1918 THE DURATION OF LIFE AND CONDITIONS ASSOCIATED WITH LONGEVITY A STUDY OF Tlii:; HYDE GENEALOGY ALEXANDER GRAHAM BELL LIBRARY G The Duration of Life and Conditions Associated with Longevity A Study of the Hyde Genealogy BY ALEXANDER GRAHAM BELL WASHINGTON, D. C. PUBLISHED BY THE GENEALOGICAL RECORD OFFICE 1601 THIRTY-FIFTH STREET 1918 LiERARf G WASHINGTON, D. C. PRESS OF JUDD & DETWEII.ER, INC. CONTENTS Page Introduction 5 CHAPTER I : Period of Birth 7 CHAPTER II : Duration of Life of Persons 8 CHAPTER III : Duration of Life of Males and Females 12 CHAPTER IV : Duration of Life of Fathers and Mothers 20 CHAPTER V: Duration of Life of Persons Compared with the Duration of Life of Their Parents 24 CHAPTER VI : Duration of Life of Fathers and Mothers Compared with the Num- ber of Children Born to Them 34 CHAPTER VII : Duration of Life of Persons Compared with the Ages of the Par- ents at Marriage 38 CHAPTER VIII: Duration of Life of Persons Compared with the Ages of the Parents When the Persons Were Born 41 CHAPTER IX : Duration of Life of Persons Compared with the Number of Years After the Marriage of the Parents the Persons Were Born 44 CHAPTER X: Duration of Life of Persons Compared with Number in Family (Siblings) 46 CHAPTER XI : Conclusion ; Summary of the Chief Points Touched Upon in the Preceding Chapters 49 Appendix A 64 415531 INTRODUCTION THE following work shows the results of an attempt to utilize one of our published genealogies for the preparation of statistics relating to the duration of life and to conditions associated with longevity. The details have been taken from the "Genealogy of the Hyde Family" by Reu- ben H. Wai worth, LL. D. (1864), a work relating to the descendants of William Hyde, one of the early settlers of Norwich, Conn., who died in 1681. At first sight it would seem that but little information of general interest or value could be obtained from the records of a single family ; but 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, &c. The descendants of the seventh generation, and there are hundreds of them, owed only l/64th of their blood to William Hyde, and 63/64ths 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 par- ticular family in which family traits might be expected to make their appearance. THE DATA FOR THE HYDE STATISTICS In many cases we find recorded in the Hyde Genealogy the year of birth and of death of the individuals, and of their fathers and mothers, and the date of marriage of the parents. From these details we can calculate the age at death of the individual, and of his parents ; the age of his parents when he was born ; the age of the parents at mar- riage, and the number of years after the marriage the person was born. We can also ascertain from the published Genealogy the sex of the individual and the number of his brothers and sisters. These details comprise all the data in the Hyde Genealogy which seemed to be capable of statistical analysis. In making this investigation the details were copied upon cards, so as to facilitate the preparation of statistical tables. In all, 8,797 cards have been prepared, each relating to an individual mentioned in the Hyde Genealogy. From these cards numerous statistical tables have been prepared in great detail ; and the tables have been preserved in the Beinn Bhreagh Recorder, a typewritten Record Book for the preservation of the more important results of the author's experiments and researches; and a few copies have been deposited in different places so as to lessen the danger of destruction by fire.* *One copy has been deposited in the Library of the Smithsonian Institution, Washington, D. C. ; another In the Genealogical Record Office for the Collection and Preservation of Gene- alogical Records Pertaining to Long Life (office in the Volta Building, 1601 Thirty-fifth street, Washington, D. C.) ; and the others are in the hands of private individuals. 6 : :'. .- INTRODUCTION As two of the copies are accessible to the public, it has been thought well not to bur- den the present work with detailed tables, but only to present short abstracts of the tables bringing out the general features and results, and to refer those who desire further details to the copies of the Beinn Bhreagh Recorder in the Smithsonian Institution and the Genealogical Record Office. A list of the detailed tables that have been preserved in the Beinn Bhreagh Recorder is given in Appendix A. GRAPHICAL DIAGRAMS In glancing over detailed statistical tables like those shown in the Beinn Bhreagh Recorder, the mind is bewildered by the multiplicity of details that strike the eye, and we fail to get a general impression of the whole. If, however, we express the figures by dots upon a graphical diagram, the details are seen in their proper relations to one another; and if we use large numbers and a small scale, we gain perspective. We obtain, as it were, a bi/rd j s-eye view of the whole territory explored, and can note the general lay of the land, and the relation of one part to another in a way we could not do when traveling through the woods and valleys upon the ground. A detailed statistical table brings us so close to the ground that we perceive details rather than general effects. A graphical diagram, on the other hand, reveals generaliza- tions and subordinates details; and the larger the numbers represented and the smaller the scale employed, the more do we subordinate detail to the general effect as a whole. For this reason it has been thought advisable in this work to use graphical diagrams upon a small scale, to illustrate the tables presented. The diagrams will express the generalizations perceived, and the tables give the details upon which the generalizations are based. The author desires to express his indebtedness to Miss Mabel B. McCurdy, Miss Gretchen Schmidt, Miss Catherine D. Mackenzie, Mr. John Smallwood, Mr. Melville Bell Grosvenor, and Mr. A. W. Clime for valuable aid in preparing the tables and diagrams. A. G. B. 1331 CONNECTICUT AVENUE, WASHINGTON, D. C., November 1, 1917. CHAPTER I Period of Birth This investigation relates to 8,797 persons mentioned in the Hyde Genealogy. In 7,437 cases the year of birth is known ; and in 1,360 cases, although the exact date of birth is unknown, the period of birth has been approximately ascertained. (See Table 1.) TABLE 1. Period of birth of 8,797 persons, showing the number and percentage born in the periods named. Period of birth- Number. Per cent. Stated 8797 100 o Before 1750 869 9 9 1750 and before 1775 1,463 16.6 1775 and before 1800 2,471 28 1 1800 and before 1825 2,592 29 5 1825 and before 1850 1268 14 4 1850 and after 134 1 5 FIG. 1 The majority of the persons noted were born in the eighteenth century or earlier (68 cases in the seventeenth century), and the remainder were born during the early part of the nineteenth century. Very few appeared after 1825. The statistics thus relate mainly to generations that have passed away, and very few of the people considered can be now alive (1917). The exact years of birth and other details may be ascertained by consulting the Beinn Bhreagh Recorder, Vol. VIII, pp. 133-138; 153-158; 187-219; 319-353; 387-414. As this investigation relates to the duration of life and to conditions associated with longevity, it is not so important for us to ascertain when the people were born as the ages at which they died. 7 CHAPTER II Duration of Life of Persons Although 8,797 persons are included in the Hyde Statistics, the age at death is stated in only 2,965 cases. Aggregate years lived, 102,641. Average duration of life, 34.61 years. (B. B. Rec., IX, p. 99.) In 5,832 cases the persons were either living when the Hyde Genealogy was pub- lished (1864) or the ages at death could not be ascertained. For an analysis of these cases by "latest age known to be living," see B. B. Rec., Vol. VIII, pp. 319-353. TABLE 2. Ages at death of 2,965 persons, by single years. Age. Number. Age. Number. Age. Number. Age- Number. Age. Number 237 20 39 40 27 60 27 80 28 1 227 21 42 41 33 61 22 81 36 2 116 22 39 42 26 62 31 82 18 3 70 23 51 43 22 63 19 83 20 4 49 24 42 44 25 64 25 84 17 5 43 25 39 45 34 65 23 85 17 6 30 26 39 46 24 66 29 86 14 7 24 27 33 47 28 67 20 87 10 8 19 28 33 48 28 68 29 88 8 9 17 29 33 49 27 69 20 89 16 10 18 30 36 50 30 70 31 90 8 11 15 31 34 51 31 71 30 91 7 12 23 32 34 52 21 72 18 92 4 13 12 33 31 53 27 73 32 93 4 14 15 34 22 54 27 74.. 24 94 . 2 15 20 35 27 55 22 75 13 95 3 16 20 36 29 56 25 76 ... 26 96 . 2 17 22 37 37 57 22 77 22 97 i 18 33 38 25 58 22 78 ... .20 98 ... 