3 3 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 47 18 350 321 296 263 234 201 193 139 110 80 47 15 350 348 285 256 223 190 1/53 71 8.82 8.36 7.93 7.50 7.09 6.70 6.31 5.98 5.64 5.32 5.02 4.74 4.47 4.23 4.01 3.79 3.58 3.39 3.22 3.06 11.15 10.59 10.04 9.51 8.99 8.49 8.00 7.53 7.07 6.63 6.20 5.78 5.38 4.99 4.62 4.25 3.89 3.56 3.27 3.06 2 2 2 2 1 1 1 1 1 1 1 1 120 84 40 4 329 288 252 201 157 113 66 15 332 277 223 168 113 62 18 0 51 72 52 73 53 74 54 75 55 76 .... 56 77. . . 57 78 58 79. . 59 80 60 81 61 82 62 83 63 84 64 85 65 86 66 87 67 88 68 89 69 90 70.. corner diagonal cells of Table 16! No more striking demonstration could be found of the overwhelming im- portance of heredity in determining duration of life. For if all the deaths which reason will justify one in suppos- ing preventable on the basis of what is now known, were prevented in fact the resulting increase in expectation of life falls seven years short of what might reasonably be expected to follow the selection of only one generation of ancestry (the parental) for longevity. So much for Bell's analysis of longevity in the Hyde 166 BIOLOGY OF DEATH family. We have seen that it demonstrates with the ut- most clearness and certainty that there is an hereditary influence between parent and offspring affecting the ex- pectation of longevity of the latter. Bell's method of handling the material does not provide any precise meas- ure of the intensity of this hereditary influence, nor does it furnish any indication of its strength in any but the direct line of descent. Of course, if heredity is a factor in the determination of longevity we should expect its effects to be manifested as between brothers and sisters, or in the avuncular relationships, and in greater or less degree in all the other collateral and more remote direct degrees of kinship. Happily, we have a painstaking analysis, with a quantitative measure of the relative in- fluence of heredity in the determination of longevity, which was carried out many years before Bell's work on the Hyde family, by the pioneer in this field, Prof. Karl Pearson. His demonstration of the inheritance of longev- ity appeared more than twenty years before that of Bell. I have called attention to the latter 's work first merely because of the greater simplicity and directness of his demonstration. We may now turn to a consideration of Pearson's more detailed results. PEAKSON'S WORK The material used by Pearson and his student, Miss Beeton, who worked with him on the problem, came from a number of different sources. Their first study dealt with three series from which all deaths recorded as due to accident were excluded. The first series included one thousand cases of the ages of fathers and sons at death, the latter being over 22.5 years of age, taken THE INHERITANCE OF DURATION 167 from Foster's "Peerage." The second series consisted of a thousand pairs of fathers and sons, the latter dying beyond the age of 20, taken from Burke 's "Landed Gentry.' The third series consisted of ages at death of one thousand pairs of brothers dying beyond the age of 20 taken from the "Peerage." It will be noted that all these series considered in this first study dealt only with inheritance in the male line. The reason for this was simply that in such books of record as the "Peerage" and "Landed Gentry'1 sufficiently ex- act account is not given of the deaths of female relatives. In a second study the material was taken from the pedi- gree records of members of thd English Society of Friends and from the Friends' Provident Association. This ma- terial included data on inheritance of longevity in the female line and also provided data for deaths of infants, which were lacking in the earlier used material. The investigation was grounded upon that important branch of modern statistical calculus known as the method of correlation. For each pair of relatives between whom it was desired to study the intensity of inheritance of longe- vity a table of double entry was formed, like the one shown here as Table 17. The figures in each cell or compartment of this table denote the frequency of occurrence of pairs of fathers and adult sons having respectively the durations of life indicated by the figures in the margins. Thus we see, examining the first line of the table, that there were 11 cases in which the average duration of life of the father was 48 years and that of the adult son 23 years. Farther down and to the right in the table there were 13 cases in which the average duration of life of the father and the son was in each case 83 years. These cases are men- 168 BIOLOGY OF DEATH tioned merely as illustrations. The whole table is to be read in the same manner. From such a table as this it is possible to calculate, by well-known mathematical methods, a single numerical constant of somewhat unique properties known as the TABLE 17 Correlation table showing the correlation between father and son in respect of duration of life DURATION OF LIFE OF FATHER 23 28 33 38 43 48 53 58 63 68 73 78 83 88 93 98 103 Totals 23 1 1 2 5 3 11 6 7 11 9 6 12 8 2 2 86 28 1 6 4 5 12 15 10 13 10 7 1 1 85 I 33 1 2 2 5 7 8 7 10 7 8 8 4 1 70 38 1 1 2 2 8 5 3 9 11 11 9 5 2 1 70 8 43 1 1 5 1 5 6 11 10 10 17 5 72 H 48 1 1 2 5 5 4 6 9 12 15 5 3 68 £ 53 1 3 5 7 3 2 11 11 14 10 1 1 1 70 ^ 58 1 3 4 5 10 8 10 5 8 9 3 2 68 £ 63 2 1 3 5 1 4 8 13 9 11 11 11 5 84 68 1 6 3 6 7 5 5 6 14 16 12 7 2 90 o 73 1 2 1 6 5 4 7 9 10 14 13 8 8 1 1 90 $ 78 1 1 2 2 4 4 4 10 5 8 9 4 3 57 tt 83 1 1 5 3 1 2 3 7 10 13 3 2 2 53 g 88 1 2 3 1 4 7 5 1 2 2 28 93 1 2 2 5 98 1 1 1 1 4 Totals 1 8 9 30 26 65 70 76 90 122 131 153 132 53 18 15 1 1000 coefficient of correlation, which measures the degree of association or mutual dependence of the two variables included in such double entry tables. This coefficient measures the amount of resemblance or association be- tween characteristics of individuals or things. It is stated in the form of a decimal which may take any value between 0 and 1. As the correlation coefficient rises to 1 we approach a condition of absolute dependence of the variables one upon the other. As it falls to zero we approach a condition of absolute independence, where the one variable has no relation to the other in the amount THE INHERITANCE OF DURATION 169 or direction of its variation. The significance of a cor- relation coefficient is always to be judged, in any partic- ular case, by the magnitude of a constant associated with it called the probable error. A correlation coeffi- cient may be regarded as certainly significant when it has a value of 4 or more times that of its probable error, which is always stated after the coefficient with a com- bined plus and minus sign between the two. The coeffi- cient is probably significant when it has a value of not less than 3 times its probable error. By "significant' in this connection is meant that the coefficient probably is not merely a random chance result. In Table 18 are the numerical results from the first study based upon the "Peerage" and "Landed Gentry.' TABLE 18 Inheritance of duration of life in male line. Data from "Peerage" and "Landed Gentry." (Beeton and Pearson). Ratio of Relatives Correlation coefficient to coefficient its probable error X y Txy r*y + Er Father ("Peerage") Son, 25 years and over .115 ±.021 5.5 Father ("Landed Gentry") Son, 20 years and over .142 ±.021 6.8 Father ("Peerage") Son, 52.5 years and over .116 ±.023 5.0 Father ("Landed Gentry") Son, 50 years and over .113 ±.024 4.7 Brother ("Peerage") Brother .260 ± .020 13.0 It is seen at once that all of the coefficients are signifi- cant in comparison with their probable errors. The last column of the table gives the ratio of the coefficient to its probable error, and in the worst case the coefficient is 4.7 times its probable error. The odds against such a correlation having arisen from chance alone are about 655 to 1. Odds such as these may be certainly taken as demonstrating that the results rep- 170 BIOLOGY OF DEATH resent true organic relationship and not mere chance. All of the other coefficients are certainly significant, hav- ing regard to their probable errors. Furthermore, they are all positive in sign, which implies that a variation in the direction of increased duration of life in one relative of the pair is associated with an increase in expectation of life in the other. It will be noted that the magnitude of the correlation between brother and brother is about twice as great as in the case of correlation of father with son. From this it is provisionally concluded that the intensity of the hereditary influence in respect of duration of life is greater in the fraternal relationship than in the parental. It evidently makes no difference, broadly speaking, so far as these two sets of material are con- cerned, whether there are included in the correlation table all adult sons, whatever their age, or only adult sons over 50 years of age. The coefficients in both cases are es- sentially of the same order of magnitude. Perhaps someone will be inclined to believe that the correlation between father and son, and brother! and brother, in respect of the duration of life arises as a result of similarity of the environments to which they are exposed. Pearson's comments on this point are penetrating, and I believe absolutely sound. He says : There may be some readers who will be inclined to consider that much of the correlation of duration of life between brothers is due to there being a likeness of their environment, and that thus each pair of brethren is linked together and differentiated from the general population. But it is difficult to believe that this really affects adult brothers or a father and his adult offspring A man who dies between 40 and 80 can hardly be said to have an environment more like that of his brother or father, who died also at some such age. than like any other member of the general popula- tion. Of course, two brothers have usually a like environment in infancy, and their ages at death, even if they die adults, may be influenced by their rearing. But if this be true, we ought to find a high correlation in ages THE INHERITANCE OF DURATION 171 at death of brethren who die as minors. As a matter of fact this correla- tion for minor and minor is 40 to 50 per cent, less than in the case of adult and adult. It would thus seem that identity of environment is not the principal factor in the correlation between ages of death, for this correlation is far less in youth than in old age. TABLE 19 Inheritance of duration of life. Data from Quaker records. (Beeton and Pearson) Relatives Correlation coefficient r*v Ratio of coefficient to its probable error r*y I Er X y Father Adult son 0.135± .021 6.4 Father Minor son .087± .022 4.0 Father Father Mother Adult daughter Minor daughter Adult son .130± .020 .052± .023 .131± .019 6.5 2.3 6.9 Mother Minor son .076=t .024 3.2 Mother Mother Adult daughter Minor daughter .149± .020 .138=t .024 7.5 5.7 Elder adult brother Adult brother Younger adult brother Adult brother .229± .019 .285± .020 12.1 14.3 Minor brother Minor brother .103± .029 3.6 Adult brother Minor brother -.026± .025 1.0 Elder adult sister Adult sister Younger adult sister Adult sister .346± .018 .332± .019 19.2 17.5 Minor sister Minor sister .175± .031 5.6 Adult sister Minor sister -.026± .029 .9 Adult brother Adult sister .232± .015 15.5 Minor brother Minor sister .144 ± .025 5.8 Adult brother Minor sister -.006± .035 .2 Adult sister Minor brother -.027± .024 1.1 The cases above the horizontal line are all direct lineal inheritance; those below the line collateral inheritance. The results regarding minors to which Pearson refers are shown in Table 19. This table gives the results of the second study made by Beeton and Pearson on inher- itance of duration of life, based upon the records of the 172 BIOLOGY OF DEATH Friends' Societies. It appears in the upper half of the table that wherever a parent, father or mother, appears with a minor son or daughter the correlation coefficients are small in magnitude. In some cases they are just barely significant in comparison with their probable errors as for example, the correlation of father and minor son, and that of mother and minor daughter. In the other cases involving minors the coefficients are so small as to be insignificant. On the other hand, in every case of correlation between parent and adult offspring of either sex, the coefficient is 6 or more times its probable error, and must certainly be regarded as significant. It- will further be noted that the magnitude of the coefficients obtained from these Quaker records, is of the same general order as was seen in the previous table based on the "Peerage" and "Landed Gentry" material. The lower part of the table gives the results for various fraternal relationships. In general the frater- nal correlations are higher that the parental. The coeffi- cients for minors or for minors with adults are very low and in most cases not significantly different from zero. In four cases — namely, adult brother with minor brother ; adult sister with minor sister ; adult brother with minor sister; and adult sister with minor brother — the coeffi- cients are all negative in sign, although in no one of the cases is the coefficient significant in comparison with its probable error. A minus sign before a correlation coefficient means that an increase in the value of one of the variables is associated with a decrease in the value of the other. So that these negative coefficients would mean, if they were significant, that the greater the age at death of an adult brother, the lower the age at death of his minor brother or sister. But the coefficients are actually sensibly equal to zero. Pearson THE INHERITANCE OF DURATION 173 points out that the minus sign in the case of these correla- tions of adult with minor exhibits the effect of the inheri- tance of the mortality of youth. Minors dying from 16 to 20 are associated with adults dying from 21 to 25. That is, minors dying late correspond to adults dying early. This situation may be a peculiarity of the Quaker material with which this work deals. There is urgent need for further study of the inheritance of the duration of life on more and better material than any which has hitherto been used for the purpose. I have under way in my own laboratory at the present time an extensive investigation of this kind, in which there will be hundreds of thousands of pairs of relatives in the individual correlation tables instead of thousands, and all types of collateral kinship will be represented. Because of the magnitude of the investigation, however, it will be still a number of years before the results will be in hand for discussion. The facts which have been presented leave no doubt as to the reality of the inheritance factor as a prime determinant of the length of the life span. At the beginning it was pointed out that it was on a priori grounds highly probable that duration of life is influenced by both heredity and environment, and that the real problem is to measure the comparative effect of these two general sets of factors. We have seen that the intensity of inheritance of duration of life, taking aver- ages, is of the order indicated by the following coefficients. Parental correlation (adult children) r = . 