7 6 



Scientific Proceedings (104). 



brothers of twin-producing mothers 4.5 per cent of the children 

 are twins. These figures indicate that the twin ratio is increased 

 4 to 7 times in twin-producing families and that the ratio of twin 

 production is about as high on the father's as on the mother's side 

 of a fraternity which contains 2 or more twins. This result dis- 

 poses of the statement that fathers play no role in twin production, 

 but at the same time raises the query — How can this be? 



It is relatively easy to understand how the sperm may in- 

 fluence the early division of the fertilized egg so as to produce 

 identical or 1 -egg twins, and we find indeed that the rate of occur- 

 rence of twin production is high, both on the paternal and on the 

 maternal side of such fraternities containing identical twins. 

 The rate in both cases is about 1 3 per cent. But if one takes only 

 those fraternities producing 2 or more sets of twins of unlike sex, 

 we still find an equality of influence on the paternal and on the 

 maternal side. 



The difficulty in the way of understanding the equality of the 

 part played by the father and the mother in the production of 

 2-egg twins arises from the assumption that such twins are due 

 merely to double ovulation. Were they due merely to double 

 ovulation, then there would be as many sets of twins born as 

 there are double ovulations that occur at the time of conception. 

 That this assumption is false is shown by several lines of evidence. 



1. Not all of the eggs that are ovulated are fertilized. In the 

 case of 4 sows' uteri examined the number of embryos (well 

 advanced in development) found was 22 and the number of recent 

 corpora lutea 34. This indicates that fully one third of the eggs 

 laid at these periods of conception were not fertilized. This 

 number is, I think, rather too high for a general statement, 

 probably not more than 10-25 per cent, of eggs ovulated at favor- 

 able periods fail of fertilization. In humans also there is evidence 

 based upon counts that have been made on the corpora lutea of 

 non-pregnant and pregnant women that the rate of occurrence 

 of 2.3 corpora lutea in the ovary is much greater than the occur- 

 rence of twins. Thus, to cite only a single study, in 33 sets of 

 ovaries with corpora lutea or follicles about to burst, 5 (or 15 

 per cent) showed that double ovulation was occurring. It seems, 

 therefore, certain that many more cases of double ovulation occur 

 than of twins. 