19 34 39 29 59 24 79 25 101 l A STUDY OF THE HYDE GENEALOGY t- w IX FIG. 2 A glance at the diagram, Fig. 2, enables us to realize at once the relatively enor- mous number of persons who died in infancy, before they were one year old. The number who died at the age of one was only slightly less; while the number who died at the ages of two and three, although considerably less, was still much greater than the number who died at any later age. The diagram conveys the general impression that there were critical periods in the lives of these persons, when the liability to death was greater than at other times. The period of early childhood was evidently a very dangerous period. After the age of three had been safely passed the danger diminished very greatly 'until the age of puberty was reached. Then came another dangerous period, the period of adolescence, reaching its maximum danger point at or about the age of twenty-three years. Beyond this point the curve pursued a substantially horizontal direction with nu- merous oscillations of slight amplitude until extreme old age was reached. The minor irregularities of the curve, Fig. 2, result largely from the small num- bers involved in the determination of any point in the diagram and from the largeness of the scale. 10 TEE DURATION OF LIFE By grouping the details together we can secure larger figures, and by plotting them upon a smaller scale we can reduce non-significant variations of the curve to a very great degree. This is well shown by Table 3, illustrated by Fig. 3, in which the ages at death are given by five-year periods instead of by single years; and in which the scale is reduced to only one-tenth of what it was in Fig. 2. TABLE 3. Aget at death of 2,965 persons, oy five-year periods. Ages at death. Number. Ages at death. Number. Stated 2,965 Under 5 699 50 and under 55 136 5 and under 10 133 55 and under 60 115 10 and under 15 83 60 and under 65 124 15 and under 20 129 65 and under 70 121 20 and under 25 213 70 and under 75 135 25 and under 30 177 75 and under 80 106 30 and under 35 157 80 and under 85 ' 119 35 and under 40 147 85 and under 90 65 40 and under 45 133 90 and under 95 25 141 7 is i if 8 3? PIG. 3 While the main features of Fig. 2 reappear in the diagram, Fig. 3, the minor irreg- ularities of the curve (Fig. 2) are practically extinguished, showing that they have little or no significance. A STUDY OF THE HYDE GENEALOGY 11 As a general result, we may note that after the age-period 20-25 years had been safely passed, Fig. 3, the number of deaths for each age-period remained remarkably con- stant until extreme old age was reached. In order to obtain a bird's-eye view of our subject, we should of course use the largest figures available and plot them upon a small scale. In Table 4 the ages at death are given by twenty-year periods instead of by five- year periods, as in Table 3, or by single years, as in Table 2. The diagram Fig. 4 thus gives us a more generalized view than the diagrams Figs. 3 or 2. TABLE 4. Ages at death of 2,965 persons, by twenty-year periods. Ages at death. Number. Per cent Stated 2,965 100.0 Under 20 1,044 35.2 20 and under 40 694 23.4 40 and under 60 525 17.7 60 and under 80 486 16.4 80 and over 216 7.3 30 20 20 10 FIG. 4 From Table 4 we learn that the majority (58.6 per cent) of the persons whose deaths are noted died before they were 40 years of age; and, as has already been stated, the average age at death of the whole was only 34.6 years. No less than 35.2 per cent died in childhood, before reaching the age of twenty; and only 7.3 per cent lived to be eighty years of age or older. A characteristic feature of the curve in Fig. 4 is the slight peak produced at the age-period 60-80 years. CHAPTER HI Duration of Life of Males and Females In the Hyde Statistics the sex is stated in 8,630 cases; 4,431, or 51.3 per cent, were males, and 4,199, or 48.7 per cent, females. (B. B. Rec., VIII, 153.) In 2,958 cases the sex and age at death were both stated. In these cases 1,606, or 54.3 per cent, were males, and 1,352, or 45.7 per cent, were females. The number of years lived by the 1,606 males aggregated 57,531, and the aggregate years lived by the 1,352 females amounted to 45,119 years. Average duration of life : Males, 35.8 years ; females, 33.4 years. The males, upon the average, lived longer than the females. In Table 5 the returns of males and females are analyzed by age at death by twenty- year groups, so as to show what proportion of the males died during the age-periods named, and what proportion of the females. TABLE 5. Ages at death of 1,606 males and 1,352 females, by twenty-year periods, showing the number and proportion who died at the age-periods named. Number. Per cent- Male. Female. Male. Female. Stated ' 1,606 1,352 100.0 100.0 Under 20. ... 545 492 33.9 36.4 20 and under 40.... 348 346 21.7 25.5 40 and under 60. . . . 298 227 18.5 16.8 60 and under 80. ... 290 196 18.1 14.5 80 and over 125 , 1 7.8 6.8 40* 20 !0 40* 30 ZO 10 20 tt FIG. 5 A STUDY OF THE HYDE GENEALOGY 13 In the diagram, Fig. 5, the continuous line refers to males, and the broken line to females ; and the most characteristic feature seems to be the crossing of the two lines in the middle of the diagram, and the consequent reversal of the relations of the sexes on either side of the crossing point. It will also be noticed that the peak in the curve at the 60-80 period of life observed in Fig. 4 is more pronounced in the case of the males than the females (see Fig. 5). The general result revealed by Fig. 5 seems to be that a larger proportion of females than of males died before reaching middle life and a smaller proportion afterwards. Here it may be well to observe that the diagram does not indicate that a larger number of females than of males died in the earlier periods, for this would not be true. A larger proportion, not a larger number. The curves represent percentages, not absolute numbers; and, as a matter of fact, the majority of the persons who died at each of the age-periods named were males. Take, for example, the case of those who died young : 545 males and 492 females died before they were twenty years old. Here the males are in excess. But 545 is 33.9 per cent of 1,606 (the total number of males) ; and 492 is 36.4 per cent of 1,352 (the total number of females). A larger proportion of the females died than of the males. The reverse seems to have been the case where the deaths occurred in infancy. (See Table 6.) Two hundred and thirty-one persons died in infancy, before they were one year old. Of these, 142, or no less than 61.5 per cent, were males ; and 89, or only 38.5 per cent, were females. Such a disproportion between the sexes arises more from the fact that there were more males than females to begin with (1,606 males, 1,352 females) than from any very much greater susceptibility of the males to death at this period. It is true that the susceptibility was greater, but not so great as suggested by the above figures ; for, after all, the proportion of males who died during the first year of life before reaching the age of one was comparatively small (8.8 per cent) ; and the proportion of females (6.6 per cent), although smaller, was not so very much less. Then, again, the proportion of males who died at the age of one year constituted 8.0 per cent of the males; and the proportion of females who died at the same age (7.2 per cent), although less, was not very much less. At the age of two the female deaths began to predominate. 14 THE DURATION OF LIFE TABLE 6. Ages at death of 1,606 males and 1,852 females by single years, showing the number and percentage who died in early childhood, under five, by single years. Ages at death. Number. Per cent. Male. Female. Male. Female- Stated 1,606 142 128 62 38 23 393 1,352 89 98 54 32 26 299 100.0 8.8 8.0 3.9 2.4 1.4 24.5 100.0 6.6 7.2 4.0 2.4 1.9 22.1 Under 1 1 and under 2 2 and under 3 ...... 3 and under 4 4 and under 5 * 2 FIG. 6 The excessive mortality of males was confined to the period of infancy before reach- ing the age of two. (See crossing point of the male and female lines in Fig. 6.) As the liability of males and females to death seems to have been different at differ- ent periods of life, it may be well to examine the matter in more detail. (See Table 7.) A STUDY OF THE HYDE GENEALOGY 15 TABLE 7. Ages at death of 1,606 males and 1,852 females, by five-year periods, showing the number and proportion who died during the age-periods named. 30# Agres at death. Number. Per cent- Male. Female. Male. Female. Stated 1,606 1,352 100.0 100.0 Under 5 .... 393 299 24.5 22.1 5 and under 10. ... 63 70 3.9 5.2 10 and under 15.... 36 47 2.2 3.5 15 and under 20. ... 53 76 3.3 5.6 20 and under 25 117 96 7.3 7.1 25 and under 30 86 91 5.4 6.7 30 and under 35.... 74 83 4.6 6.1 35 and under 40 71 76 4.4 5.6 40 and under 45 .... 63 70 3.9 5.2 45 and under 50. ... 82 59 5.1 4.4 50 and under 55.... 82 54 5.1 4.0 55 and under 60. ... 71 44 4.4 3.2 60 and under 65 69 55 4.3 4.1 65 and under 70. . . . 79 42 4.9 3.1 70 and under 75.... 81 54 5.1 4.0 75 and under 80. ... 61 45 3.8 3.3 80 and unde* 85 71 48 4.4 3.6 85 and under 90 37 28 2.3 2.1 90 and under 95. ... 