L. DONCASTRR 5 



two families ('12.25, 12.40) there were exceptions to the ordinary 

 sex-limited transmission. These will be referred to more fully below. 

 It will be noticed that in several families the same reference number 

 is used, followed either by ' (2) ' or by ' B.' Such families as '12.8 and 

 '12.8 {2) wera reared from eggs of the same parents, 'li2.8 being from 

 the first eggs laid, '13.8 {'3) from eggs laid later. The object of this 

 was to determine whether the earlier and later eggs gave similar 

 results as regards sex-ratio. The results show that there is never any 

 important difference between the ratios of the sexes from the two lots. 

 When a family is distinguished with 'B' (e.g. '12.21, '12.21B), it was 

 reared from parents which were brother and sister of those of the 

 family with the same number without ' B.' Thus the male parent 

 of ' 12.21 B was brother of the male parent of '12.21, the female 

 parent sister of that of '12.21. Such pairs of families are in each 

 case closely similar in the 1912-13 families, as they were in previous 

 years, but the fact that two sisters mated to males of closely similar 

 ancestry sometimes give very different progeny (e.g. '12.1 and '12.3) 

 makes it doubtful whether the resemblance between the offspring of 

 two sisters mated to two brothers will be found in every case. 



There were no cases in 1912-13 of females belonging to bisexual 

 broods having only female offspring, such as occurred in three families 

 in the preceding year. Matings of females belonging to broods in 

 which males were very scarce usually gave a smaller preponderance 

 of females, when the sex-ratio differed from equality, than occurred 

 in the mother's family (e.g. '12.16, '12.20); there were no cases of a 

 great increase in the ratio of females to males. It seems, therefore, 

 that when the tendency to produce female offspring is not strong 

 enough to cause all the progeny to be female, it suffers a progressive 

 decrease in intensity in subsequent generations. 



Complete or nearly complete sterility occurred in six matings with 

 females of unisexual families, out of a total of 21 such matings, and 

 in four out of 23 other cases. The infertility of females of unisexual 

 families is thus not much greater than that of related moths belonging 

 to bisexual broods. 



It may be mentioned incidentally that among the larvae of family 

 '12.27 B, seven were almost totally black. The moths when hatched 

 showed no peculiarity. Two of them paired together gave a brood of 

 23, which were all nearly or quite black. On the other hand, a male 

 and a female derived fi-om black larvae, paired with moths derived from 

 normal larvae, both gave broods consisting only of normals. The black 