1365 Fraternal correlation (adults) r=.2831 Now we have to ask this question : What are the values of parental and fraternal correlation for characters but slightly if at all affected in their values by the environ- ment? Happily, Pearson has provided such values in his 174 BIOLOGY OF DEATH extensive investigations on the inheritance of physical characters in man. In Table 20 are given the values of the parental correlations for the four physical characters — stature, span, forearm length, and eye color. Now it is obvious TABLE 20 Parental inheritance of physical characters in man. (Pearson) Pair Organ Correlation Father and Son . . . . Stature .51 Father and Son . . . . Span 45 Father and Son . . . . Forearm 42 Father and Son . . . . Eye color 55 Father and Daughter . . . . Stature 51 Father and Daughter . . . . Span 45 Father and Daughter . . . . Forearm 42 Father and Daughter . . . . Eye color 44 Mother and Son . . . . Stature 49 Mother and Son . . . . Span 46 Mother and Son . . . . Forearm 41 Mother and Son . . . . Eye color 48 Mother and Daughter . . . . Stature 51 Mother and Daughter . . . . Span 45 Mother and Daughter . . . . Forearm 42 Mother and Daughter . . . . Eye color 51 that the differences of environmental forces impinging upon the various members of a homogeneous group of middle class English families (from which source the data for these correlations were drawn) can by no pos- sibility be great enough to affect sensibly the stature, the arm-length, or the eye color of the adults of such families. It would be preposterous to assert that the resemblance between parents and offspring in respect of eye color is due solely, or even sensibly, to similarity of environment. It is due to heredity and substantially nothing else. Now the average value of the 16 parental coefficients for the inheritance of physical characters shown in the table is r= .4675 THE INHERITANCE OF DURATION 175 Table 21 shows the coefficients for the fraternal in- heritance of six physical characters, cephalic index (the ratio of head length and head breadth) and hair color having been added to those given in the parental table. Again it is seen that the coefficients have all about the TABLE 21 Fraternal inheritance of physical characters in man. (Pearson) Pair Organ Correlation Brother and Brother Stature .51 Brother and Brother Span 55 Brother and Brother Forearm .49 Brother and Brother Eye color 52 Brother and Brother Cephalic index 49 Brother and Brother Hair color .59 Sister and Sister Stature 54 Sister and Sister Span .56 Sister and Sister Forearm 51 Sister and Sister Eye color .45 Sister and Sister Cephalic index .54 Sister and Sister Hair color 56 Brother and Sister Stature 55 Brother and Sister Span 53 Brother and Sister Forearm .44 Brother and Sister Eye color 46 Brother and Sister Cephalic index . . . . , 43 Brother and Sister Hair color .56 same values, and it is as apparent as before that the resemblance between brother and sister, for example, in eye-color, or arm length, or shape of head cannot for a moment, because of the nature of the characters them- selves, be supposed to have arisen because of the simi- larity of environment, The average value of all these fraternal coefficients is r= .5156 From these data, with the help of a method due to Pearson, it is possible to determine the percentage of the 176 BIOLOGY OF DEATH death rate dependent upon the inherited constitution, and the percentage not so dependent. If pN be the number of deaths in N cases which depend in no way upon the inherited constitution of the individual, then (1-p) will represent the chance of an individual dying because of his inherited constitutional makeup, and (1-p)2 will be the chance of a pair of individuals, say two brothers, both dying from causes determined by inheritance. If further r denotes the observed correlation between individuals in respect of duration of life, and rQ the correlation between the same kin in respect of such measured physical charac- ters as those just discussed, in the determination of which it is agreed that environment can play only a small part, we have the following relation: Substituting the ascertained values we have 1. From parental correlations. 0.1365 = .4675 (1-p)2 (1-p) 2 = .292 (1-p) - .54 2. From fraternal correlations 0.2831 = .5156 (1-p) 2 (1-p) = .74 From these figures it may be concluded, and Pearson does so conclude, that from 50 to 75 per cent, of the general death rate within the group of the population on which the calculations are based, is determined funda- mentally by factors of heredity and is not capable of essential modification or amelioration by any sort of environmental action, however well intentioned, however costly, or however well advertised. Mutatis mutandis the same conclusion applies to the duration of life. I have THE INHERITANCE OF DURATION 177 preferred to state the conclusion in terms of death rates, as it was originally stated by Pearson, because of the bearing it has upon a great deal of the public health propaganda so loosely flung about. It need only be re- membered that there is a perfectly definite functional relation between death rate and average duration of life in an approximately stable population group, expres- sible by an equation, in order to see that any conclusion as to the relative influence of heredity and environment upon the general death rate must apply with equal force to the duration of life. • THE SELECTIVE DEATH BATE IN MAN If the duration of life were inherited it would logical- ly be expected that some portion of the death rate must be selective in character. For inheritance of duration of life can only mean that when a person dies is in part determined by that individual's biological constitution or makeup. And equally it is obvious that individuals of weak and unsound constitution must, on the average, die earlier than those of strong, sound, and vigorous con- stitution. Whence it follows that the chances of leaving offspring will be greater for those of sound constitution than for the weaklings. The mathematical discussion which has just been given indicates that from one-half to three-fourths of the death rate is selective in char- acter, because that proportion is determined by hereditary factors. Just in proportion as heredity determines the death rate, so is the mortality selective. The reality of the fact of a selective death rate in man can be easily shown graphically. In Figure 44 are seen the graphs of some data from European royal families, where no neglect of children, 12 178 BIOLOGY OF DEATH degrading environmental conditions, or economic want can have influenced the results. These data were com- 30 £ >• 20 15 10 MOTHER AND CHILDREN FATHER AND CHILDREN 16 36 56 66 \ \ \ \ 66 and over AGE. AT DEATH OF PARENTS FIG. 44. — Diagram showing the influence of age at death of parents upon the percentage of offspring dying under 5 years. (After Ploetz). piled by the well-known German eugenist, Professor Ploetz of Munich. The lines show the falling per- THE INHERITANCE OF DURATION 179 centage of the infantile death rate as the duration of life of the father and mother increases. Among the chil- dren of short-lived fathers and mothers, at the left end of each line, is found the highest infant mortality, while among the offspring of long-lived parents the lowest infant mortality occurs, as shown at the right-hand end of the diagram. The results so far presented regarding a selective death rate and inheritance of duration of life, have come from selected classes : the aristocracy, royalty or Quakers. None of these classes can be fairly said to represent the general population. Can the conclusion be transferred safely from the classes to the masses? To the determina- tion of this point one of Pearson's students, Dr. E. C. Snow, addressed himself. The method which he used was, from the necessities of the case, a much more complicated and indirect one than that of Pearson and Ploetz. Its essen- tial idea was to see whether infant deaths weeded out the unfit and left as survivors the stronger and more resis- tant. All the infants born in a single year were taken as a cohort and the deaths occurring in this cohort in suc- cessive years were followed through. Resort was had to the method of partial or net correlation. The variables correlated in the case of the Prussian data were these : 1. a?0 = Births in year a given cohort started. 2. Xi = Deaths in the first two years of life. 3. a?2 = Deaths in the next eight years of life. 4. #3 = Deaths in the ten years of all individuals not included in the particular cohort whose deaths are being followed. In the case of the English data the variables were : x0 = Births in specified year. Xi =• Deaths in the first three years of life of those born in specified year. xz = Deaths in fourth and fifth years of life of those born in specified year. d?3 = The "remaining" deaths under 5. 180 BIOLOGY OF DEATH The underlying idea was to get the partial or net correlation between x1 and x2, while x0 and x3 are held constant. If the mortality of infancy is selective, its amount should be negatively correlated to a significant degree with the mortality of the next eight years when the births in each district considered are made con- stant and when the general health environment is made constant. Under the constant conditions specified a negative correlation denotes that the heavier the infan- TABLE 22 Snow's results on selective death rate in man. English and Prussian rural districts Data Actual correlation r!2.03 Expected correlation if no selection Males: English Rural Districts (1870) (1871) (1872) -0.4483 - .3574 - .2271 -0.0828 - .1014 - .0807 Prussian Rural Districts (1881) (1882) - .9278 - .6050 - .0958 - .0765 Females : English Rural Districts (1870) (1871) (1872) - .4666 - .2857 - .5089 - .0708 - .0505 - .0496 Prussian Rural Districts (1881) (1882) - .8483 - .6078 - .0933 - .0705 tile death rate in a cohort of births the lighter will be the death rate of later years, and vice versa. The last variable, x3, is the one chosen, after careful consideration and many trials, to measure variation in the health envi- ronment. If any year is a particularly unhealthy one — an epidemic year for example — then this unhealthiness should be accurately reflected in the deaths of those mem- bers of the population not included in the cohort under review. THE INHERITANCE OF DURATION 181 Snow's results for English and Prussian rural dis- tricts are set forth in Table 22. From this table it is seen that in every case the correlations are negative, and therefore indicate that the mortality of early life is selec- tive. Furthermore, the demonstration of this fact is completed by showing that the observed coefficients are from 3 to 10 times as great as they would be if there were no selective character to the death rate. The coefficients for the Prussian population, it will be noted, are of a distinctly higher order of magnitude than those for the English population. This divergence is probably due chiefly to differences in the quality of the fundamental statistical material in the two cases. The Prussian ma- terial is free from certain defects inherent in the English data, which cannot be entirely got rid of. The difference in the coefficients for the two successive Prussian cohorts represents, in Snow's opinion, probably a real fluctua- tion in the intensity of natural selection in the one group as compared with the other. How significant Snow's results are is shown graphically in Figure 45. Snow's own comments on his results are significant. He says: * The investigations of this memoir have been long and laborious, and the difficulties presented by the data have been great. Still, the general result cannot be questioned. Natural selection, in the form of a selective death-rate, is strongly operative in man in the early years of life. Those data which we believe to be the best among those we have used — the Prus- sian figures — show very high negative correlation between the deaths in the first two years of life and those .in the next eight, when allowance is made for difference in environment. We assert with great confidence that a high mortality in infancy (the first two years of life) is followed by a corresponding low mortality in childhood, and conversely. The English figures do not allow such a comprehensive survey to be undertaken, but, so far as they go, they point in the same direction as the Prussian ones. The migratory tendencies in urban districts militate against the detection of selective influences there, but we express the belief that those influences 182 BIOLOGY OF DEATH are just as prevalent in industrial as in rural communities, and could be measured by other means if the data were forthcoming. 