14 11 0.9 0.8 3 4 0.2 0.3 10 15 20 25 30 35 40 IS 10 Wf 30 35 FIG. 7 In Table 7 the ages at death are shown by five-year periods instead of by twenty- year periods, as in Table 5; and it will be observed that characteristic fluctuations of the male and female curves appear in Fig. 7 that do not show on the more generalized diagram, Fig. 5. The diagram, Fig. 7, reveals characteristic likenesses and differences of the sexes in respect to death : Likenesses: The analysis by five-year periods brings out two well-marked periods of life when the danger of death was greater than at other times for both sexes. The dan- 16 THE DURATION OF LIFE ger was greatest at the very beginning of life and all through the period of early child- hood; and, in both sexes, a secondary maximum appeared at the age-period 20-25 years. This corresponds to the danger peak shown in Fig. 2 at or about the age of 23 years. Differences: Beginning at the beginning of life, the danger of death was greater for the male than the female during the period of infancy (under two years). After the age of two the female was in greater peril than the male during the whole period of childhood, at the age of puberty, and during the period of adolescence up to the age-period 20-25 years, when the proportion who died was about the same in the two sexes. There is this difference, however, that the age-period 20-25 years forms, in the male curve, quite a sharp peak, which is absent in the female curve. This is suggestive of some cause of death operating more powerfully upon males than females at this period of life. The danger of death continued to be greater for the female during the whole of the child-bearing period, up to the age-period 40-45, after which the two lines crossed. This formed the central crossing place, and then the male line came to the top. After 45 a larger proportion of males than of females died at each age-period until the age-period 90-95 was reached. Then, again, the two lines crossed; and the few win- ners in life's race who survived the age of 95 were largely females. It is true that this last reversal in the relation of the sexes rests upon very slim evidence, so far as the Hyde Statistics are concerned. Only seven persons in all, three males and four females, are noted who lived to be 95 years of age or older ; but the grad- ual approximation of the male and female lines after middle life affords additional evi- dence that a real crossing point was being approached. Other evidence also is avail- able which apparently demonstrates that such a reversal of the sexes at this period of life is a real phenomenon. The records pertaining to long life, collected by the Genealogical Eecord Office, in- clude returns from 1,470 nonagenarians and centenarians. (B. B. Rec., XVIII, p. 303.) For details, concerning the individual cases, see B. B. Rec., XVI, 70, 190, 217, 270, 382; XVII, 24, 106, 185, 265, 335, 350; XVIII, 2, 158, 281, 300. A STUDY OF THE HYDE GENEALOGY 17 TABLE 8. 1,470 Nonagenarians and Centenarians, showing their ages by five-year periods, and the percentage male and female. lumber. Per cent- Total. Male. Female. Total. Male. Female. Stated 1,470 708 762 100.0 48.2 51.8 90 and under 95.... .95 and under 100 879 302 289 448 149 111 431 153 178 100.0 100.0 100.0 51.0 49.3 38.4 49.0 50.7 61.6 "0* FIG. 8 These returns show that the majority of the persons who died between the ages of 90 and 95 were males, whereas the majority of those who lived to be 95 or older were females. Of course this, by itself, does not necessarily mean that a larger proportion of females than of males survived the age of 95 ; but we may safely infer so from the well- known fact that females constitute only a minority of the whole population born. With this point in mind, the fact that more females than males live to be, older than 95 indicates that the proportion of females living to this great age is really greater than the proportion of males, and thus affords confirmatory evidence that the third crossing point in Fig. 7, at or about the age-period 90-95, represents a real reversal in the relation of the sexes at this period of life in spite of the smallness of the figures which support it in the Hyde Statistics. . The tables heretofore presented relate to deaths; but it might be well before con- cluding this chapter upon sex to give some attention to the proportion alive at each age. 18 THE DURATION OF LIFE TABLE 9. Duration of Ufe of 1,606 males and! 1,352 females, showing the number and propor- tion who were alive at the ages mentioned. Ages. Number living:- Per cent living. Male. Female. Male. Female. Living at birth 1,606 1,352 100.0 100.0 Living at 5 years.. 1,213 1,053 75.5 77.9 Living at 10 years.. 1,150 983 71.6 72.7 Living at 15 years.. 1,114 936 69.4 69.2 Living at 20 years.. 1,061 860 66.1 63.6 Living at 25 years.. 944 764 58.8 56.5 Living at 30 years. . 858 673 53.4 49.8 Living at 35 years . . 784 590 48.8 43.7 Living at 40 years.. 713 514 44.4 38.1 Living at 45 years.. 650 , 444 40.5 32.9 Living at 50 years . . 568 385 35.4 28.5 Living at 55 years.. 486 331 30.3 24.5 Living at 60 years.. 415 287 25.9 21.3 Living at 65 years.. 346 232 21.6 17.2 Living at 70 years.. 267 190 16.7 14.1 Living at 75 years.. 186 136 11.6 10.1 Living at 80 years.. 125 91 7.8 6.8 Living at 85 years. . 54 43 3.4 3.2 Living at 90 years. . 17 15 1.1 1.1 Living at 95 years.. 3 4 0.2 0.3 -GO* Jf f 10 It tO 15 * M 40 .r 48.1 per cent (nearly one-half of the whole), had parents who lived to be eighty or >lder. Fathers (not mothers), 337; mothers (not fathers), 246; both parents, 184; total, 767 cases. The importance of heredity as a factor in producing longevity is so obvious as to justify a more detailed examination of the cases referred to. 30 THE DURATION OF LIFE TABLE 15. Analysis of 1,594 cases in which the age at death and the fathers' and mothers' ages at death are all stated, showing the number of persons having fathers and mothers who died at the age periods named, and the average duration of life of the persons and of their fathers and mothers. Age of parents at death- Number of cases. Aggregate years lived Average years lived By persons. By fathers. By mothers- By persons. By fathers. By mothers. Fathers- Mothers. Stal 60 60 60 ed. ...... 1,594 128 120 74 64,717 4,202 4,010 2,689 113,057 6,518 6,140 3,726 105,214 5,688 8,526 6,421 40.6 32.8 33.4 36.3 70.9 50.9 51.2 50.3 66.0 44.4 71.0 86.8 60 60-80 80+ 60-80 60-80 60-80 60 60-80 80+ 251 328 172 8,984 12,477 7,737 16,947 23,467 12,140 10,814 23,278 14,700 35.8 38.0 45.0 67.5 71.5 70.6 43.1 71.0 85.5 80+ 80+ 80+ 60 60-80 80+ 131 206 184 5,541 9,384 9,693 11,154 17,391 15,574 5,912 14,423 15,452 42.3 45.5 52.7 85.1 84.4 84.6 45.1 70.0 84.0 NUMBER OF CASES CONDENSED TABLE AVERAGE DURATION OF LIFE A6 AT OfATH 6O-80 80-+- 6O A&f AT DEATH 60-8O 8O-I- > a Jo 131 206 184 fr N 0) *A ^8 251 328 172 $0 K(l 128 120 74 51 42.3 45.5 52.7 + o< ODjg h K 35.8 38.0 45.0 08 ** 32.8 33.4 363 | A STUDY OF THE HYDE GENEALOGY 31 Plan for a three dimensional diagram illustrating Table 15 A'gf AT Of ATM 310 BIO Tto e)o en K . i nli 11.1.11111 ml Huh MI In. 71. i ill, ti. I MM I.. n ft . l . l 3 , 4K> 1 1 1 1 I 1 1 1 1 1 1 1 1 1 1 1 II 1 1 I 1 II 1 1 1 1 1 1 H 1 1 1 1 1 1 1 1 1 1 1 H K : 41 Si 1J *- -8 -S 3S 3J8 i'x iii i- iii ar r "Tr TT ar ' ,r "^r*^; Mornt*!s Ast AT FIG. 15 FIG. 15a Photograph of a stereogram or three dimensional diagram illustrating Table 15 NOTE. Three dimensional diagram: The horizontal components indicate the average dura- tion of life of the fathers and mothers, and the vertical component the average duration of life of the offspring. For example, take the lowest dot on the left-hand side of the diagram, Fig. 15o. The ver- tical height of this dot above the plane of the paper is 32.8, upon the scale adopted, and this means that the persons in this group lived on the average 32.8 years. The horizontal com- ponents of this dot are shown in Fig. 15 and indicate upon the same scale the average ages at death of the fathers and mothers ; fathers 50.9 years, mothers 44.4 years. Or, again, take the highest dot on the right-hand side of the diagram 15a. The vertical height is 52.7, indicating that the persons represented in this group lived on the average 52.7 years. By reference to Fig. 15 the horizontal components are seen to be fathers 84.6 years, mothers 84.0 years. In Fig. 15a this point forms quite an elevated peak, indicating that the offspring of parents, both of whom lived to be old, were very much longer lived on the average than the offspring of parents who died at earlier periods of life. A STUDY OF THE HYDE GENEALOGY 33 The 1,594 persons referred to in Table 15 lived on the average 40.6 years, their fathers 70.9 years; and their mothers 66.0 years. In order to secure large enough numbers upon which to calculate averages, the 1,594 persons have been divided into only nine groups, arranged according to the ages at death of the parents. For example, the condensed table is divided into nine compartments or boxes. It will be observed that the number 128 has been placed in one of these compartments ; and, by reference to the margin, it will be found that this compartment is devoted to cases having fathers and mothers who did not live to be sixty years of age. The figures mean that there were 128 persons who had fathers and mothers who both died under sixty. By reference to the detailed part of the table it will be found that the aggregate years lived by these 128 persons amounted to 4,202 years; and dividing this by 128, we reach the conclusion that the average age at death in these cases was 32.8 years. In the condensed table this result (32.8) has been placed in the proper compartment relating