EUGENICS RECORD OFFICE. 147 



in one and the same chorion. Frequently a mother will have more 

 than one pair of twins, a condition that may be called repeating. This 

 phenomenon indicates a structural or physiological condition of the 

 ovary which readily permits double ovulation. It is easy to under- 

 stand how such an idiosyncrasy of the mother would tend to reappear 

 in her daughter and thus the tendency to twinning show itself as a 

 hereditary trait. 



A statistical study of the close relatives of twin-repeating mothers, 

 combined as a population, shows that in this population the ratio of 

 twin production rises to 4.5 per cent, which indicates that such mothers 

 belong to strains in which the factors for twinning are four times as 

 effective as in the population at large. If, on the other hand, one con- 

 siders as a population the close relatives of fathers of twins, then one 

 still finds that the incidence of twins in this selected population is much 

 above that in the population at large, namely, 4.2 per cent, a ratio nearly 

 as high as that found for the relatives of the mothers. These pro- 

 portions, calculated from the extensive records of the Eugenics Record 

 Office, lead to the inquiry : How is it possible that there shall be a 

 paternal inheritance in twin production that is as real, and nearly as 

 potent, as the maternal. Also, the tendency for twin production is even 

 stronger in identical than in two-egg twins, since the rate of twin pro- 

 duction in families that produce identical twins is about 13 per cent. 



The frequently denied possibility of inheritance of twinning through 

 the father's side of the house depends on a tacit assumption which 

 seems never to have been challenged by students of twin inheritance. 

 This assumption is that the determining, essential fact in twin pro- 

 duction as contrasted with single-birth production is the constant 

 double ovulation in the first case and the constant single ovulation in 

 the second. Whenever two eggs are simultaneously ovulated at a 

 period when fertilization occurs, there will be twins. Under such an 

 hypothesis it wdll be difficult to understand how the results are in- 

 fluenced by any tendency toward twin production from the father's 

 side of the house. 



Where the above hypothesis, however, fails is that about 8 per cent 

 of ovulations are double, according to counts made by Leopold and 

 other gynecologists. Thus the proportion of twins actually born is 

 less than one-fifth, probably only one-seventh, as great as the pro- 

 portion of double ovulations. To secure light on the question of what 

 has happened to reduce the proportion of twin births so far below that 

 of twin ovulations, a comparison has been made in the uteri of pregnant 

 swine between the number of embryos in the course of development 

 and the number of recent corpora lutea, each one of which indicates 

 one ovulation. A preliminary study made diu"ing the winter revealed 

 that there was a regular deficiency of from one to seven advanced 

 embryos, as compared with the number of corpora lutea. During the 