1.0 i- bl IS 7 .6 MALES .3 .1 .o EXPECTED CORRELATION - NO SELECT/ON 1 1670 1871 l^7^ 1661 1682 ENGLISH PRUSSIAN FIG. 45. — Snow's results on selective death rate in man. The cross-hatched area may be taken, in comparison with the small clear area at the bottom, as indicating the influence of the selective death rate in increasing the correlations. Our investigation substantiates for a general population the results found by Pearson and Ploetz for more restricted populations, and disagrees with many statements of health officers. It is with great reluctance that THE INHERITANCE OF DURATION 183 we point out this disagreement, and assert a doctrine which, in the present sentiment of society, is bound to be unpopular. We have no feelings of antagonism towards the efforts which have been made in recent years to save infant life, but we think that the probable consequences of such actions, so far as past experience can indicate them, should be completely under- stood. All attempts at the reduction of mortality of infancy and childhood should be made in the full knowledge of the facts of heredity. Everybody knows the extreme differences in constitutional fitness which exist in men and women. Few intelligent people can be ignorant of the fact that this constitutional fitness is inherited according to laws which are fairly definitely known. At the same time marriage is just as prevalent among those of weak stocks as among those of the vigorous, while the fertility of the former is certainly not less than that of the latter. Thus a propor- tion of the infants born every year must inevitably belong to the class referred to in the report as "weaklings," and, with Pearson's results before us, we are quite convinced that true infantile mortality (as distinct from the mortality due to accident, neglect, etc. — no small proportion of the whole) finds most victims from among this class. Incidentally we would here suggest that no investigation into the causes of infant and child mortality is complete until particulars are gathered by the medical officers of the constitutional tendencies and physical characters of the parents. Our work has led us to the conclusion that infant mortality does effect a "weeding out" of the unfit; but, though we would give this conclusion all due emphasis, we do not wish to assert that any effort, however small, to the end of reducing this mortality is undesirable. Nobody would suggest that the difference between the infant rates in Oxfordshire and Glamorgan- shire (73 and 154 per 1,000 births respectively, in 1908) was wholly due to the constitutional superiority of the inhabitants of the former county. The "weeding-out" process is not uniform. In the mining districts of South Wales, accident, negligence, ignorance and insanitary surroundings account for much. By causing improvements under these heads it may be possible to reduce the infant mortality of Glamorganshire by the sur- vival of many who are not more unfit than are those who survive in Oxfordshire, and the social instincts of the community insist that this should be done. This work of Snow's aroused great interest, and soon after it appearance was controverted, as it seems to me quite unsuccessfully, by Brownlee, Saleeby and others. Happily the results of Pearson, Ploetz and Snow on the selective death rate have recently been accorded a confirmation and extension to still another group of 184 BIOLOGY OF DEATH people — the Dutch — in some investigations carried out by Dr. F. S. Cram of the Prudential Life Insurance Com- pany, with the assistance of the distinguished mathe- matical statistician, Mr. Arne Fisher. The Dutch Government publishes annually data which undoubtedly furnish the best available material now exist- ing in the world for the purpose of determining whether or not there is a positive or negative correlation between infant mortality and the mortality in the immediately subsequent years of life. Fisher's mathematical analy- sis embraces a very large body of material, including nearly a million and a half births, and nearly a quarter of a million deaths of males occurring in the first five years of life. The Holland data make it possible to develop life tables for every cohort of births and this has been done in the 16 cohorts of males during the years 1901-1916. The data also make it possible to work up these life tables for urban areas and for rural areas. After carefully eliminating secular disturbances the Holland material appears to prove quite conclusively for the rural districts that there is a definite negative corre- lation, of significant magnitude, between infant mortality and the mortality in the immediately subsequent years of life. The only place where positive correlation appears is in the four large cities of the country with more than a hundred thousand inhabitants each. Fisher makes the following point (in a letter to the present writer) in ex- planation of these positive correlations. He says : The larger cities are better equipped with hospital and clinical facilities than the smaller cities and the rural districts. More money is also spent on child welfare. Is it therefore not possible that many feeble lives who in the course of natural circumstances would have died in the first year of life are carried over into the second year of life by means of medical skill? But medicine cannot always surpass nature, and it THE INHERITANCE OF DURATION 185 might indeed be possible that among cohorts with a low mortality during the first two years of life there will be an increase of death rate in the following three years of life. Altogether, we may regard the weight of present evi- dence as altogether preponderant in favor of the view that the death rate of the earliest period of life is selec- tive— eliminating the weak and leaving the strong. From our present point of view it adds another broad class of evidential material to the proof of the proposition that inheritance is one of the strongest elements, if not indeed the dominating factor, in determining the duration of life of human beings. CHAPTER VII EXPERIMENTAL STUDIES ON THE DURATION OF LIFE INHEKITANCE OF DURATION OF LIFE IN DROSOPHILA IN the last chapter there was presented indubitable proof that inheritance is a major factor in determining the duration of life in man. The evidence, while entirely convincing and indeed in the writer's opinion critically conclusive, must be, in the nature of the case, statistical in its nature. Experimental inquiries into the duration of human life are obviously impossible. It is always important, however, as a general principle, and particu- larly so in the present instance, to check one's statistical conclusions by independent experimental evidence. This can be successfully done, when one's problem is longevity, only by choosing an animal whose life-span relative to that of man is a short one, and in general the briefer it is, the better suited will the animal be for the purpose. An organism which rather completely fulfils the re- quirements of the case, not only in respect of the short- ness of the life span, but also in other ways, such as ease of handling, feeding, housing, etc., is the common "fruit" or "vinegar" fly, Drosophila melanogaster. This insect, which every one has seen hovering about bananas and other fruit in fruit shops, has lately attained great fame and respectability as a laboratory animal, as a result of the brilliant and extended investigations of Morgan and his students upon it, in an analysis of the mechanism of heredity. DrosopMa is a small fly, per- 186 STUDIES ON THE DURATION OF LIFE 187 haps one fourth as large as the common house fly. It has striking red eyes, a brownish body, and wings of length and form varying in different strains. It lives normally on the surface of decaying fruit of all sorts, but because of a more or less well-marked preference for FIG. 46. — Male and female fruit fly. (Drosophila melanog aster). (From Morgan). banana it is sometimes called the "banana" fly. While it lives on decaying fruit surfaces its food is mainly not the fruit itself, but the yeast which is always growing in such places. The life cycle of the fly is as follows : The egg laid, by the female on some fairly dry spot on the food develops in about 1 day into a larva. This larva or maggot crawls about and feeds in the rich medium in which it finds itself for about 3 to 4 days and then forms a pupa. From the pupa the winged imago or adult form emerges in about 4 or 5 days. The female generally begins to lay eggs within the first 24 hours after she is hatched. So 188 BIOLOGY OF DEATH then we have about 8 to 10 days as the minimum time duration of a generation. The whole cycle from egg to egg, at ordinary room temperature, falls within this 10- day period with striking accuracy and precision. The duration of life of the adult varies in an orderly manner from less than 1 day to over 90 days. The span CO I 1,000 900 '600 700 *» too A6L IN DAYS FIG. 47. — Life lines for Drosophila melanog aster; showing the survivors at different ages out of 1000 born at the same time. of life of Drosophila quantitatively parallels in an extra- ordinary way that of man, with only the difference that life's duration is measured with different yardsticks in the two cases. Man's yardstick is one year long, while Drosophila' s is one day long. A fly 90 days old is just as decrepit and senile, for a fly, as a man 90 years old is in human society. This parallelism in the duration of life of Drosophila and man is well shown in Figure 47, which represents a life table for adult flies of both sexes. The survivor- ship, or lx figures, are the ones plotted. The curves deal STUDIES ON THE DURATION OF LIFE 189 only with flies in the adult or imago stage, after the com- pletion of the larval and pupal periods. The curve is based upon 3,216 female and 2,620 male flies, large enough numbers to give reliable and smooth results. We note at once that in general the curve has the same form as the corresponding lx curve from human mortality tables. The most striking difference is in the absence from the fly curves of the heavy infant mortality which characterizes the human curve. There is no specially sharp drop in the curve at the beginning of the life cycle, such as has been seen in the lx curve for man in an earlier chapter in this book. This might at first be thought to be accounted for by the fact that the curve begins after the infantile life of the fly, but it must be remembered that the human I x line begins at birth, and no account is taken of the mortal- ity in utero. Really the larval and pupal stages of the fly correspond rather to the foetal life of a human being than to the infant life, so that one may perhaps fairly take the curves as covering comparable portions of the life span in the two cases and reach the conclusion that there is not in the fly an especially heavy incidence of mortality in the infant period of life, as there is in man. The explana- tion of this fact is, without doubt, that the fly when it emerges from the pupal stage is completely able to take care of itself. The baby is, on the contrary, in an almost totally helpless condition at the same relative age. It is further evident that at practically all ages in Drosophila the number of survivors at any given age is higher among the female than among the males. This, it will be recalled, is exactly the state of the case in human mortality. The speed of the descent of the Drosophila curve slows off in old age, just as happens in the human life curve. The rate of descent of the curve in early middle life is somewhat more rapid with the flies than 190 BIOLOGY OF DEATH in the case of human beings, but as will presently appear there are some strains of flies which give curves almost identical in this respect with the human mortality curves. In the life curves of Figure 47, all different degrees of inherited or constitutional variation in longevity are in- cluded together. More accurate pictures of the true state of affairs will appear when we come, as we presently shall, to deal with groups of individuals more homoge- neous in respect of their hereditary constituents. Having now demonstrated that the incidence of mor- tality is in general similar in the fly Drosopkila to what it is in man, with a suitable change of unit of measure, we mayproceed to examine some of the evidence regarding the inheritance of duration of life in this organism. The first step in such an examination is to determine what degree of natural variation of an hereditary sort exists in a general fly population in respect of this characteristic. In order to do this it is necessary to isolate individual pairs, male and female, breed them together and see whether, between the groups of offspring so obtained, there are genetic differences in respect of duration of life which persist through an indefinite num- ber of generations. This approaches closely to the pro- cess called by geneticists the testing of pure lines. In such a process the purpose is to reduce to a minimum the genetic diversity which can possibly be exhibited in the material. In a case like the present, the whole amount of genetic variation in respect of duration of life which can appear in the offspring of a single pair of parents is only that which can arise by virtue of its prior existence in the parents themselves individually, and from the combina- tion of the germinal variation existing in the two parents one with another. We may call the offspring, through successive generations, of a single pair of parents a line STUDIES ON THE DURATION OF LIFE 191 of descent. If, when kept under identical environmental conditions, such lines exhibit widely different average durations of life, and if these differences reappear with constancy in successive generations, it may be justly concluded that the basis of these differences is heredi- tary in nature, since by hypothesis the environment of all the lines is kept the same. In consequence of the environmental equality, whatever differences do appear must be inherently genetic. The manner in which these experiments are performed may be of interest. An experiment starts by placing two flies, brother and sister, selected from a stock bottle, together in a half -pint milk bottle. At the bottom of the bottle is a solidified, jelly-like mixture of agar-agar and boiled and pulped banana. On this is sown, as food, some dry yeast. A bit of folded filter paper in the bottle fur- nishes the larvae opportunity to pupate on a dry sur- face. About ten days after the pair of flies have been placed in this bottle, fully developed offspring in the imago stage begin to emerge. The day before these off- spring flies are due to appear, the original parent pair of flies are removed to another bottle precisely like the first, and the female is allowed to lay another batch of eggs over a period of about nine days. In the original bottle there will be offspring flies emerging each day, having developed from the eggs laid by the mother on each of the successive days during which she was in the bottle. Each morning the offspring flies which have emerged during the preceding twenty-four' hours are transferred to a small bottle. This has, just as the larger one, food material at the bottom and like the larger one is closed with a cotton stopper. All of the offspring flies in one of these small bottles are obviously of the same age, because they were born at the same time, 192 BIOLOGY OF DEATH using this term "born.' to denote emergence from the pupal stage as imagines. Each following day these small bottles are inspected. Whenever a dead fly is found, it is removed and a record made in proper form of the fact that its death occurred, and its age and sex are noted. Finally, when all the flies in a given small bottle have died, that bottle is discarded, as the record of the duration 1.000 900 600 700