to fathers and mothers who both died under sixty. (See condensed table for average duration of life.) Contrast with this the result obtained with the 184 persons whose parents, both of them, lived to be eighty years of age or older. The average duration of life in these cases was 52.7 years. Where both parents died under sixty, the persons on the average lived only 32.8 years (128 cases). Where both parents lived to be eighty or older, the persons on the average lived 52.7 years (184 cases). Where the parents died at the intermediate age-periods noted, the average duration of life of the persons was intermediate. The results indicate that the duration of life of the person was directly correlated with the duration of life of both parents, and that the influence of the father was greater than that of the mother. Where the father lived to be eighty or older and the mother died under sixty, the persons lived on the average 42.3 years (131 cases). Where the mothers lived to be eighty or older and the fathers died under sixty, the persons lived on the average 36.3 years (74 cases). The persons whose parents both lived to be old were the longest lived on the average. CHAPTER VI Duration of Life of Fathers and Mothers Compared with the Number of Children Born to Them FATHERS In several 'thousand cases (5,024) the age of the father at death was known (see Table 10) ; but the number of fathers was not known because many of the persons were the children of the same father. A special examination of the card catalogue of the Hyde Genealogy has been made to eliminate duplicate fathers and thus ascertain the actual number of fathers involved. Each card contained the name of a person, the name of his father, the name of his mother, and the total number of children born to these parents. It was found, however, that duplication of the fathers could not be en- tirely avoided, because some of the fathers had married more than once and had families by each marriage. This difficulty was overcome by noting the number of fathers and the number of mar- riages, and calculating the average number of children per father and per marriage. There were -only 795 fathers in all, but they had contracted 843 marriages resulting in offspring. (See Table 16.) TABLE 16. 8^3 fertile marriages of 795 males, showing the number of children produced from fathers who died at the age-periods named, and the average number of children per father and per marriage. Father's age at death. Number of fathers. Number of marriages. Number of children. Number of children- Per father. Per marriage- Stated 795 69 207 343 176 843 71 213 369 190 5,041 192 1,243 2,356 1,250 6.3 2.8 6.0 6.9 7.1 6.0 2.7 5.8 6.4 6.6 Under 40 40 and under 60 60 and under 80 34 A STUDY OF THE HYDE GENEALOGY 35 a is n 5? FIG. 16 Table 16 relates to 843 fertile marriages of 795 males; 5,041 children were pro- duced. Average : 6.3 children per father, or 6.0 children per marriage. Aggregate years lived by the 795 fathers, 51,668 years. Average duration of life of fathers, 65.0 years. The fathers who died under forty years of age had on the average 2.8 children apiece (69 cases). The fathers who lived to be eighty or older had on the average 7.1 children apiece (176 cases). The fathers who died at the intermediate age-periods had on the average families of intermediate size. There was thus a direct correlation between the duration of life of the fathers and the number of children born to them. Here a complication arises, due to the fact that some of the fathers married more than once and had children by each marriage. Fig. 16 shows that the average number of children born to the older fathers was somewhat increased from this cause, the curve for the average number of children per father being sensibly higher than that for the aver- age number of children per marriage. The latter curve, however, is of the same general character as the former, showing a distinct rise in the average number of children in the older cases. Fathers died under forty. Average, 2.7 chil- dren per marriage (71 cases). Fathers died between forty and sixty. Aver- age, 5.8 children per marriage (213 cases). Fathers died between sixty and eighty. Average, 6.4 children per marriage (369 cases). Fathers lived to be eighty or older. Average, 36 THE DURATION OF LIFE The effect of considering the children of second and third marriages as distinct families reduced the average size of the family in the case of fathers who married more than once; and yet the curve for the average number of children per marriage continu- ally goes up, showing that, quite independently of this complicating cause, there was a distinct correlation between the size of the family and the duration of life of the father. The longer-lived fathers, on the average, had the largest families. MOTHERS The correlation between the duration of life of the mother and the number of chil- dren born to her is still more obvious than in the case of the fathers, because the com- plicating condition referred to above is absent. It is safe to say that no females had children after they reached middle life, the con- clusion of the child-bearing period (say forty-five years) ; so that in the cases of women who married again late in life there would be no offspring by these later marriages. We cannot suppose, for example, that a mother who died at fifty would have had more children had she lived to be sixty, seventy, or eighty years of age or older, even though she might have married half a dozen times, and young husbands, too. Table 17 relates to 671 fertile marriages of 671 females; 4,022 children were pro- duced. Average, 6.0 children per marriage. Aggregate years lived by the 671 mothers, 40,805 years. Average duration of life of mothers, 60.8 years. The Hyde Genealogy gives us some account of the second marriages of males, espe- cially where they bore the name of Hyde; but few cases of second fertile marriages of females are recorded. In Table 17, therefore, we are unable to distinguish between the number of children per mother and the number of children per marriage. A STUDY OF THE HYDE GENEALOGY 37 TABLE 17. 671 fertile marriages of 671 females, showing the number of children produced from mothers who died at the age-periods named, and the average number of children per marriage. Number of Mother's age at death- Number of Number of children per marriages. children. marriage- Stated 671 4,022 6.0 Under 40 130 439 3.4 40 and under 60. ... 164 1,018 6.2 60 and under 80 244 1,612 6.6 133 953 7.2 6 40 60 80 FIG. 17 The mothers who died under forty had on the average 3.4 children per marriage (130 cases) . The mothers who died between forty and sixty had on the average 6.2 children per marriage (164 cases). The mothers who died between sixty and eighty had on the average 6.6 children per marriage (244 cases). The mothers who lived to be eighty or older had on the average 7.2 children per marriage (133 cases). CHAPTER VII Duration of Life of Persons Compared with the Ages of the Parents at Marriage FATHERS There were 2,605 cases where the person's age at death and the father's age at mar- riage were both stated. These 2,605 persons lived 91,126 years. Average duration of life of persons, 35.0 years. TABLE 18. 2,605 cases where the person's age at death and the father's age at marriage were both stated, showing the number having fathers who married at the age-periods named, the aggregate years lived by the persons, and their average ages at death. Father's age at marriage- Number of persons. Aggregate years lived. Average years lived- Stated 2,605 91,126 35.0 Under 25 1,072 39,299 36.6 25 and under 30.... 963 33,049 34.3 30 and under 35 346 12,322 35.6 35 and under 40. . . . 114 3,481 30.5 40 and over 110 2,975 27.0 FIG. 18 26- Fig. 18 shows that the persons whose fathers married early in life lived on the average to be older than the others. The persons whose fathers married before they were twenty-five lived on the average 36.6 years (1,072 cases). The persons whose fathers married after they were forty years of age lived on the average 27.0 years (110 cases). The persons whose fathers married at the intermediate age-periods lived on the average to intermediate ages. 38 A STUDY OF THE HYDE GENEALOGY 39 There was thus a correlation of an inverse kind between the age at death of the person and the age at marriage of his father. The average duration of life was greater, as the father was younger at marriage. There were only 23 cases where the father married before he was twenty ; but these persons, on the average, were the longest lived of all. Average duration of life, 45.8 years. Although there is an irregularity in the curve shown in Fig. 18, where the fathers married between thirty and thirty-five years of age, the general trend of the curve is dis- tinctly downwards with increase of age at marriage; and the indications are very clear that in the case of the fathers early marriage was a condition favorable to the produc- tion of long-lived children. MOTHERS There were 2,309 cases where the person's age at death and the mother's age at marriage were both stated. Average duration of life of the persons, 34.6 years. TABLE 19. 