DEPARTMENT OF EXPERIMENTAL EVOLUTION. 109 



tistical analysis, or physiological chemistry or anatomy or microscopy 

 per se, but only for the assistance they can give to our main problem. 



In a department like ours it is necessary to keep an eye single to our 

 main purpose and avoid diversion of funds to subsidiary matters, how- 

 ever deserving of investigation. It is necessary to plan our work more 

 and more for a common purpose and to apply to all of our investiga- 

 tions the principle of cooperation. 



In detail our future plans are as follows : To make a series of prelimi- 

 nary reconnaissance studies on pedigrees of human traits, including in- 

 stincts and temperament, as particular opportunity arises. Statistical 

 studies on human mate-selection, differential fecundity, and the sex 

 ratio are also planned. On the experimental side it is proposed to push 

 the study of inheritance of instincts (in dogs), of tumor-growth (in 

 mice), of sex ratio, of the meaning of sex and sex intergrades, of fecun- 

 dity, of steriUty, of particular traits in animals and plants, including 

 rabbits, mice, pigeons, bantam fowl, Portulaca, and Datura. The 

 present status of studies on this material is more fully described in the 

 following pages and in the publications of the department. 



The main results of the year's work may now be passed briefly in 

 review. The Director completed his assignment at Washington to the 

 service of measuring 100,000 veterans at demobilization, and secured 

 (1) standard measurements for the use of the Army in making uniforms 

 (and incidentally for clothing manufacturers in general) and (2) a mass 

 of anthropological data concerning the American population compara- 

 ble to and greater in amount than that secured by Dr. B. A. Gould 

 at the close of the Civil War. In the conduct of the present work 

 leading anthropologists and anatomists of the country were enUsted. 

 A discussion of these measurements, together with those made on 

 2,000,000 men at mobiHzation, is now ready for the printer. 



An attempt was made by your Director to throw Ught upon the an- 

 cient problem of the meaning of human multiple births. A study of 

 original records showed an inheritable tendency on the maternal side 

 toward double ovulation, but also a nearly equal hereditary tendency 

 toward twin production on the part of the male. Also, the method 

 of inheritance appeared irregular. This led to a study of plural births 

 in pigs and to the discovery (which proved to be only the confirmation 

 of a discovery made by Hammond in 1914) that an important propor- 

 tion of fetuses fail of full development in utero, probably because of lack 

 of vital factors, while another fairly large percentage of eggs ovulated 

 fail (even under favorable conditions) of fertilization. The proportion 

 of these failures will be less, the more active, abundant, and freer from 

 lethal factors the sperm is. Fathers of twins, experience indicates, 

 belong to exceptionally fertile strains. Thus it comes about that 

 fathers of twins are about as apt to belong to twin-producing strains as 

 mothers of twins, and that twins depend on constitutional, hereditary 

 factors on both sides of the house. 