•a~s aS I-H "eS •<-> 0 H Diff. 800-899 10 10 900-999 10 6 1000-1099 6 9 1100-1199 1 64 39 One concludes from these figures that tethelin can be regarded as having lengthened the span of life to a de- gree which is just significant statistically. One would expect, from the variation of random sampling alone, to get as divergent results as these about l1^ times in every 100 trials with samples of 64 and 39, respectively. In any event it is apparent that, making out the best case possible, the differences in average duration of life 222 BIOLOGY OF DEATH produced by administration of tethelin are of a wholly different and smaller order than those which have been shown, in the earlier portion of this chapter, to exist be- tween pure strains of Drosophila which are based upon hereditary differences. Putting together all the results which have been re- viewed in this and the preceding chapter, it appears to be clearly and firmly established that inheritance is the factor of prime importance in determining the normal, natural duration of life. In comparison with this factor, the influence of environmental forces (of sub-lethal im- mediate intensity of course) appears in general to be less marked. CHAPTER VIII NATURAL DEATH, PUBLIC HEALTH, AND THE POPULATION PEOBLEM. SUMMAKY OF KESULTS I have attempted to review some of the important biological and statistical contributions which have been made to the knowledge of natural death and the duration of life, and to synthesize these scattered results into a coherent unified whole. In the present chapter I shall endeavor to summarize, in the briefest way, the scattered facts which have been passed in review, and to follow a presentation of the general results to which they lead with some discussion of what we may reasonably regard the future as having in store for us, so far as may be judged from our present knowledge of the trend of events. What are the general results of our review of the gen- eral biology of death? In the first place, one perceives that natural death is a relatively new thing, which appeared first in evolution when differentiation of cells for partic- ular functions came into existence. Unicellular ani- mals are, and always have been, immortal. The cells of higher organisms, set apart for reproduction in the course of differentiation during evolution, are Immortal. The only requisite conditions to make their potential im- mortality actual are physico-chemical in nature and are now fairly well understood, particularly as a result of the investigations of Loeb upon artificial parthenogenesis and related phenomena, The essential and important 223 224 BIOLOGY OF DEATH somatic cells of the body, however much differentiated, are also potentially immortal; but the conditions neces- sary for the actual realization of the potential immor- tality are, in the nature of the case, as has been shown by the brilliant researches of Leo Loeb, Harrison and Carrel on tissue culture, such as cannot be realized so long as these cells are actually in and a part of the higher metazoan body. The reason why this is so, and why in consequence death results in the metazoa, is that, in such organisms the specialization of structure and function necessarily makes the several parts of the body mutually dependent for their life upon each other. If one organ or group, for any accidental reason begins to function abnormally and finally breaks down, the balance of the whole is upset and death eventually follows. But the individual cells, themselves, could go on living indefinitely, if they were freed, as they are in cultures, of the neces sity of depending upon the proper functioning of other cells for their food, oxygen, etc. So then we see emerging, as our first general result, the fact that natural death is not a necessary or inevit- able consequence of life. It is not an attribute of the cell. It is a by-product of progressive evolution — the price we pay for differentiation and specialization of structure and function. This first result indicates logically, in any particu- lar organism such as man, the great importance of a quantitative analysis of the manner in which dif- ferent parts of the body break down and lead to death. Such an analysis, carefully worked through, demonstrates that this breaking down is not a haphazard process, but a highly orderly one resting upon a fundamental biolog- ical basis. The progress of the basic tissue elements NATURAL DEATH, PUBLIC HEALTH 225 of the body along the evolutionary pathway appears to be an important factor in determining the time when the organ systems in which they are chiefly involved shall break down. Those organ systems that have evolved farthest away from original primitive conditions are the soundest and most resistant, and wear the longest under the strain of functioning. So then, the second large result is that it is the way potentially immortal cells are put together in mutually dependent organ sys- tems that immediately determines the time relations of the life span. But it was possible to penetrate more deeply into the problem than this by finding that the duration of life is an inherited character of an individual, passed on from parent to off spring, just as is eye color or hair color, and with a relatively high degree of precision. This has been proved in a variety of ways, first directly for man (Pearson) and for a lower animal, Drosophila, (Hyde, Pearl) by measuring the degree of hereditary transmis- sion of duration of life, and indirectly by showing that the death rate was selective (Pearson, Snow, Bell, Ploetz) and had been, since nearly the beginning of recorded his- tory, at least. It is heredity which determines the way the organism is put together — the organization of the parts. And it is when parts break down and the organ- ization is upset that death comes. So the third large re- sult is that heredity is the primary and fundamental determiner of the length of the span of life. Finally, it is possible to say probably, though not as yet definitely because the necessary mass of experimen- tal evidence is still lacking, but will, I believe, be shortly provided, that environmental circumstances play their 15 226 BIOLOGY OF DEATH part in determining the duration of life largely, if not in principle entirely, by influencing the rate at which the vital patrimony is spent. If we live rapidly, like Loeb and Northrop 's Drosophila at the high temperatures, our lives may be more interesting, but they will not be so long. The fact appears to be, though reservation of final judgment is necessary till more returns are in, that heredity determines the amount of capital placed in the vital bank upon which we draw to continue life, and which when all used up spells death ; while environment, using the term in the broadest sense to include habits of life as well as physical surroundings, determines the rate at which drafts are presented and cashed. The case seems in principle like what obtains in respect of the duration of life of a man-constructed machine. It is self-evident that if, of two automobiles of the same make leaving the factory together new at the same time, one is run at the rate of 1,000 miles per year and the other at the rate of 10,000 miles per year, the useful life of the former is bound to be much longer in time that that of the latter, accidents being excluded in both cases. Again, a very high priced car, well-built of the finest material, may have a shorter duration of life than the poorest and cheapest machine, provided the annual mileage output of the former is many times that of the latter. The first three of these conclusions seem to be firmly grounded. The last rests, at present, upon a less secure footing. Because it does, it offers an extremely promis- ing field for both statistical and experimental research. We need a wide variety of investigations, like those of Loeb and Northrop, of Slonaker and of Rubner, on the experimental side. On the statistical side, well-conceived NATURAL DEATH, PUBLIC HEALTH 227 and careful studies, by the most refined of modern meth- ods, upon occupational mortality seem likely to yield large returns. PUBLIC HEALTH ACTIVITIES Fortunately, it is possible to get some light on the environmental side from existing statistical data by con- sidering, in a broad general way, the results of public health activities. Any public health work, of course, deals, and can deal in the present state of public senti- ment and enlightenment, only with environmental matters. Attempts at social control of the germ-plasm — the innate inherited constitutional make-up — of a people, by eugenic legislation, have not been conspicuously successful. And there is a good deal of doubt, having regard to all factors necessarily involved, whether they have always been even well-conceived. As an animal breeder of some years' experience, I have no doubt whatever that almost any breeder of average intelligence, if given omnipotent control over the activities of human beings, could, in a few generations, breed a race of men on the average con- siderably superior — ~by our present standards — to any race of men now existing in respect of many qualities or attributes. But, as a practical person, I am equally sure that nothing of the sort is going to be done by legislative action or any similar delegation of powers. Before any sensible person or society is going to entrust the control of its germ-plasm to politics or to science, there will be demanded that science know a great deal more than it now does about the vagaries of germ-plasms and how to control them. Another essential difficulty is one of stan- dards. Suppose it to be granted that our knowledge of 228 BIOLOGY OF DEATH genetics was sufficiently ample and profound to make it possible to make a racial germ-plasm exactly whatever one pleased; what individual or group of individuals could possibly be trusted to decide what it should be? Doubtless many persons of uplifting tendencies would promptly come forward prepared to undertake such a responsibility. But what of history? If it teaches us anything, it is that social, moral and political standards are not fixed and absolute, but vary, and vary radically in both space and time. And further, history teaches that a great many of the most valuable people, in the highest and best sense, whom the world has ever known, were so constituted physically, morally; or otherwise, as to make it certain that under a strict eugenic regime they never would have existed at all. One cannot but feel that man's instinctive wariness about experimental interferences with his germ-plasm is in considerable degree, well-founded. But because of the altogether more impersonal na- ture of the case, most men individually and society in general are perfectly willing to let anybody do anything they like in the direction of modifying the environment in what is believed, or hoped to be, the direction of improve- ment, or trying to, quite regardless of whether science is able to give any slightest inkling on the basis of ascer- tained facts as to whether the outcome will be good, bad or indifferent. Hence many kinds of weird activities and propaganda flourish like the proverbial bay tree. Of all organized activities looking towards the direct modification of the environment to the benefit of mankind, that group comprised under the terms sanitation, hygiene NATURAL DEATH, PUBLIC HEALTH 229 and public health have, by all odds, the best case when measured in terms of accomplishment. Man's expecta- tion of life has increased as he has come down through the centuries (cf. Pearson and Macdonell.) A large part of this improvement must surely be credited to his improved understanding of how to cope with an always more or less inimical environment and assuage its asper- ities to his greater comfort and well-being. To, fail to give this credit would be manifestly absurd. But it would be equally absurd to attempt to main- tain that all decline in the death-rate which has occurred has been due to the efforts of health officials, whether conscious or unconscious, as is often asserted and still more often implied in the impassioned outpourings of zealous propagandists. The open-minded student of the natural history of disease knows perfectly well that a large part of the improvement in the rate of mortality cannot possibly have been due to any such efforts. To illustrate the point, I have prepared a series of illustra- tions dealing with conditions in the Registration Area of the United States in the immediate past. All these diagrams (Figures 52, 53, and 54) give death-rates per 100,000 from various causes of death in the period of 1900-1918, inclusive, both sexes for simplicity being taken together. The lines are all plotted on a logarithmic scale. The result of this method of plotting is that the slope trend of each line is directly comparable with that of any other, no matter what the absolute magnitude of the rates concerned. It is these slopes, measuring im- provement in mortality, to which I would especially direct attention. 