2,809 cases where the person's age at death and the mother's age at marriage were both stated, showing the number having mothers who married at the age-periods named, the aggregate years lived by the persons, and their average ages at death. Mother's age at marriage. Number of persons. Aggregate years lived- Average years lived. Stated 2,309 79,976 34.6 Under 20 712 25,837 36.3 20 and under 25 1,070 37,185 34.7 25 and under 30 421 13,635 32.4 30 and over 106 3,315 31.3 30 30 20 20 25 .ZS 30 FIG. 19 20 In Fig. 19 we have proof of a correlation, of an inverse kind, between the duration of life of the person and the age at marriage of the mother. The persons whose mothers married before they were twenty lived on the average 36.3 years (712 cases). The persons whose mothers married after they were thirty years of age lived on the average 31.3 years (106 cases). The persons whose mothers married at the intermediate age-periods lived on the average to intermediate ages. 40 THE DURATION OF LIFE The curve in Fig. 19 forms almost a straight line, sloping downwards from left to right, and indicates very clearly that the average duration of life of the persons was greater, as the mother was younger at marriage. We have too few cases to decide the lower limit of mothers' age at marriage consistent with the production of long-lived offspring; hut the persons whose mothers married at sixteen and seventeen were the longest lived of any of the groups examined. The persons whose mothers married at seven- teen lived on the average 39.3 years (141 cases). The persons whose mothers married at six- teen lived on the average 40.8 years (46 cases). In Tables 18 and 19 we have undoubted proof that the persons whose parents mar- ried young were, on the average, longer lived than the persons whose parents married later in life. It thus appears that early marriage was a condition distinctly favorable to longevity in the offspring. CHAPTER VIII Duration of Life of Persons Compared with the Ages of the Parents When the Persons Were Born FATHERS In 2,756 cases the person's age at death and the age of the father when the person was born were both stated. Average duration of life of the persons, 35.4 years. TABLE 20. 2,756 cases where the person's age at death and the age of the father when the person was born were both stated, showing the number born to fathers of the ages speci- fied, the aggregate years lived by the persons, and their average ages at death. Father's age when person was born. Number of persons. Aggregate years lived. Average years lived. Stated 2,765 97,793 35.4 Under 25. 180 7,132 39.6 25 and under 30. 601 23,000 38.3 30 and under 35. 642 23,640 36.8 35 and under 40. 572 19,747 34.5 40 and under 45. 409 12,894 31.5 45 and under 50. 235 7,383 31.4 50 and over 128 3,997 31.7 s FIG. 20 The youngest fathers, on the average, had the longest-lived offspring. The persons whose fathers were under twenty- five years of age when the persons were born lived on the average 39.6 years. The persons whose fathers were between forty and forty-five years of age when the persons were born lived on the average 31.5 yean. The persons whose fathers were of inter- mediate ages when the persons were born lived on the average to intermediate ages. 41 42 THE DURATION OF LIFE There was thus an apparent correlation, of an inverse kind, between the duration of life of the person and the age of the father when the person was born, down to fathers of 40-45 years of age; and there the correlation stops. (See Fig. 20.) The line is prac- tically horizontal beyond the age-period 40-45. The evidences of correlation are so marked upon the whole that we would naturally expect the curve to go down steadily to the very end, and thus indicate that the offspring of the oldest fathers were on the average the shortest lived ; but the curve does not do so. The general trend of the curve in Fig. 20 is distinctly downwards from left to right ; and the horizontality of the last part of the curve is suggestive of some cause operating to interfere with the normal descent of the curve in the case of the older fathers. Here an explanation suggests itself from a consideration of the fact that children are the product of two parents, not one alone. It is very obvious, for example, that where the fathers were over fifty years of age when their children were born the mothers were certainly much younger. The fathers who were over forty years of age when their children were born had on the average wives who were younger than themselves. (See B. B. Eec., Vol. XX, pp. 2-36.) MOTHERS In 2,386 cases the person's age at death and the age of the mother when the person was born were both stated. Average duration of life of the persons, 34.9 years. TABLE 21. 2,386 cases where the person's age at death and the age of the mother when the person was born were both stated, showing the number born to mothers of the ages speci- fied, the aggregate years lived by the persons, and their average ages at death. Mother's age when person was born. Number of persons. Aggregate years lived. Average years lived. Stated 2,386 83,116 34.9 Under 25 25 and under 30 30 and under 35 35 and under 40. ... 525 603 561 422 275 20,322 22,073 18,829 14,161 7,731 38.7 36.6 33.6 33.6 28.1 IS 3S Fro. 21 20 \ A STUDY OP THE HYDE GENEALOGY 43 The youngest mothers, on the average, had the longest-lived offspring. The persons whose mothers were under twenty-five years of age when the persons were born lived on the average 38.7 years. The persons whose mothers were forty years of age or older when the persons were born lived on the average 28.1 years. The persons whose mothers were of inter- mediate ages when the persons were born lived on the average to intermediate ages. There was thus a distinct correlation, of an inverse kind, between the duration of life of the person and the age of the mother when the person was born ; and in this case the curve slopes downwards to the very end, indicating not only that the youngest mothers had the longest-lived offspring, but that the oldest mothers had the shortest-lived off- spring. (See Pig. 21.) There is, however, in this case a curious irregularity of the curve where the mothers were between thirty-five and forty years of age when their children were born, resembling somewhat a similar irregularity in Fig. 18 referring to the father's age at marriage. Whether there is any connection between these irregularities is not clear; but in both cases the absolute numbers are sufficiently large to give significance to the averages. These irregularities therefore mean something, but the interpretation is not obvious. The youngest mothers, on the average, had the longest-lived children; but we have too few cases to enable us to decide the extreme lower limit of age consistent with the production of long-lived offspring. The following cases, however, are significant: The persons whose mothers were only eighteen years of age when the persons were born lived on the average 44.1 years (19 cases). The persons whose mothers were only seven- teen years of age when the persons were born lived on the average 44.6 years (13 cases). Thus both in the case of the fathers and mothers it becomes obvious that youth is a condition favorable to the production of long-lived offspring. The younger fathers and mothers had on the average the longest-lived children. CHAPTER IX Duration of Life of Persons Compared with the Number of Years After the Marriage of the Parents the Persons Were Born There were 2,757 cases in which the age at death of the person was known, and also the number of years after the marriage of the parents the person was born. Average duration of life of these persons, 34.4 years. TABLE 22. 2,757 cases where the person's age at death and the number of years after the mar- riage of the parents the person was born were both stated, showing the number of persons born the specified years after the marriage of their parents, the aggregate years lived by the persons, and their average ages at death. Number of years after marriage of parents person was born. Number of persons. Aggregate years lived. Average years lived- Stated 2,757 94,749 34.4 years 62 2,168 35.0 304 10,588 34.8 Less than 2. .. 366 12,756 34.9 2 and less than 4... 413 14,221 34.4 4 and less than 6. . . 352 13,197 37.9 6 and less than 8. . . 332 12,083 36.4 8 and less than 10... 260 8,951 34.4 10 and less than 12... 218 6,973 32.0 12 and less than 14. . . 196 6,801 32.0 14 and less than 16... 170 5,598 32.9 16 and less than 18. . . 144 4,765 33.1 18 and less than 20. .. 119 4,114 34.6 20 and more 187 5,290 28.3 Less than 10. . . 1,723 VJW 61,208 35.5 10 and less than 20. .. 