40 



The Journal of Heredity 



naires for the two periods. If the 

 ratio between the married and the un- 

 married persons used in the computa- 

 tions be made the same as in Table 5, 

 the averaoes are 2.63 children ( 1870- 

 1884), and 2.36 (1885-1899). 



It is clear that there has not been a 

 considerable or significant decrease in 

 the size of the families within the 

 group studied. 



The Families of Clergymen 



The Allegheny College statistics fur- 

 nish a limited opportunity for a study 

 of the families of clergymen. The 

 College has always been strongly re- 

 ligious. Many of its alumni are min- 

 isters. Some of them hold, and have 

 held, high offices in the Church. Table 

 7 summarizes the data for these men. 

 There are college presidents, a Y. M. 

 C. A. worker, a missionary, and so 

 forth, but all except eight are known 

 to have been actively engaged in the 

 Christian ministry. Of the married 

 individuals, only those are tabulated 

 whose total number of children to 

 date, dead and living, was furnished 

 by the questionnaires. 



A second computation of the aver- 

 age number of children per family was 

 made, discarding all cases where the 

 wife is dead or younger than 45 years. 

 The average for the earlier period 

 was 4.7 children (19 families), and 3.1 

 (25 families) for the later period. 



Compare Table 7 with Tables 3, 5 

 and 6. The most striking feature 

 brought out by this comparison is the 

 somewhat larger size of minister's 

 families, especially for the classes of 

 1870-1884. The superiority of the 

 clergymen almost disappears, however, 

 in the second period. The drop in 

 the size of the families by an average 

 of over one and a half children each, 

 suggests that voluntary limitation came 

 into vogue amf)ng the individuals of 

 the later period. 



c. The third and most important 

 question is, has the group under obser- 

 vation ])roduced enough children to 

 replace itself? Will the number of 



children who reach early manhood, the 

 approximate age at which their parents 

 graduated from college, be less, equal 

 to, or greater than the total number 

 of parents plus their unmarried class- 

 mates ? 



Many of the children of the earlier 

 graduates are doubtless middle-aged by 

 this time, but on the other hand a 

 considerable number of the offspring 

 of the later graduates are probably still 

 3'oung children. Therefore, direct ob- 

 servations on these children would not 

 furnish a satisfactory answer to this 

 question. 



The nearest approach to an accurate 

 answer is found by employing a mor- 

 tality table. In such a table a hypo- 

 thetical number of individuals (10,000 

 or 100,000 for instance) is represented 

 as starting life, say at age. The 

 probable number that will survive death 

 is then shown for each succeeding year 

 until extreme old age. I have employed 

 the Northeastern States Mortality 

 Table. ^' The data used in constructing 

 this table were "the death returns for 

 the five calendar years 1908-1912, in- 

 clusive, and the census returns as of 

 June 1, 1900, and April 15, 1910, for 

 the New England States and the three 

 Middle Atlantic States, New York, 

 New Jersey, and Pennsylvania." It 

 will be observed that this table is a 

 recent one covering the territory m 

 which a very large part of the gradu- 

 ates have lived. 



The numerical maintenance of this 

 limited section of the population ob- 

 viously depends not only upon the 

 replacement of the parents (one or both 

 of whom in each family is a graduate 

 of Allegheny College), but also of the 

 unmarried classmates of the parents. 

 A few of the men (14) married women 

 graduates. These women are counted 

 among the wives of the male graduates 

 and not under the heading of women 

 graduates in Tables 8 and 10. Thus 

 they are not counted twice. When- 

 ever a male alumnus married a second 

 time, the second wife is, of course, 

 counted as one of the parental group. 