230 BIOLOGY OF DEATH 1.000 100 1 g 10 O.I CONTROLLABLE. CAUSES OF DEATH <* THC LUNGS V — i i i i i I i I i i i i i 1900 01 02 03 O4 O5 O6 O7 Od 09 IO II iZ 13 /4 15 16 17 Id YLAR FIG. 52. — Trend of death rates for four causes of death against which public health activities have been particularly directed. In figure 52 are given the trends of the death-rates for four diseases against which public health and sani- tary activities have been particularly and vigorously NATURAL DEATH, PUBLIC HEALTH 231 directed, with, as we are accustomed to say, most grati- fying results. The diseases are: 1. Tuberculosis of the lungs. 2. Typhoid fever. 3. Diphtheria and croup. 4. Dysentery. We note at once that the death-rates from these diseases have all steadily declined in the 19 years under review. But the rate of drop has been slightly unequal. Remembering that the slopes are comparable, where- ever the lines may lie, and that an equal slope means a relatively equally effective diminution of the mortality of the disease, we note that the death-rate from tuber- culosis of the lungs has decreased slightly less than any of the other three. Yet it may fairly be said that so strenuous a warfare, or one engaging in its ranks so many earnest and active workers, has probably never in the history of the world been waged against any disease as that which has been fought in the United States against tuberculosis in the period covered. The rates of decline of the other three diseases are all practically identical. Figure 53 shows entirely similar trends for four other causes of death — namely: 1. Bronchitis (acute and chronic). 2. Paralysis without specified cause. 3. Purulent infection and septicaemia. 4. Softening of the brain. Now it will be granted at once, I think, that public health and sanitation can have had, at the utmost, ex- tremely little, if anything, to do with the trend of mor- tality from these four causes of death. For the most part they certainly represent pathological entities far beyond the present reach of the health officer. Yet the 232 BIOLOGY OF DEATH 1.000 r= NON - CONTROLLED CAUSES OF DEATH }00 o o o cr o 10 O.I I I I I I I I I I I I I I I I I I I I 190001 02 03 O4 05 O6 07 03 O9 10 II 12. 13 14- 15 16 17 13 YEAR FIG. 53. — Trend of death rates from four causes of death upon which no direct attempt at control has been made. outstanding fact is that their rates of mortality have de- clined and are declining just as did those in the control- lable group shown in Figure 52. It is of no moment NATURAL DEATH, PUBLIC HEALTH 233 I.OOO IOO § 8' 10 O.I I I 1 I I I LJ I I I I I 1900 01 02. O3 04 05 06 07 O8 O9 IO II 12. 13 14 15 16 17 Id YLAR FIG. 54. — Trend of combined death rate from the four causes shown in Figure 52 as compared with the four causes shown in Figure 53. to say that the four causes of death in the second group ,are absolutely of less importance than some of those in the first group, because what we are here discussing is not relative force of mortality from different causes, 234 BIOLOGY OF DEATH but rather the trend of mortality from particular causes. The rate of decline is just as significant, whatever the absolute point from which the curve starts. It is difficult to carry in the mind an exact impression of the slope of a line, so, in order that a comparison may be made, I have plotted in Figure 54, first, the total rate of mortality from the four controllable causes of death taken together and, second, the total rate of mortality from the four uncontrolled causes taken together. The result is interesting. The two lines were actually nearer together in 1900 than they were in 1918. They have diverged because the recorded mortality from the uncon- trolled four has actually decreased faster in the 19 years than has that from the four against which we have been actively fighting. The divergence is not great, however. Perhaps we are only justified in saying that the mortality in each of the two groups has notably declined, and at not far from identical rates. Now the four diseases in this group, I chose quite at random from among the causes of death whose rates I knew to be declining, to use as an illustration solely. I could easily pick out eight other causes of death which would illustrate the same point. I do not wish too much stress to be laid upon these examples. If they may serve merely to drive sharply home into the mind that it is only the tyro or the reckless propagandist, long ago a stranger to truth, who will venture to assert that a declining death- rate in and of itself marks the successful result of human effort, I shall be abundantly satisfied. It has been objected that the decline shown by the four "non-controlled" causes in the example just dealt with is due wholly, or nearly so, to changes in the practice of physicians relative to the reporting of the cause of NATURAL DEATH, PUBLIC HEALTH 235 death, and that, therefore, the decline is spurious. I have not been able to find that there is any good evidence that this is the fact ; that, in short, changes in reporting prac- tice have affected the "non-controlled" group more than the "controllable" group. But another kind of example may be cited to illustrate the same general point. Suppose we compare the course of mortality from certain well- defined causes, about the reporting of which there can be no controversy, in (a) a group of countries standing in an advanced position in matters of public health, sanitation, etc., and (b) a group of countries relatively backward and undeveloped in these respects. Such a comparison is im- possible to make over any long period of time because of lack of comparable data. I have succeeded in getting com- parable statistics on two diseases, namely typhoid fever and diphtheria, for the period 1898 to 1912 inclusive, for the following countries : A. Countries having (in period B. Countries having (in period covered) highly developed covered) less highly developed public health and sanitation. public health and sanitation Australia than those in group A. Austria Italy England and Wales Jamaica Germany Koumania Without going into detailed comparisons, which might be thought invidious, it is evident on the face of the case, I think, that the countries in the A group were, on the average during the period covered, much more advanced in all practical public health matters than were the coun- tries in group B. In Figures 55 and 56 are shown the trends of the weighted average death rates from typhoid fever and diphtheria respectively in the two groups of countries. It is evident from these diagrams that the death rates 236 BIOLOGY OF DEATH from these two causes declined, during the period cov- ered, in both the A and the B groups of countries and at not far from the same rate. There is no such large difference as would be expected if organized human inter- ference with the natural history of disease always played 100 10 i TYPHOID \ 1898 99 1900 01 0^ 03 04 05 O6 O7 08 09 10 II 12. YLAR FIG. 55. — Course of the weighted average death rate, for the countries in the A (solid line) and B (broken line) groups, from typhoid fever. the role of immediate and large importance which the propagandist asserts that it does. To guard against the possibility of any misunder- standing, let me say quite specifically and categorically, that the above is not intended in any way to convey the idea that public health work is not desirable, or that a NATURAL DEATH, PUBLIC HEALTH 237 laissez-faire policy would be better, or that public health efforts have not been enormously valuable in connection with typhoid fever and diphtheria. My purpose is quite other, being solely a desire to emphasize two things, viz: 1. That the trend of human mortality in time is an too 9O SO 70 60 -50 40 30 ^o to DIPHTHERIA 1898 99 19OO Ol 0^ 03 04 O5 06 O7 08 09 YEAR 10 II 12 Fia. 56. — Like figure 55, but for diphtheria and croup. extraordinarily complex biological phenomenon, in which many factors besides the best efforts of health officials are involved. 2. That for many causes of death a vast lot needs to be added to our knowledge of etiology, in the broadest sense, before really efficient control can be hoped for. This knowledge can come only through scientific investi- 238 BIOLOGY OF DEATH gation, and not through the complacent acceptance of the propagandist's assurance that "if what knowledge we now have is applied, all will be well."* Many others have, of course, perceived that, in the natural history of disease, mortality from particular causes may decline over long periods of time without any relation to what health departments have done, or tried to do about it. For example, Given has recently pointed out that there is no evidence that anything that man has done has affected, in either one way or the other, the decline in the mortality of tuberculosis, which has been continuous for nearly three-quarters of a century. Pearson has discussed the same point. There is much in our public health work that is worthy of the highest praise. When based upon a sound founda- tion of ascertained fact it may, and does, proceed with a step as firm and inexorable as that of Fate itself, to the wiping out of preventable mortality. Two recent ex- amples may be cited here, by way of specific illustration of what real and reasonably complete scientific knowledge can accomplish in public health work. Both examples are taken from the work of the International Health Board of the Rockefeller Foundation, with the permission of its director, Mr. Wickliffe Rose. The first concerns malaria. The life cycle of the malaria parasite is definitely known, and furnished a * One can but wonder if the many scientific men, who permit, and to some extent approve, such assertions, have ever thought of the menace to the continued support of research in science in general which inheres in this attitude of mind. The support of research comes finally back always to society in general — to the "average citizen" in short. Is it the part of wisdom to leave his education as to the meaning and significance of science for his happiness and well-being, so entirely in the hands of the propagandist as we now do? Has anti-vivisection taught no lesson? NATURAL DEATH, PUBLIC HEALTH 239 definite scientific basis for control procedure. "It is well understood, not only by scientists, but also by intel- ligent laymen, that the spread of the infection may be prevented by mosquito control, by protecting people from being bitten by mosquitoes, or by destroying the parasite in the blood of the human carrier. It has been shown, moreover, by repeated demonstrations, that by applica- tion of any one of these measures, or of any combination of them, the amount of malaria in a community may be reduced indefinitely. There are few diseases that pre- sent so many vulnerable points of attack and none per- haps the control of which may be made more definite or certain.' (Rose). In 1916 the International Health Board undertook some experiments in control at Crossett, Ark. In des- cribing the work Rose says : "Effort has been made to test the feasibility of malaria control in small communities by resort to such simple anti-mosquito measures as would fall within the limits of expenditure that such communities might well afford. The habits of the three mosquitoes — A. quadrimaculatus Say, A. punctipennis Say, and A. cruzians Wiedermann — which are responsible for the infection in these communities have been made the subject of constant study with a view to eliminating all unnecessary effort, and thereby reducing cost. "Experiment at Crossett, 1916 — The first of these tests was undertaken at Crossett, a lumber town of 2,129 inhabitants, situated in Ashley County in south-eastern Arkansas, about 12 miles north of the Louisiana line. Crossett lies at the edge of the so-called "uplands," in a level, low-lying region (elevation 165 feet), with sufficient undulation to provide reason- ably good natural drainage. Climatic conditions and abundant breeding places favor the propagation of anopheles. Malaria, in its severe form, is widely prevalent as an endemic infection, and according to the estimate of local physicians, is the cause of about 60 per cent, of all illness through- out the region. Within the town itself the malaria rate was high, and was recognized by the lumber corporation and the people as a serious menace to health and working efficiency. "The initial step in the experiment was a survey of the community to determine the malaria incidence, to ascertain in the species of mosquitoes 240 BIOLOGY? OF DEATH responsible for the spread of the infection, and to locate the breeding places of these mosquitoes. Breeding places were exhibited on a community map, and organized effort was centered on their destruction or control. The program of simple measures excluded all major drainage. Barrow pits and shallow ponds were filled or drained; streams were cleared of undergrowth when necessary to let the sunlight in; their margins and beds were cleared of vegetation and obstruction; and they were trained to a narrow channel, thus providing an unobstructed off-flow. Artificial ^con- tainers were removed from premises ; water barrels on bridges were treated with nitre cake. All remaining breeding places were regularly treated by removing vegetation, opening up shallow margins to give free access to small fish, and spraying once a week with road oil by means of automatic drips or a knapsack sprayer. All operations were under the supervision of a trained lay inspector. Care was exercised to eliminate all unnecessary effort and to secure, not the elimination of the last mosquito, but a rea- sonably high degree of control at a minimum cost." The results are shown in Figure 57, as measured by a number of physicians' calls for the treatment of ma- laria in the community. The second example shows the effectiveness of con- trol of yellow fever, another disease for which definite scientific knowledge exists as to etiology and mode of transmission. Nothing could more convincingly demonstrate than does Figure 58 the effectiveness with which this disease can be controlled. The diagram shows the results of the International Health Board's yellow fever work in Guayaquil in 1918-1920. ! THE POPULATION PKOBLEM Turning to another phase of the problem, it is appar- ent that if, as a result of sanitary and hygienic activi- ties and natural evolution, the average duration of human life is greater now than it used to be and is getting greater all the time, then clearly there must be more people on the earth at any time, out of a given number NATURAL DEATH, PUBLIC HEALTH 241 too S90 seo S40 620 SDO 460 460 440 420 400 SSo 360 34O Malaria Control at CROSSETT -ARKANSAS Calls for Malaria 1915 1916 1917 1918 I 1 5S0 1 1 AX> 1 4 *» >, 420 400 300 360 340 320 B! U 5 1 • I liH'H 5TO 1 '•'• 300 I 1 | II J 230 tto 240 220 ZOO 180 160 140 120 IOO 80 60 40 20 O 200 MO i 9 B i 1 II S V INI s. J; Ai^«l .11 1 r J ) «0 ito 1*0 too to *> 0 X 1 : f •III HlH! T - ' ~ 1 II fl •1 •H 1 11 1 1 ! illH I'H I • III LIB-M | •I | ill 1 1 II II III 1 1 i t i f fl (9 I i i B B p ^ _ G 1 8 \MAKCH APRIL | I II I ^ 9 O ^ 1 MAKCH i APRIL MAV i T «! I 1 s MAKCH 1 1 | X -5 i 1 § <; 1 5: 5 MAKCH APKIL. X 1 X ? i kl 8 V. O § Monthly Distribution of Calls Population, 2G2<) '9/5 1)16 iff /9/8 Total Cat Is ijis 2500 /9/s "»- 741 /9/7 - 200 /9/s - 73 JAM. 45 4O 63 FEB. 45 39 7 2 MARCH S& 59 13 4 j/*/*. oo 61 12 8 H»X 80 114 31 2 ytw£ 120 98 15 6 vt/ir 200 95 95 AU6. 