208 



SCIENCE 



[N. S. Vol. XLVII. No. 1209 



we find that the number of children per 

 married woman graduate of Smith College 

 was only 1.3, of Vassar, 1.6, of Bryn 

 Mawr, 1.7 and of Holyoke, 1.8. Even 

 more significant is the ratio of children 

 per graduate, which, for all of these col- 

 leges, is less than one, due to the fact that 

 less than 50 per cent, of the graduates of 

 women's colleges marry. ^^ Professor Cat- 

 tell'" further shows in his study of 643 

 American men of science that the families 

 from which the scientific men had come 

 had an average of 4.7 children, while these 

 scientific men who were married and whose 

 families were completed had an average of 

 only 2.3 children, these figures including 

 all the children born. "We shall later see 

 how far from adequate such an average 

 number of children is for maintaining the 

 status quo in the respective groups. 



In the preceding discussion I have re- 

 ferred frequently to the number of children 

 per family in groups of our population and 

 have intimated that some families were 

 small, that others were fairly normal and 

 that others were large. You may now 

 properly ask me what the average number 

 of children per family should be. The 

 answer to the question depends upon the 

 point of view, namely, upon the amount 

 of natural increase one would wish to see 

 in a generation. As there is likely to be 

 a difference of opinion on this point, I 

 shall attempt to show only what the num- 

 ber of children in a family must be, under 

 present conditions, in order that the popu- 

 lation may remain stationary; anything 

 above this figure will mean natural in- 

 crease, anything below will mean a de- 

 crease in population. 



18 Nearing, Nellie Seeds, ' ' Education and Fe- 

 cundity, ' ' Quarterly Publications, American Statis- 

 tical Association, Boston, June, 1914, p. 126. 



19 Oattell, J. McKeen, ' ' ramiliea of American 

 Men of Science," Scientific Monthly, March, 1917, 

 p. 248. 



It is obvious that the basis of every fam- 

 ily is two individuals. The question then 

 resolves itself as follows: How many chil- 

 dren must be born to every family in order 

 that two individuals may be raised to ma- 

 turity? The number of children bom 

 must be more than two for a number of 

 reasons. The first is the fact of mortality. 

 The death rate is exceptionally high in the 

 period of childhood, amounting in the first 

 year of life to about 10 per cent, of the 

 babies born. If we begin with 100,000 at 

 birth and trace them through from year to 

 year, we find that about 78,000 are alive at 

 the average age at marriage. This is ac- 

 cording to the mortality rate that prevails 

 over a large portion of the United States. 

 The rest have died. This fact alone would 

 make it necessary that every marriage re- 

 sult in an average of 2^ children in order 

 that two persons may attain the average 

 age of marriage and replace their parents 

 in the population. But this assumes first 

 that all persons marry, and second, that 

 every marriage is productive. As a mat- 

 ter of fact all persons do not marry. In 

 our own country from 12 to 15 per cent, 

 do not marry until after the reproductive 

 period, if at all. A considerable proportion 

 of marriages, over 7 per cent., are sterile. 

 When we make the necessary corrections in 

 our figures, the average number of chil- 

 dren per family which must be bom is in- 

 creased to close to four^° in order that the 

 stock may maintain itself without increase 

 or decrease. 



I say without hesitation that a large 

 part of the native stock of this country is 

 in this sense not maintaining itself. It is 

 not producing the four children that are 

 necessary to perpetuate the stock. This 

 fact is true primarily for those families in 



20 Sprague, Eobert J., ' ' Education and Eace 

 Suicide," Journal of Heredity, Washington, D. C, 

 April, 1915, p. 158. 