847 28,251 33.3 187 5,290 28.3 44 ft 4 6 6 8 ! Pro. 22 1} 14 14 16 A STUDY OF THE HYDE GENEALOGY 45 The persons who were born less than ten years after the marriage of their parents were on the average longer lived than those born later in married life. (See first dia- gram in Fig. 22.) Persons born less than ten years after the marriage of their parents. Average duration of life, 35.5 years (1,723 cases). Persons born between ten and twenty years after the marriage of their parents. Average duration of life, 33.3 years (847 cases). Persons born twenty or more years after the marriage of their parents. Average duration of life, 28.3 years (187 cases). Table 22 throws some light upon the duration of life of first-born children as com- pared with others. (See second diagram in Fig. 22.) Persons born the same year as the marriage of their parents or the year after may safely be assumed to be first-born children. Average duration of life in these cases, 34.9 years (366 cases). Thus the persons who were undoubtedly first-born children were fully equal in vitality to the average of the whole, or rather they lived upon the average as long in fact, a little longer. In Fig. 22 the horizontal line shows the average duration of life of the whole 2,757 persons (34.4 years). Persons born less than two years after the marriage of their parents. Average duration of life, 34.9 years. This is slightly above the average of the whole (366 cases). Persons born two and three years after the marriage of their parents. Average duration of life, 34.4 years, just the average of the whole (413 cases). Persons born four and five years after the marriage of their parents. Average duration of life, 37.9 years. This is considerably above the average of the whole ; in fact, the highest average of any of the groups shown in Table 22 (362 cases). Persons born six and seven years after the marriage of their parents. Average duration of life, 36.4 years. This is above the average of the whole, but not quite up to the average of the last group named (332 cases) . Persons born eight and nine years after the marriage of their parents. Average duration of life, 34.4 years, just the average of the whole (260 cases). Persons born less than ten years after the marriage of their parents, the summation of the above groups, lived on the average 35.5 years (1,723 cases). While this is above the average of the whole, only two of the above groups show averages markedly greater than the average of the whole. 46 THE DURATION OF LIFE Why persons born four and five years after the marriage of their parents should on the average have been longer lived than the persons born earlier or later in married life is a mystery that needs explanation. The children, too, born six and seven years after the marriage of their parents were also longer lived than the children born later in married life. There is another unexplained feature about Fig. 22. It is very obvious that upon the whole the children born after ten years of married life were shorter lived than those born less than ten years after the marriage of their parents ; but why the curve should here go up instead of down, so that the children born eighteen and nineteen years after the marriage of their parents should prove to be as long lived as the average of the whole, is a mystery indeed. There were 119 cases of persons born eighteen and nineteen years after the marriage of their parents, and this seems a sufficiently large number to make the average mean something. Average duration of life, 34.6 years, slightly above the average of the whole. There can be no question here of old fathers with young wives. These persons were born eighteen and nineteen years after the marriage of their parents, from which it is obvious that the mothers as well as the fathers were eighteen and nineteen years older than when they were married. The mothers certainly were not young. The regularity of the oscillation of the curve (second diagram in Fig. 22) strongly confirms the belief that we are here dealing with a real oscillation, whatever it means, and not merely with an accidental irregularity due to basing averages upon too small num- bers. An examination of Table 22 will show that the absolute numbers involved in the determination of each point in the diagram are sufficiently large to make the averages mean something. CHAPTER X Duration of Life of Persons Compared with Number in Family (Siblings) SIBLINGS It is unfortunate that we have no word in the English language, excepting the very awkward and recently invented term "sibling," to express the relationship of the children of the same parents to one another independently of sex. Suppose a person had two brothers and three sisters, then he had five siblings, and the total number of siblings in the family was six. The person was one of a family of six children. In order to avoid the inconvenience of employing a little-understood term, we shall in the present investigation use the expression "number in family" for "number of sib- lings in family"; the "family" consisting of the person and his brothers and sisters. A STUDY OF THE HYDE GENEALOGY 47 There were 2,964 cases in which the person's age at death and the number in family were both stated. (See Table 23.) TABLE 23. Duration of life of persons compared with number in family (siblings). Number In family (siblings). Total persons. Persons' age at death- -20 20-40 40-60 60-80 80+ Stated 2,964 41 85 126 313 584 694 683 396 168 1,044 24 36 60 113 207 229 224 133 78 693 9 21 30 80 143 175 152 84 29 525 2 16 18 60 107 123 122 73 22 486 4 8 12 45 93 117 119 71 29 216 2 4 6 15 34 50 66 35 10 1 2 Less than 3 5 and less than 7 9 and less than 11 11 and less than 13*. 13 or more Percentages. Stated 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 35.2 58.5 42.4 35.5 36.1 47.6 33.0 32.8 33.6 46.4 23.4 -22.0 24.7 24.5 25.5 23.8 25.2 22.2 21.2 17.3 17.7 4.9 18.8 18.3 19.2 14.3 17.7 17.9 18.4 13.1 16.4 9.7 9.4 15.9 14.4 9.5 16.9 17.4 17.9 17.3 7.3 4.9 4.7 5.8 4.8 4.8 7.2 9.7 8.9 5.9 1 2 3 and less than 5 11 and less than 13 7 9 FIG. 23 48 TEE DURATION OF LIFE Fig. 23 contains a series of seven diagrams. The first refers to persons belonging to small families containing only one or two children. The last relates to persons belong- ing to very large families containing thirteen or more children; and the intermediate diagrams to persons belonging to families of intermediate size. In each diagram the first dot expresses the percentage of persons who died young, under twenty ; the last dot the percentage of persons who lived to be eighty years of age or older; and the intermediate dots the percentage of persons who died at the inter- mediate age-periods named in Table 23. Both in the case of very small families (containing only one or two children) and in the case of very large families (containing thirteen or more children), the proportion of persons who died young was very large and the proportion who lived to be old small. (See Fig. 23.) The majority of the persons who were the only children of their parents died young (58.5 per cent) , but there were only 41 cases. There were 126 persons belonging to small families containing only one or two children. No less than 47.6 per cent of these persons died In childhood under twenty, and only 4.8 per cent lived to be eighty or older. There were 168 persons belonging to very large families containing thirteen or more children. No less than 46.4 per cent of these persons died in childhood under twenty, and only 5.9 per cent lived to be eighty or older. The proportion who died young was least, and the proportion who lived to be old greatest, where the persons belonged to families containing nine and ten children. There were 683 persons belonging to families containing nine and ten children. 32.8 per cent of these persons died young under twenty; and this is the smallest pro- portion dying in childhood of any of the groups shown In Table 23. 9.7 per cent of these persons lived to be eighty or older; and this is the largest proportion living to old age in any of the groups shown in Table 23. The proportion who lived to be old increased with the size of the family up to fam- ilies containing nine and ten children, and fell again in the case of the larger families. CHAPTER XI Conclusion ; Summary of the Chief Points Touched Upon in the Preceding Chapters The 8,797 persons dealt with in the Hyde Statistics were born chiefly in the eigh- teenth century and the early part of the nineteenth century. Only a small proportion appeared after 1825. "Most of these persons were either living in 1864, when the Hyde Genealogy was published, or their ages at death could not be ascertained ; so that the full duration of life is only known in about one-third of the cases. There were 2,965 persons whose ages at death were known. Average duration of life of these persons, 34.6 years. 