350 91 33 7 s£A>r SCO 54 22 u ocr 600 46 14 8 uov. 350 20 23 7 /up 100 4 15 !O Percentage of Reduction, ijis-ijis 97.1 Per Capita Cost: 19/6 - 1.24 19/7 - -63 /9/S - -53 FIG. 57. — Record of malaria control by anti-mosquito measures, Crossett, Ark. 1916-1918. 242 BIOLOGY OF DEATH MONTHS I-' i . 5 ;. — Disappearance of yellow fever from Guayaquil, Ecuador, as a result of control nfirmis^irm nf Tnfprnft tinnnl FT^olth NATURAL DEATH, PUBLIC HEALTH 243 born, than was formerly the case. It is furthermore plain that if nothing happens to the birth-rate there must eventually be as many persons living upon the habitable parts of the globe as can possibly be supported with food and the other necessities of life. Malthus, whom every one discusses but few take the trouble to read, pointed out many years ago that the problem of popu- lation transcends, in its direct importance to the welfare of human beings and forms of social organization, all other problems. Lately we have had a demonstration on a ghastly gigantic scale of the truth of Malthus' conten- tion. For, in last analysis, it cannot be doubted that one important underlying cause of the great war, through which we have just passed, was the ever-growing pres- sure of population upon subsistence. Any system or form of activity which tends, by how- ever slight an amount, to keep more people alive at a given instant of time than would otherwise remain alive, adds to the difficulty of the problem of population. We have just seen that this is precisely what our public-health activities aim to do, and in which they succeed in a not inconsiderable degree. But someone will say at once that, while it is true that the death-rate is falling more or less generally, still the birth-rate is falling concomi- tantly, so we need not worry about the population prob- lem. It is evident that if we regard the population problem in terms of world-area, rather than that of any particular country, its degree of immediacy depends upon the ratio of births to deaths in any given time unit. If we examine, as I have recently done, these death-birth ratios for different countries, we find that they give us little hope of any solution of the problem of population 244 BIOLOGY OF DEATH by virtue of a supposed general positive correlation be- tween birth-rates and death-rates. The relation of birth-rate and death-rate changes to population changes is a simple one and may be put this way. If, neglecting migration as we are justified in doing in the war period and in considering the world prob- lem, in a given time unit the percentage 100 Deaths Births has a value less than 100, it means that the births exceed the deaths and that the population is increasing within the specified time unit. If, on the other hand, the per- centage is greater than 100, it means that the deaths are more frequent than the births and that the population is decreasing, again within the specified time unit. The TABLE 29 Percentage of Deaths to Births Year 77 non-invaded departments of France Prussia Bavaria England and Wales 1913 97 per cent. 58 per cent. 57 per cent. 1914 110 per cent. 66 per cent. 74 per cent. 59 per cent. 1915 169 per cent. 101 per cent. 98 per cent. 69 per cent. 1916 193 per cent. 117 per cent. 131 per cent. 65 per cent. 1917 179 per cent. 140 per cent. 127 per cent. 75 per cent. 1918 1919 198 per cent. 154 per cent. 132* per cent. 146 per cent. 92 per cent. 73 per cent. 1920 42* per cent. First three-fourths of year only. ratio of deaths to births may be conveniently designated as the vital index of a population. From the raw data of births and deaths, I have cal- culated the percentage which the deaths were of the births for (a) the 77 non-invaded departments of France; (b) NATURAL DEATH, PUBLIC HEALTH 245 Prussia; (c) Bavaria; and (d) England and Wales, from 1913 to 1920 by years. The results are shown in Table 29. The points to be especially noted in Table 29 are : 1. In all the countries here dealt with the death-birth ratio in general rose throughout the war period. This means that the proportion of deaths to births increased so long as the war continued. 2. But in England it never rose to the 100 per cent, mark. In other words, in spite of all the dreadful effects of war, England's population wenti on making a net increase throughout the war. 3. Immediately after the war was over, the death- birth ratio began to drop rapidly in all countries. In England in 1919 it had dropped back from the high figure of 92 per cent, in 1918 to 73 per cent. In France it dropped from the high figure of 198 in 1918 to 154 in 1919, a lower figure than France had shown since 1914. In all the countries the same change is occurring at a rapid pace. Perhaps the most striking possible illustration of this is the history of the death-birth ratio of the city of Vienna, shown in Figure 4, with data from the United States and England and Wales for comparison. Prob- ably no single large city in the world was so hard hit by the war as Vienna. Yet observe what has happened to its death-birth ratio. Note how sharp is the decline in 1919 after the peak in 1918. In other words, we see how promptly the growth of population tends to regulate itself back towards the normal after even so disturbing an upset as a great war. In the United States, the death-birth ratio was not affected at all by the war, though it was markedly altered by the influenza epidemic. The facts are shown in Fig- ure 59 for the only years for which data are available. 246 BIOLOGY OF DEATH The area covered is the United States birth registration area, We see that with the very low death-birth ratio of 56 in 1915, there was no significant change till the influenza year 1918, when the ratio rose to 73 per cent. £25 200 I7S I 8 100 75 SO STATES' 19/3 &>5 1917 1920 YEAR FIG. 59. — Showing the change in percentage which deaths were of births in each of the years 1912 to 1919 for Vienna ( - ); 1915 to 1919 for the United States ( --- );and 1912 to 1920 for England and Wales ( ----- ). But in 1919, it promptly dropped back to the normal value of 57.98, almost identical with the 1917 figure of 57.34. In England and Wales, the provisional figure indi- cates that 1920 will show a lower value for the vital index than that country has had for many years. So we see that neither a highly destructive war, nor the most destructive epidemic since the Middle Ages, serves more than to cause a momentary hesitation in the steady onward march of population growth. NATURAL DEATH, PUBLIC HEALTH 247 The first thing obviously needed in any scientific approach to the problem of population is a proper mathe- matical determination and expression of the law of popu- lation growth. It has been seen that the most devastating calamities make but a momentary flicker in the steady progress of the curve. Furthermore, population growth is plainly a biological matter. It depends upon, in last analysis, only the basic biological phenomena of fertility and mortality. To the problem of an adequate mathe- matical expression of the normal growth of populations, my colleague, Dr. Lowell J. Reed, and I have addressed ourselves for some time past, The known data upon which we have to operate are the population counts given by successive censuses. Various attempts have been made in the past to get a mathematical representation of these in order to predict successfully future populations, and to get estimates of the population in inter-censal years. A noteworthy attempt of this sort is Pritchett's fitting of a parabola of the third order to the United States popu- lation from 1790 to 1880 inclusive. This gave a fairly good result over the period, but was obviously purely empirical, expressed no real biological law of change, and in fact failed badly in prediction after 1890. We have approached the problem from an a priori basis, set up a hypothesis as to the more important biological factors involved, and tested the resulting equation against the facts for a variety of countries. The hypothesis was built up around the following considerations : 1. In any given land area of fixed limits, as by political or natural boundaries, there must necessarily be an upper limit to the number of persons that can be sup- ported on the area. To take an extreme case, it is obvious 248 BIOLOGY OF DEATH that not so many as 25,000 persons could possibly stand upon an acre of ground, let alone live on it. So, similarly, there must be for any area an upper limiting number of persons who can possibly live upon it. In mathematical terms this means that the population curve must have an upper limiting asymptote. 2. At some time in the more or less remote past the population of human beings upon any given land area must have been nearly or quite zero. So the curve must have somewhere a lower limiting asymptote. 3. Between these two levels we assume that the rate of growth of the population, that is, the increase in numbers in any given time unit, is proportional to two things, namely: a. The absolute amount of growth (or size of population) already attained ; b. The amount of as yet unutilized, or reserve, means or sources of subsistence still available in the area to support further population. These hypotheses lead directly to a curve of the form shown in Figure 60, in which the position of the asymp- totes and of the point of inflection, when the population is growing at the most rapid rate, are shown in terms of the constants. It is seen that the whole history of a population, as pictured by this curve, is something like this: In the early years following the settlement of a country1 the population growth is slow. Presently it begins to grow faster. After it passes the point where half the available resources of subsistence have been drawn upon and utilized, the rate of growth becomes slower, until finally the maximum population which the area will support is reached. NATURAL DEATH, PUBLIC HEALTH 249 This theory* of population growth makes it possible to predict what the maximum population in a given area will be, and when it will be attained. Furthermore, one can tell exactly when the population is growing at the maximum rate. To test the theory, we have only to fit Fio. 60. — Showing a theoretical curve of population growth. this theoretical curve to the known facts of population for any country by appropriate mathematical methods. If the hypothesis fits well all the known facts for a variety of countries in different stages of population growth, it may well be regarded as a first approximation to a sub- stantially correct hypothesis and expressive of the bio- logical law according to which population grows. In making this test the statistician has somewhat the same * The mathematical hypothesis here dealt with is essentially the same as that of Verhulst, put forth in 1844. As Pearl and Reed pointed out in their first paper on the subject it is a special case of a much more general law. A comprehensive general treatment of the problem we are publishing shortly in another place. The generalization in no way alters the conclusions drawn here from a few illustrative examples. 