Edward N. Wentworth 



lUt 



seven. In fact the interpretation on a three-factor basis corresponds 

 very well to the distribution of these pairs. Assuming twenty to be 

 the niininiuni fecundit}', since it is approximately the lowest that 

 segregated out, and one hundred and forty the upper limit with rather 

 wide fluctuations, in all groups the intervening number divides readily 

 into six parts of twenty each. An hypothesis involving three pairs 

 would permit of six factors in the individual homozygous for high 

 fecundity, these factors to be of twenty each. If this were the case, 

 then the expectation and actual results would be as follows, the actual 

 pairs being somewhat arbitrarily distributed owing to their lying in- 

 termediate between the groups of twenty. 



_ _ _ 90 _ — — 



Circumstances prevented the breeding out of these different groups, 

 so no definite confirmation of the above hypothesis can be offered. 

 This does not account for the many progenies that numbered above 

 140. It is possible that at least four of these factors are concerned 

 in the race instead of three, but the particular pair under consideration 

 apparently bore only three of theni. The data used in Tables I^IV 

 inclusive more nearly fit a four-factor hypothesis than the three-factor 

 reported for Table VI. No evidence of sex linkage of fecundity factors 

 appeared. 