35.2 per cent of these persons died before they were twenty years of age, and 7.3 per cent lived to be eighty or older. There were critical periods in the lives of these persons when the danger of death was greater than at other times. The danger was greatest in infancy, especially during the first year of life; and a second danger period appeared during adolescence, reaching its maximum danger point at or about the age of twenty-three years. Both the males and females show an increase in the number and proportion of deaths occurring during adolescence, and in both cases the maximum appears at the age-period 20-25 years. But the death curve for males exhibits quite a sharp peak at this point which is absent from the female curve, suggesting some cause of death at this period of life affecting males more powerfully than females. Unfortunately the statistics do not give us any information concerning causes of death. There were 1,606 males whose ages at death were known. Average duration of life of these males, 35.8 years. There were 1,352 females whose ages at death were known. Average duration of life of these females, 33.4 years. The males upon the average lived longer than the females, and yet more females than males lived to extreme old age. A larger proportion of females than of males died in childhood before reaching the age of twenty years, in spite of an excessive mortality of males during the period of infancy. Thus a larger proportion of males than of females lived to adult life. 49 50 TEE DURATION OF LIFE A larger proportion of females than of males died between the ages of twenty and forty. Females, of course, were exempt from military service; but they were exposed, during the child-bearing period, to dangers which did not affect the male population at all. A larger proportion of females than of males lived to be ninety-five years of age or older. The few winners in life's race were largely females. There were 795 fathers whose ages at death were known. Average duration of life of these fathers, 65.0 years. There were 671 mothers whose ages at death were known. Average duration of life of these mothers, 60.8 years. The fathers, on the average, lived longer than the mothers; and yet more mothers than fathers lived to extreme old age. Heredity was deeply involved in the production of longevity. Where neither parent lived to be eighty, about 5 per cent of the known offspring lived to be eighty or older (5.3 per cent). Where one parent, but not the other, lived to be eighty or older, about 10 per cent of the known offspring lived to be eighty or older (9.8 per cent). Where both parents lived to be eighty or older, about 20 per cent of the known off- spring lived to be eighty or older (20.6 per cent). The influence of the father seemed to be somewhat greater than that of the mother. Where the father (but not the mother) lived to be eighty or older, 11.3 per cent of the known offspring lived to be eighty or older. Where the mother (but not the father) lived to be eighty or older, 7.7 per cent of the known offspring lived to be eighty or older. Where the father died over eighty and the mother died under sixty, the children pro- duced by these parents lived on the average 42.3 years. Where the mother died over eighty and the father died under sixty, the children pro- duced by these parents lived on the average 36.3 years. The average duration of life of offspring was greatest where the parents were long lived, least where the parents were short lived, and intermediate where the parents died at intermediate age-periods. A STUDY OF THE HYDE GENEALOGY 51 There was thus a direct correlation between the duration of life of parents and the duration of life of their offspring. The average duration of life of the offspring was greater as the parents were younger at marriage. The average duration of life of the offspring was greater as the parents were younger when the children were born. The children born during the first ten years of married life were, on the average, longer lived than those born later. Thus virility on the part of the parents had something to do with the duration of life of the offspring. Children born between four and eight years after the marriage of their parents were, upon the average, longer lived than those born earlier or later in married life. First-born children were fully up to the average of the whole in vitality and lived as long. The majority of the persons who were the only children of their parents died young; but there were only 41 cases. Both in very small families (containing only one or two children) and in very large families (containing thirteen or more children) the proportion who died young was very large and the proportion who lived to be old small. The proportion who lived to be old increased with the size of the family up to fam- ilies containing nine and ten children, and fell again in the case of larger families. Parents who died early in life had comparatively small families, because on the average they did not live long enough to develop their full reproductive powers. Fathers who died under forty had on the average 2.8 children apiece. Mothers who died under forty had on the average 3.4 children apiece. The average number of children produced increased with the duration of life of the parents, even in cases where the parents had passed the reproductive period of life. Where the fathers died between forty and sixty, 6.0 children per father were produced. Where the fathers died between sixty and eighty, 6.9 children per father were produced. Where the fathers lived to be eighty or older, 7.1 children per father were produced. Where the mothers died between forty and sixty, 6.2 children per mother were produced. Where the mothers died between sixty and eighty, 6.6 children per mother were produced. Where the mothers lived to be eighty or older, 7.2 children per mother were produced. 52 THE DURATION OF LIFE There was thus a direct correlation between the duration of life of the parents and the number of offspring born to them. The longer-lived parents were the most fertile. Only a small proportion of the population born lived to be old ; but a large propor- tion came from long-lived parents. There were 1,606 males whose ages at death were known. 7.8 per cent of these males lived to be eighty or older. There were 1,352 females whose ages at death were known. 6.8 per cent of these females lived to be eighty or older. There were 5,024 persons whose fathers' ages at death were known. 24.9 per cent of these persons had fathers who lived to be eighty or older. There were 3,958 persons whose mothers' ages at death were known. 23.0 per cent of these persons had mothers who lived to be eighty or older. The males of one generation became, of course, the fathers of the next generation. At least some of them did ; but some did not marry and others had no offspring. If, then, the above percentages should hold good for two successive generations, it is obvious that less than 7.8 per cent of the males of one generation would become the fathers of 24.9 per cent of the whole population born in the next generation, and that less than 6.8 per cent of the females would become the mothers of 23.0 per cent of all the children born in the next generation. There were 1,594 cases where the ages at death of the persons and the ages at death of both parents were known. 8.7 per cent of these persons lived to be eighty or older. 32.7 per cent of these persons had fathers who lived to be eighty or older. 27.0 per cent of these persons had mothers who lived to be eighty or older. 48.1 per cent of these persons had parents one or both of whom lived to be eighty or older. If these proportions should hold true for two successive generations, then less than 8.7 per cent of the people of one generation would become the fathers or mothers of nearly half of the population born in the next generation (48.1 per cent). The above illustrations indicate that long-lived people, though few in number, may profoundly affect the composition of the whole population born in the next generation. The statistics indicate that a tendency to longevity is an inheritable characteristic, capable of being handed down from parents to children. A STUDY OF THE HYDE GENEALOGY 53 What is really inherited is probably a tough, wiry constitution that enables the fortunate possessor to survive the multitudinous ills that flesh is heir to and live on to the extreme limit of human life. From this point of view the attainment of old age is extremely significant. The last survivors of a whole generation are people who have by the very fact of their surviving to old age proved themselves to be resistant to disease. They have been exposed to the diseases and accidents that have cut off the vast majority of their fellows before their prime, but have not succumbed. The statistics indicate that this disease-resistant quality is inherited by offspring; and, through the superior fecundity of the long lived, distributed very generally through- out the population. Here we have evidence of the existence of a natural process at work among human beings tending to improve the vigor and vitality of succeeding generations. A. G. B. APPENDIX A The Hyde Material in the Beinn Bhreagh Recorder Two series of detailed tables have been preserved in the Beinn Bhreagh Recorder: 1. "Statistics from the Hyde Genealogy," compiled at various times by Miss Mabel B. McCurdy, Miss Gretchen Schmidt, Miss Catherine D. Mackenzie, Mr. John Small- wood, Mr. Melville Bell Grosvenor, and Mr. A. W. Clime. All the tables were prepared under the personal direction and supervision of Alexander Graham Bell, who did his best to verify the results. 