250 BIOLOGY OF DEATH kind of problem that confronts the astronomer calculat- ing the complete orbit of a comet. The astronomer never has more than a relatively few observations of the posi- ns IZS 75 SO ZS UNITED STATES L_L__LJ~ 20 4O 6O 8O I6OO 20 4O 60 8O I9OO ^0 4O 6O SO ZOOO £O 4O 6O SO ^IOO VEARS FIG. 61. — Showing the curve of growth of the population of the United States. For further explanation of this and the two following diagrams, see text. tion of the comet. He has, from Newtonian principles, a general mathematical expression of the laws of motion of heavenly bodies. He must then construct his whole curve from the data given by the few observations. So, similarly, the statistician has but a relatively few popu- lation observations because census taking has been prac- tised along present lines only a little more than a century. According to the stage in historical development of the country dealt with, he may have given an early, a late, or a middle short piece of the population " orbit" or his- tory. From this he must construct, on the basis of his general theory of "population orbits, " the whole history, past and future, of the population in question. To demonstrate how successful the population curve shown in Figure 60 is in doing this, three diagrams are presented, each illustrating the growth of the population NATURAL DEATH, PUBLIC HEALTH 251 in a different country. The heavy solid portion of each curve shows the region for which census data exist. The lighter broken part of the curve shows the portions out- side this observed range. The circles show the actual, known observations. The first curve deals with the popu- 41360 30 is " » FRANCE: I6OO 20 AO 60 6O 1700 20 4O 6O QO IQOO ZQ 4O 6O SO I9OO 2O 40 6O 6O 2OOO YEARS FIG. 62. — Showing the curve of growth of the population of France. lation of the United States. Here the observations come from the first part of the curve, when the population was leaving the lower asymptote. First should be noted the extraordinary accuracy with which the mathematical theory describes the known facts. It would be extremely difficult, by any process, to draw a curve through the ob- served circles and come nearer to hitting them all than this one does. Before considering the detailed consequences of this United States curve in relation to the whole population history of the country, let us first examine some curves for other countries, where the observed data fell in quite different portions of the "population orbit." Fig-ure 62 252 BIOLOGY OF DEATH gives the curve for France. Since before the time when definite census records began, France has been a rather densely populated country. All the data with which we had to work, belong therefore, towards the final end of the whole population history curve. The known popula- tion data for France and for the United States stand at opposite ends of the whole historical curve. One is an old country whose population is nearing the upper limit; the other a new country whose population started from near the lower asymptote only about a century and a half ago. But it is seen from the diagram that the general theory of population growth fits perfectly the known facts regarding France 's population in the 120 years for which records exist. While there are some irregularities in the observation, due principally to the effects of the Franco- Prussian war, it is plain that on the whole it would be practically impossible to get a better fitting line through the observational circles than the present one. We have seen that the general theory of population describes with equal accuracy the rate of growth in a young country, with rapidly increasing population, and an old country, where the population is approaching close to the absolute saturation point. Let us now see how it works for a country in an intermediate position in respect of population. Figure 63 shows the population history of Serbia. Here it will be noted at once that the heavy line, which denotes the region of known census data, lies about in the middle of the whole curve. Again the fit of theory to observation is extraordinarily close. No better fit, by a general law involving no more than 3 con- stants, could possibly be hoped for. I think that these three examples, which could be multiplied to include practically every country for which NATURAL DEATH, PUBLIC HEALTH 253 accurate population data exist, furnish, a cogent demon- stration of the essential soundness and accuracy of this theory of population growth. Indeed, the facts warrant, I believe, our regarding this as a first approximation to the true natural law of population growth. We now are 4.388 P SERVIA ..L.4-T- 1700 20 4O 60 60 1800 20 40 6O 6O I9OO 2O 4O 60 80 2000 2O 4O 6O 60 2JOO YEARS FIG. 63. — Showing the curve of growth of the population of Serbia. approaching the proper mathematical foundation on which to build sociological discussions of the problem of population. As a further demonstration of the soundness of this theory of population growth, let attention be directed for a moment to an example of its experimental verification. To a fruit fly (Drosophila) in a half pint milk bottle, such as is used in experimental work on these organisms, the interior of the bottle represents a definitely limited uni- verse. How does the fly population grow in such a uni- verse? We start a bottle with a male and female fly, and a small sample, say 10, of their offspring of different ages (larvae and pupae). The results are shown in Fig- 254 BIOLOGY OF DEATH ure 64. The circles give the observed population growth, obtained by census counts at 3-day intervals. There can be no doubt that this population has grown in accordance with the equation. The two final observations lie below the curve, because of the difficulty experienced, in this 3ZS 3OO Z7S Z50 ZZ& 20° /7f ISO 125 IOO 75 SO /• GROWTH Of DROSOPHILA POPULATION z 16 OCT. II NOV. Fio. 64. — Showing the growth of a Drosophila population kept under controlled experimental conditions. particular experiment, of keeping the food supply in good condition after so long a period from the start. Let us return to the further discussion of the popu- lation problem of the United States in the light of the curve. The first question which interests one is this: When did or will the population curve of this country pass the point of inflection and exhibit a progressively diminishing instead of increasing rate of growth? It is easily deter- mined that this point occurred about April 1, 1914, on the assumption that our present numerical values reliably rep- resent the rate of population growth in this country. In other words, so far as we may rely upon present nu- merical values, the United States has already passed its period o,f most rapid population growth, unless there NATURAL DEATH, PUBLIC HEALTH 255 comes into play some factor not now known, and which has never operated during the past history of the country, to make the rate of growth more rapid. The latter con- tingency appears improbable. The 1920 census confirms the result, indicated by the curve, that the period of most rapid population growth was passed somewhere in the last decade. The population at the point of inflection works out to have been 98,637,000, which was, in fact, about the population of the country in 1914. The upper asymptote given by the equation has the value of 197,274,000 roughly. This means that the maxi- mum population which continental United States, as now areally limited, will have, will be roughly twice the pres- ent population; provided no fundamental new factor comes into play in the meantime, different in its magni- tude and mode of operation from any of the factors which have influenced population growth in the past. This state of affairs will be reached in about the year 2,100, a little less than two centuries hence. Perhaps it may be thought that the magnitude of this number is not suffi- ciently imposing. It is so easy, and most writers on population have been so prone, to extrapolate population by geometric series or by a parabola or some such purely empirical curve, and arrive at stupendous figures, that calm consideration of real probabilities is most difficult to obtain. While we regard the numerical results as only a rough first approximation, it remains a fact that if anyone will soberly think of every city, every village, every town in this country having its present population multiplied by 2, and will further think of twice as many persons on the land in agricultural pursuits, he will be bound, we think, to conclude that the country would be 256 BIOLOGY OF DEATH fairly densely populated. It would have about 66 per- sons per square mile of land area. It will at once be pointed out that many European countries have a much greater density of population than 66 persons to the square mile, as, for example, Belgium with 673, the Netherlands with 499, etc. But it must not be forgotten that these countries are far from self- supporting in respect of physical means of subsistence. They are, or were before the war, economically self- supporting, which is a very different thing, because, by their industrial development at home and in their colo- nies, they produce money enough to buy physical means of subsistence from less densely populated portions of the world. We can, of course, do the same thing, pro- vided that by the time our population gets so dense as to make it necessary, there still remain portions of the globe where food, clothing material and fuel are produced in excess of the needs of their home populations. Now 197,000,000 people will require, on the basis of our present food habits, about 260,000,000 million calories per annum. The United States, during the seven years 1911-1918, produced as an annual average, in the form of human food, both primary and secondary (i.e., broadly vegetable and animal), only 137,163,606 million calories per year. So that, unless our food habits radically change, and a man is able to do with less than 3,000 to 3,500 calories per day, or unless our agricultural production radically increases, which it appears not likely to do for a variety of reasons which cannot be here gone into, it will be necessary, when even our modest figure for the asymptotic population is reached, to import nearly or quite one-half of the calories necessary for that population. It seems improbable that the population will go on increasing at NATURAL DEATH, PUBLIC HEALTH 257 any very rapid rate after such a condition is reached. East has shown that the United States has already entered upon the era of diminishing returns in agriculture in this country. Is it at all reasonable to suppose that by the time this country has closely approached the asymptote here indicated, with all the competition for means of sub- sistence which the already densely populated countries of Europe will then be putting up, there can be found any portion of the globe producing food in excess of its own needs to an extent to make it possible for us to find the calories we shall need to import? Altogether we believe it will be the part of wisdom for anyone disposed to criticize our asymptotic value of a hundred and ninety-seven and a quarter millions because it is thought too small, to look further into all the rele- vant facts. This point of view is sustained in a recent paper by East in which the future agricultural resources of the country are particularly examined. The relation of this already pressing problem of popu- lation to the problem of the duration of life is obvious enough. For every point that the death rate is lowered (or, what is the same thing, the average duration of life increased) the problem of population is made more imme- diate and more difficult unless there is a corresponding decrease in the birth-rate. Is it to be wondered at that most thoughtful students of the problem of population are advocates of birth control? Or is it remarkable that Major Leonard Darwin, president of the Eugenics Education Society in England, should say in a carefully considered memorandum to the new British Ministry of Health: "In the interests of posterity it is most desirable that parents should now limit the size of their families by any means held by them to be right (provided such 17 258 BIOLOGY OF DEATH means are not injurious to health, nor, like abortion, an offense against public morals) to such an extent that the children could be brought up as efficient citizens and with- out deterioration in the standards of their civilization; and that parents should not limit the size of the family for any other reasons except on account of definite hered- itary defects, or to secure an adequate interval between births." I am able to make no prediction as to how civilized countries will solve (if they do solve) the problems arising out of the impending saturation with human popu- lation of the portion of the earth's surface habitable by man. The certainty and assurance with which various ones of my friends advance solutions excites my wonder and admiration. But what impresses me even more is that scarcely any two of them agree on the nature of the panacea. To some it is birth control, to others synthetic foods derived from the atmosphere or else- where, and so on. For myself, I am content if I have succeeded, in even a, small measure, in indicating that population growth pre- sents a problem fast becoming urgent; a problem that in its overwhelming significance and almost infinite rami- fications touches upon virtually every present human ac- tivity and interest, and in particular upon the activities comprised in the terms public health and hygiene. BIBLIOGRAPHY The following list of literature in no sense aims at completeness within the field covered. It is, in the main, made up only of the sources which have been consulted in the preparation of the present volume. It is hoped, however, that even with this limitation it may serve as a useful introduction to the literature for any who may wish to pursue further their reading on the subjects here dealt with. AMMA, K. Uber die Differenzierung der Keimbahnzellen bei den Kope- poden. Arch. f. Zellforsch, Bd. VI, IS 11. BATAILLON. La parthenogenese experimentale des amphibiens. Rev. gen. d. Sci. T. XXII, p. 786, 1911. See also Comp. rend. Acad. Sci. Paris, T. CL, 996, 1910; T. CLII, 920, 1911; T. CLII, pp. 1120 and 1271, 1911; T. CLVI, 812, 1913; Arch, de Zool. exper. et gen.t T. XLVI, p. 103, 1910. BEETON, M. and PEARSON, K. Data for the problem of evolution in man. II. — A first study of the inheritance of longevity, and the selective death rate in man. Proc. Roy. Soc. Vol. LXV, pp, 290-305, 1899. BEETON, M. and PEARSON, K. On the inheritance of the duration of life, and on the intensity of natural selection in man. Biometrika, Vol. 1, pp. 50-89, 1901. BELL, A G. The duration of life and conditions associated with longevity. A study of the Hyde genealogy. Washington, 1918. pp. 57, 4to. ( Privately printed ) . BENEDICT, HARRIS M. Senile changes in leaves of Vitis vulpina L. and certain other plants. Cornell Agr. Expt. Stat. Mem. 7, pp. 273- 370, 1915. BERTILLON, J. Morbidity and mortality according to occupation. Jour. Roy. Stat. Soc., Vol. LV, pp. 559-600, 1892. BOGDANOW, E. A. tiber das Ziichten der gewuhnlichen Fleischfliegen (Calliphora vomitaria) in sterilisierten Nahrmitteln. Arch, f. d. ges. Physiol. 