THE LIMITS OF HEREDITY 195 



the hereditary influence of the father on twin produc- 

 tion is approximately equal to that of the mother. In 

 thirty families containing twins classed as identical, 

 both the fathers and the mothers belonged to fra- 

 ternities in which there were 13 per cent, twin births. 

 Hence the inheritance factor in the production of 

 identical twins is nearly three times as strong as in 

 the production of twins in general, and again the in- 

 fluence of the father appears to be as great as that 

 of the mother. 



The sex of twins is also used as a means of deter- 

 mining how many are probably biovular or uniovular 

 in origin. In 160 pairs of twins in repeater families 

 (having two or more pairs of twins), in which the sex 

 was given, fifty- four w^ere of unlike sex and 106 of 

 like sex. Since the expectation for two-egg twins 

 is an equalit}' of like and unlike sexed twins, the 

 excess of fifty-two pairs of like-sexed twins was 

 probably uniovular in origin. This indicated that 

 about I in 3 of the twin births are of identical twins. 

 Davenport has also examined families in which one 

 parent of twins has married twice. In fourteen cases 

 the father married twice, in fifteen the mother, and 

 in one family both father and mother. Where the 

 father had married twice, there were twins of both 

 marriages in tw^o cases, or 14 per cent., and the corre- 

 sponding figure where the mother had remarried was 

 21 per cent., again indicating by these high values 

 the hereditary- nature of twinning. 



Until recently it would have appeared absurd that 

 the father could have any hereditary' influence on 

 twinning, but Davenport points out how our present 

 knowledge provides the basis for such a result, at least 

 as regards biovular or fraternal twins. In the first 

 place, there is evidence to show that the simultaneous 

 release of two eggs is much more frequent than are 

 twin births. There is statistical evidence from the 