2. "Studies of the Hyde Statistics," by Alexander Graham Bell. The following is a list of the tables referred to, together with a description showing the nature of the material contained in them : Statistics. Studies. Table. Vol. Page. Table. Vol. Page. 1 VIII 133 VIII 287 2 VIII 153 A IX 98 3 VIII 187 B IX 172 4 VIII 319 C IX 177 5 VIII 387 D IX 182 6 IX 20 E IX 232 7 IX 143 F IX 265 8 IX 244 G IX 280 9 X 11 H IX 315 10 X 318 I IX 327 11 XI 107 J X 55 12 XII 28a K XI 46 13 XII 239 L XI 361 14 XVII 42 M XII 344 15 XVIII 275 N XII 415 16 XVIII 334 O XX 94 17 XX 2 P XX 118 18 XX 22 Q XX 153 19 XX 25 B XX 159 20 XX 29 S XX 176 21 XX 134 T XX 180 U XX 187 V XX 192 W XX 195 X XX 201 A STUDY OF THE HYDE GENEALOOY 55 A Description of the Detailed Tables Preserved in the Beinn Bhreagh Recorder, Showing the Nature of the Material Contained in Them 1. STATISTICS FROM THE HYDE GENEALOGY TABLE 1. Period of birth. B. B. Rec., VIII, 133-138. TABLE 2. Sex: By period of birth. B. B. Rec., VIII, 153-158. TABLE 3. Age at death: By sex and period of birth. B. B. Rec., VIII, 187-219. TABLE 4. Duration of life: By sex and period of birth. (This includes the "latest age known to be living.") B. B. Rec., VIII, 319-353. TABLE 5. Number in family : By sex, period of birth, and age at death. B. B. Rec., VIII, 387-^15. TABLE 6. Age at death: By age of father and mother at death and by sex. B. B. Rec., IX, 20-28. TABLE 7. Age at death: By father's age when person was born. B. B. Rec., IX, 143-171. TABLE 8. Age at death : By mother's age when person was born. B. B. Rec., IX, 244-264. TABLE 9. Age at death : By father's age at marriage. B. B. Rec., X, 11-35. TABLE 10. Age at death : By mother's age at marriage. B. B. Rec., X, 318-334. TABLE 11. Analysis of 2,287 cases in which the age at death and the father's age at death are both stated, with average ages at death. B. B. Rec., XI, 107-124. TABLE 12. Analysis of 1,806 cases in which the age at death and the mother's age at death are both stated, with average ages at death. B. B. Rec., XII, 28o-42. TABLE 13. Analysis of 1,594 cases in which the age at death and the fathers' and mothers' ages at death are all stated, with average ages at death of the persons and of their fathers and mothers. B. B. Rec., XII, 239-269. TABLE 14. Analysis of 1,594 cases in which the ages at death of the persons and of their fathers and mothers are all stated: Arranged by the persons' ages at death. B. B. Rec., XVII, 42-75. TABLE 15. Analysis of 5,024 cases in which the father's age at death is known, with aggre- gate years lived by the fathers and average age at death of the fathers. B. B. Rec., XVIII, 275-280. (Incorrectly given in the Recorder as Table 14.) TABLE 16. Analysis of 3,958 cases in which the mother's age at death is known, with aggre- gate years lived by the mothers and average age at death of the mothers. B. B. Rec., XVIII, 334-340. TABLE 17. Age at death : And fathers' and mothers' ages at marriage. B. B. Rec., XX, 2-21. TABLE 18. Number of persons having fathers and mothers who married at the ages named. B. B. Rec., XX, 22-24. TABLE 19. Analysis of 2,169 cases, showing the number of persons having fathers and mothers who married at the ages stated and the aggregate number of years lived "by the persons. B. B. Rec., XX, 25-28. TABLE 20. Number of persons having fathers and mothers who married at the ages named, arranged according to the difference of age of the parents; the aggregate years lived by the persons and their average ages at death. B. B. Rec., XX, 29-36. 56 TEE DURATION OF LIFE TABLE 21. Analysis of 488 fertile marriages where the ages at death of the married part- ners were both known and the number of children produced, showing the number of mar- riages, the number of children, and the number of children per marriage. B. B. Rec., XX, 135-140. Abstracts of the detailed tables in the Beinn Bhreagh Recorder, with graphical dia- grams. B. B. Rec., XIX, 242-268. 2. STUDIES OF THE HYDE STATISTICS TABLE . Number living at each age, number died at each age, and per cent died at each age. (A study of Table 3.) B. B. Rec., VIII, 287-298. TABLE A. Average age at death. (A study of Table 3.) B. B. Rec., IX, 98-103. TABLE B. Average age of males at death. (A study of Table 3.) B. B. Rec., IX, 172-176. TABLE O. Average age of females at death. (A study of Table 3.) B. B. Rec., IX, 177-181. TABLE D. Total: By age at death of both parents and by sex. (A study of Table 6.) B. B. Rec., IX, 182-184. TABLE E. Age at death: By father's and mother's ages at death. (A study of Table 6.) B. B. Rec., IX, 232-237. TABLE F. Age of father when person was born: By average age at death. (A study of Table 7.) B. B. Rec., IX, 265-279. . TABLE G. Age at death: By average age of father when person was born. (A study of Table 7.) B. B. Rec., IX, 280-300. TABLE H. Age of mother when person was born: By average age at death. (A study of Table 8.) B. B. Rec., IX, 315-326. TABLE I. Age at death: By average age of mother when person was born. (A study of Table 8.) B. B. Rec., IX, 327-346. TABLE J. Number of years after marriage of parents person was born : By average age at death. (A special study.) B. B. Rec., X, 55-90. TABLE K. Mother's age at marriage: By average age at death. (A study of Table 10.) B. B. Rec., XI, 46-87. TABLE L. Father's age at marriage: By average age at death. (A study of Table 9.) B. B. Rec., XI, 361-387. TABLE M. Number in family: By age at death. (A study of Table 5.) B. B. Rec., XII, 344- 353. TABLE N. Sex: By age at death. (A study of Table 3.) B. B. Rec., XII, 415-421. TABLE O. Analysis of 1,594 cases in which the ages at death of the persons and of both of their parents were stated, showing the number and percentages of persons who lived to be eighty years of age or older. (A study of Table 13.) B. B. Rec., XX, 94-99. TABLE P. Analysis of 1,594 cases in which the ages at death of the persons and of both of their parents were stated, showing the number of persons having fathers and mothers who died at the age-periods named, and the average duration of life of the persons and of their fathers and mothers. (A study of Table 13.) B. B. Rec., XX, 118-121. TABLE Q. An analysis of 843 fertile marriages of 795 males, showing the number of children produced from fathers who died at the age-periods named and the average number of chil- dren per father and per marriage. (A study of Table in B. B. Rec., XIX, 267.) B. B. Rec., XX, 153-158. TABLE R. An analysis of 671 fertile marriages of 671 females, showing the number of chil- dren produced from mothers who died at the age-periods named and the average number of children per marriage. B. B. Rec., XX, 159-162. A STUDY OF THE HYDE GENEALOGY 57 TABLE S. Analysis of 2,605 cases where the person's age at death and the father's age at marriage were both stated, showing the number having fathers who married at the age- periods named, the aggregate years lived by the persons, and their average ages at death. (A study of Table in B. B. Rec., XIX, 261.) B. B. Rec., XX, 176-179. TABLE T. Analysis of 2,309 cases where the person's age at death and the mother's age at marriage were both stated, showing the number having mothers who married at the age- periods named, the aggregate years lived by the persons, and their average ages at death. (A study of Table in B. B. Rec., XIX, 262.) B. B. Rec., XX, 180-182. TABLE U. Analysis of 2,756 cases where the person's age at death and the age of the father when the person was born were both stated, showing the number born to fathers of the ages specified, the aggregate years lived by the persons, and their average ages at death. (A study of Table in B. B. Rec., XIX, 263.) B. B. Rec., XX, 187-191. TABLE V. Analysis of 2,386 cases where the person's age at death and the age of the mother when the person was born were both stated, showing the number born to mothers of the ages specified, the aggregate years lived by the person, and their average ages at death. (A study of Table in B. B. Rec., XIX, 264.) B. B. Rec., XX, 192-194. TABLE W. Analysis of 2,757 cases where the person's age at death and the number of years after the marriage of the parents the person was born were both stated, showing the num- ber of persons born the specified years after the marriage of their parents, the aggregate years lived by the persons, and their average ages at death. (A study of Table in B. B. Rec., XIX, 265.) B. B. Rec., XX, 195-200. TABLE X. Duration of life of persons compared with number in family (siblings). (A study of Table in B. B. Rec., XIX, 266.) B. B. Rec., XX, 201-204. Conclusion of the Hyde Investigation. B. B. Rec., XX, 229-242. A complete copy of the Beinn Bhreagh Recorder is preserved in the Library of the Smithsonian Institution, Washington, D. C., where it is accessible to the public. The card catalogue used in the preparation of the foregoing tables has been pre- served in the Genealogical Eecord Office, Volta Building, 1601 Thirty-fifth street, "Wash- ington, D. C., where it is available for further research work by others. A. G. B. GENERAL LIBRARY UNIVERSITY OF ALL BOOKS MAY BE RECALLED AFTER 7 DAYS "niiF AS STAMPED BELOW SENT ON ILL OCT29 U. C. BERKELEY FOR MNO.DDO,50 m ,H/94 UNIVERSITY OF CALIFORNIA, BERKELEY BERKELEY, CA 94720 U. C. BERKELEY LIBRARIES ui - : - - UNIVERSITY OF CALIFORNIA LIBRARY