1906 BOGDANOW, E. A. Uber die Abhiingigkeit des Wachstums der Fliegen- larven von Bakterien und Fermenten und tiber Variabilitat und Verer- bung bei den Fleischfliegen. Arch. f. Anat. u. Physiol., (Physiol. Abth.) 1908, pp. 173-199, 1908. BULLOCH, W. and GREENWOOD, M. The problem of tuberculosis considered from the standpoint of disposition. Proc. Roy. Soc. Med., May, 1911, Vol. IV, Epidem. Sect., pp. 147-184. 259 260 BIBLIOGRAPHY BURROWS, M. T. The tissue culture as a physiological method. Trans. Cong. Amer. Phys. and Surg., Vol. IX, pp. 77-90, 4 plates, 1913. CARREL, A. On the permanent life of tissues outside of the organism. Jour. Exper. Med. Vol. XV, pp. 516-528, 2 plates, 1912. CARREL, A. Present condition of a strain of connective tissue twenty- eight months old. Jour. Exper. Med. Vol. XX, pp. 1-2, 2 plates, 1914. CARREL, A. and BURROWS, M. T. Cultivation of tissues in vitro and its technique. Jour. Exper. Med. Vol. XIII, pp. 387-396, 1911. CARREL, A. and EBELING, A. H. The multiplication of fibroblasts in vitro. Jour. Exp. Med. Vol. 34, pp. 317-337, 1921. CARREL, A. and EBELING, A. H. Age and multiplication of fibroblasts. Ibid. Vol. 34, pp. 599-623, 1921. CHILD, C. M. Senescence and Rejuvenescence, Chicago, 1915, pp. 481. COHNHEIM, J. Vorlesungen iiber allgemeine Pathologic, 2te Aufl., Berlin, 1882. COLLIS, E. L. and GREENWOOD, M. The Health of the Industrial Worker. London, 1921, 450 pp. CONKLIN, E. G. The size of organisms and of their constituent parts in relation to longevity, senescence and rejuvenescence. Pop. Sci. Mo. August, 1913, pp. 178-198. CRUM, F. S. The effect of infant mortality on the after-lifetime of sur- vivors. Trans, llth Ann. Meet. Amer. Child. Hyg. Ass., 1920, 17 pp. DAWSON, J. A. An experimental study of an amicronucleate Oxytricha. I. — Study of the normal animal, with an account of cannibalism. Jour. Exp. Zool. Vol. 29, pp. 473-512, 2 pi., 1919. DAWSON, M. M. Practical Lessons in Actuarial Science. 2 vols. New York, 1905. DELAGE, Y. L'Heredite et les grandes probl^mes de la biologie. Paris, 1903. DELCOURT, A., and GUYENOT, E. De la possibilite d'etudier certains Dip- teres en milieu defini (Drosophila). C.R. Ac. Sci., Paris, T.5, pp. 255-257, 1910. DELCOURT, A., and GUYENOT, E. Variation et milieu. Lignees de Droso- philes en milieu sterile et defini. C.R. IV Conf. Int. Gen., pp. 478- 487, 1911. DELCOURT, A., and GUYENOT, E. Genetique et milieu. NScessite de la determination des conditions; sa possibilite chez les Drosophiles. Technique. Bull. Scient. France Belg. T. XLV, pp. 249-333, 1911. DOFLEIN, FR. Das Problem des Todes und der Unsterblichkeit bei den Pflanzen und Tieren. Jena, 1919. DONALDSON, H. H. The Growth of the Brain: a Study of the Nervous System in Relation to Education. London, 1895, 374 pp. DUBLIN, L. I. A life table for the city of New Haven. Amer. Pull. Health Jour. Vol. VIII, pp. 580-581, 1918. BIBLIOGRAPHY 261 DUBLIN, L. I., with the collaboration of KOPF, E. W. and VAN BUREN, G. H. Mortality Statistics of Insured Wage-earners and their Families. New York, 1919, pp. viii and 397. EAST, E. M. Population. 8ri. Monthly, June 1920, pp. 603-624. EAST, E. M. The agricultural limits of our population. Ibid. Vol. XII, June 1921, pp. 551-557. EAST, E. M. and JONES, D. F. Inbreeding and Outbreeding. Their Gene- tic and Sociological Significance. Philadelphia, 1919, pp. 285. In the bibliography of this book will be found references to numerous papers by East and his students on heterosis in maize, etc. EBELING, A. H. A strain of connective tissue seven years old. Jour. Esoper. Med. Vol. XXX, pp. 531-537, 5 plates, 1919. ENRIQUES, P. La morte. Scientia. T. I. 1907. ENRIQUES, P. Duemila cinquecento generazione in un infusorio, senza conjugazione ne parthenogenesi, ne depressione. Rend, della R. Accad. d. Sci. dell'Ist. di Bologna, 1916, pp. (of reprint) 12. ERDMANN, R. and WOODRUFF, L. L. The periodic reorganization process in Paramecium caudatum. Jour, Eocper. Zool. Vol. 20, pp. 59-97, 1916. FISHER, ARNE. On the construction of mortality tables by means of compound frequency curves. Scandinavian Insurance Magazine, 1920, passim. FORSYTH, C. H. Vital and monetary losses in the United States due to preventable deaths. Quart. Pull. Amer. Stat. Assoc. Vol. 14, pp. 758-789, 1915. FRIEDENTHAL, H. tiber die Giltigkeit der Massenwirkung fiir den Ener- gieumsatz der lebendigen Substanz. II Teil. Zentralbl. f. Physiol. Bd. 24, pp. 321-327, 1910. GIVEN, D. H. C. Some deductions from the statistics on the prevention of pulmonary tuberculosis. British Med. Jour., Feb. 12, 1921, p. 225. GLOVER, J. W. United States Life Tables, 1910. Bureau of the Census, 1916, pp. 65. GREENWOOD, M. Infant mortality and its administrative control. Eugen- ics Rev. October, 1912, pp. (of reprint) 1-23. GREENWOOD, M. and BROWN, J. W. An examination of some factors influenc- ing the rate of infant mortality. Jour. Hyg. Vol. XII, pp. 5-45, 1912. GROTH und HAHN. Die Saiiglingsverhaltnisse in Bayern. Miinchen (Lind- auer) 8onderabdr. a.d. Ztschr. d.k. Bayer, Stat. Landesamts. Jahrg. 1910. GURNEY, J. H. The Comparative Ages to which Birds Live. Ibis, 1899, p. 19. GUYENOT, E. Recherches sur la vie aseptique et la developpement d'un organisme en fonction du milieu. Thesis, Paris, 330 pp , 4 plates. 262 BIBLIOGRAPHY GUYER, M. E. The development of unfertilized frog eggs injected with blood. Science, N.S. Vol. XXV, p. 910, 1907. HALLEY, EDM. An estimate of the degrees of mortality of mankind drawn from tables of births and burials in the city of Breslau, with an ^attempt to ascertain the price of annuities on lives. Phil. Trans. No. 196, p. 596. HALLEY, EDM. Some further considerations on the Breslau bills of mor- tality. Ibid. No. 198, p. 654. (Both of the above papers are reprinted in adequate abstract form in Mem. Roy. Soc. By Mr. M. Baddam, Vol. Ill, pp. 34-43; and pp. 52-53, 1739). HARPER, M. Zur Entwicklung der Geschlechtsorgane von Chironomus. Zool Jahrbuch.; Abth. f. Anat. u. Ont. Bd. XXXI, 1911. HARRISON, R. G. The outgrowth of the nerve fiber as a mode of proto- plasmic movement. Jour. Exper. Zool. Vol. IX, pp. 787-846, 1910. HARRISON, R. G. The life of tissues outside the organism from the embryo- logical standpoint. Trans. Cong. Amer. Phys. and Surg. Vol. IX, pp. 63-76, 12 plates, 1913. HARTMANN, M. Uber die dauernde rein agame Ziichtung von Eudorina elegans und ihre Bedeutung fiir das Befruchtungs — und Todproblem. Sitz.-Ber. d.kgl. Akad. d. Wiss. Berlin, Bd. 52, 1917. HEGNER, R. W. The origin and Early History of the Germ Cells in Some Chrysomelid Beetles, Jour, of Morphol., XX, 1909. HEGNER, R. W. Experiments with Chrysomelid Beetles: III, The Effects of Killing Parts of the Eggs of Leptinotarsa decemlineata. Biol. Bull. XX, 1911. HEGNER, R W. The History of the Germ Cells in the Paedogenetic Larva? of Miastor. Science XXXVI. 1912. HEGNER, R. W. Studies on Germ Cells: I, The History of the Germ Cells in Insects with Special Reference to the Keimbahn-Det erminants ; II, The Origin and Significance of the Keimbahn-Determinants in Ani- mals." Jour, of Morphol., XXV. 1914. HEGNER, R. W. Studies on Germ Cells: III, The Origin of the Keimbahn- Determinants in a parasitic Hymenopteron, Copidosoma. Anat. Anz., XLVI. 1914, HEGNER, R. W. The Germ-Cell Cycle in Animals. New York. 1914. HENDERSON, R. Mortality Laws and Statistics. New York, 1915, pp. v and 111. HERON, D. On the relation of fertility in man to social status, and on the changes in this relation that have taken place during the last fifty years. Draper's Company Res. Mem. I, pp. 1-22, 1906. HERSCH, L. L'inegalite devant la mort d'apres les statistiques de la Ville de Paris. Effets de la situation sociale sur la mortalite. Rev. d'Econ. pol. Nr. 3 and 4, 1920, 54 pp. BIBLIOGRAPHY 263 HERTWIG, R. Ueber Korrelation von Zell- und Kerngrosse und ihre Be- deutung fur die geschlechtliche Differenzierung und die Teilung der Zelle. Biol Centralbl. Bd. XXIII, 1903. HERTWIG, R. Ueber neue Probleme der Zellenlehre. Arch f. Zellforsch. Bd. I. 1908. HODGE, C. F. Changes in ganglion cells from birth to senile death. Jour. Physiol. Vol. XVII, 129, 1894. HOWARD, W. T. Senescence and natural death. Cleveland Med. Jour. Vol. IX, pp. 730-751, 1910. HYDE, R. R. Inheritance of the length of life in Drosophila, ampelophila. Rept. Indiana Acad. Sci. 1913, pp. 113-123. JENNINGS, H. S. Behavior of the Lower Organisms. New York, 1906, 366 pp. JENNINGS, H. S. Assortative mating, variability and inheritance of size, in the conjugation of Paramecium. Jour. Exper. Zool, Vol. XI, pp. 1-134, 1912. JENNINGS, H. S. Age, death and conjugation in the light of work on lower organisms. Pop. Sci. Monthly, June, 1912. JENNINGS, H. S. Life and Death, Heredity and Evolution in Unicellular Organisms. Boston, 1920, pp. 233. JICKELI, C. F. Die Unvollkommenheit des Stoffwechsels. Berlin, 1902. JOLLOS. Die Fortpflanzung der Infusorien und die potentielle Unster- blichkeit der Einzelligen. Biol. Centralbl., 1916. KASSOWITZ, M. Allgemeine Biologic. Wien, 1899. KORSCHELT, E. Lebensdauer, Altern und Tod. Zweite Aufl. Jena, 1922, 307 pp., 8 vo. LANKASTER, E. R. On comparative longevity in man and the lower animals. London, 1870. LEGRAND, M. A. La longevite M., 44, 265 211, 214, 215, 223, 226, 263, 264 Maller j^ 44 Loeb, L., 59, 64, 65, 67, 224, 264 Logarithmic plotting, 114 London, 205-208 Longevity, body size and, 26 evolutionary progress in, Nascher, L, 26, 27, 265 87-94 of animals, 22 of parents, 158, 160 Lowell Institute, 9, 27 Nerve cells, senile changes in, 27-29 Nervous system, 107, 108, 130, 131 Netherlands, 256 Non-controlled causes of death, 232, 233 Macdonell, W. R., 87, 89-93, 229, 264 Northrop, J. H., 200, 201, 209-211, Malaria, 238-241 Malthus, T. R., 243 Mammals, longevity of, 22 Man, longevity of, 23-26, 80-94 Marmoset, 68 Mendelian inheritance, 194, 197, 198 Mesoderm, 138-149 Metabolic activity, 211-217 Metazoa, 31, 33, 40, 46, 71 214, 215, 226, 264, 265 Nucleus, 29, 30, 44 Occupation, 216 Ogle, W., 95 Orbits, 250 Organ systems in mortality, 107, 108 most fatal, 136 Oxytricha hymenostoma, 73 274 INDEX Quaker records, 171, 173 Rat, 61, 212, 213, 218 Paralysis, 231, 232 Paramccium, 30-32, 35, 40, 72 Parental correlations, 171, 172, 174, Rabbit, 68 176 Parents and offspring, duration of Ratio> death-birth, 243-246 life of, 155-157 Rav> L- A-> 69> 70> 220> 221> 266 longevity of, 158, 160 Reed> L- J-> 247' 249> 266 Paris, 202-206 Parr, T., 24 Parthenogenesis, artificial, 51-58, 223 Reproduction, organic, 33, 41 Pascal, 82 bv budding, 37 Pearl, R., 106, 201, 225, 249, 265 bv fission, 32, 33, 41 Pearson, K., 19, 87-91, 93-101, 166, clonal, 37 sexual, 37-40, 41 Reptile, longevity of, 22 Respiratory system, 107, 108, 110, Registration Area, U. S., 106, 108, 109, 139, 140, 164, 229, 245, 246 169-177, 179, 182, 183, 225, 229, 238, 259, 266 Peerage, 167, 169, 172 Pennaria, 62 112, 119, 120, 136, 137 Physical characters, inheritance of, Results, summary of, 223-227 174, 175 Pituitary gland, 220-222 Pixell-Goodrich, Mrs., 28 Planaria dorotocephala, 34, 35 Plants, senility in, 44 Richards, H. A., 86, 87, 266 Ritter, W. E., 75, 266 Robertson, T. B., 69, 70, 220, 221, 266 Rockefeller Foundation, 238 Institute, 52, 61 Ploetz, A., 178, 179, 182, 183, 225, R6ie of bacteria in duration of life, 266 Population, 240-258 Potassium cyanide, 53, 54 Poverty, 202-208 Premature birth, 121, 123 Preventability of diseases, 162 Pritchett, A. S., 247, 266 Progress, evolutionary, in longevity, ROundworm, 39 43, 199-202 Roman Africa, expectation of life in, 92 Rome, expectation of life in, 90-92 Romeis, B., 219, 266 Rose, W., 238, 239, 241, 266 Roumania, 235 87-94 Prolonging life, 17, 54, 218, 221 Prostate, 126, 219 Protozoa, 30-33, 40, 41, 46 Royal families, 177 Rubner, M., 213, 214, 226, 267 Saleeby, 183 immortality of. 30-33, 41, Sanitation, 227, 235 64, 71 Sao Paulo, 106, 108-111, 139, 140 Sea urchin, 52, 54, 57 Selection, effect of, on expectation of life, 94 Selective death rate, 177-185 Seneca, 102 Prussia, 244, 245 Puberty glands, 217-219 Public health work, effects of, 112, 227-242 Purulent infection, 231, 232 INDEX 275 Senescence, 27-30, 46, 70-7-S Table., life., 70-82 theories of, 43-50 Temperature, 208-217 Senile changes in nerve cells, 27-29 Tethelin, 70, 220-222 Senility as cause of death, 109 Theories of death, 43-50 in plants, 44, 74, 75 Theory of population growth, 249 Septicaemia, 231, 232 Thyroid gland, 61 Serbia, 252, 253 Tissue culture in vitro, 58-78 Serum, influence on tissue culture, Transplantation of tumors, 64, 65 Tuberculosis, 161, 204, 208, 230, 231, 238 Tumor transplantation, 64, 65 Typhoid fever, 230, 231, 235, 230 76, 77 Sex organs, 107, 108, 111, 121-125, 217-219 Sexual reproduction, 37-41 Shell, J., 26, 27 Skeletal system, 107, 108, 112, 127, United States, growth of, 250-252. 128 254-257 Skin. 107, 108, 110, 112, 131, 132 Urostyla grandis, 72 Slonaker, J. R., 212, 213, 218, 228, 267 Slotopolski, B., 33, 267 Snow, E. C., 179-183, 225, 267 Softening of the brain, 231, 232 Soma, 40 Van Buren, G. H., 113, 260 Variation, genetic, 190 Venereal diseases, 123, 124 Verhulst, P. F., 249, 267 Verworn, M., 44, 267 Somatic cells, immortality of, 58-78 Vienna, 245, 246 Span, 174, 175 Spiegelberg, W., 87 Spiritualism, 18-20 Spleen, 61 Sponges, 62 Stature, 174, 175 Steinach, E., 217-219, 267 Stenostotmim, 35, 36 Stevenson, T. H. C., 206-208, 267 Still births, 205 Voronoff, 217 Waller, A. D., 216, 267 Walworth, R. H., 152, 267 War, 243 Wedekind, 33, 267 Weismann, A., 26, 43, 65, 267 Whale, longevity of, 22 Wilson, H. V., 62, 267 Wittstein, 99 Strongylocentrotus purpuratus, 55, Woodruff, L. L., 30, 33, 72, 73, 267, 56 Summary of results, 223-227 Survivorship lines of Drosophila, 188, 192, 195 Syphilis, 123 268 Woods, F. A., 38, 39, 268 Yellow fever, 240, 242 Young, T. E.. 23-25, 268