Heredity of Race-Characters in Silkworm (Bombyx mort) 761 
univoltinism is subpotent, while in Brood J, bivoltinism 1s totally 
prepotent. Whether the bivoltin will remain prepotent in the 
offspring of Brood J and univoltinism will remain prepotent in 
the offspring of A is yet to be determined. 
DISCUSSION OF RESULTS 
The contrasted characters that are made the basis of this series 
of experiments are “univoltin-producing” and _bivoltin-produc- 
ing.” These characters are known to silk-worm breeders as 
racial characters. The univoltin race breeds true in pure lots, 
with the exception of an occasional sport-bivoltin as previously 
noted. Unfortunately I was unable to obtain any pure bivoltin 
silkworms except the single individual with which the experiment 
began. The first generation of the cross U2 x B & follows the 
racial character of the female parent. If the character that fails 
in appearance in the first generation (the bivoltin) be considered 
the “‘recessive” in Mendelian terms, then in the second genera- 
tion (in matings within hybrids) there should be, according to 
Mendelian hypothesis, a ratio of three univoltin broods to one 
bivoltin brood and thereafter the bivolts should breed true. While, 
in the present instance, the first generation brood follows but 
one of the divergent parent-characters (univolt when the female 
is univolt, bivolt when the female is bivolt) in the second genera- 
tion, the broods fail to follow both parents in the expected pro- 
portions. Therefore the Mendelian law does not hold in this 
case. 
If we consider that the phenomenon may be Mendelian com- 
plicated by a distinct correlation between sex and character, we 
still must account for the abeyance of the character for several 
generations and its later reappearance in certain strains. If this 
can be accounted for through the successive failure of the meet- 
ing of the right “determiners” then, according to Mendelian 
expectation we must look for the “breeding-true’”’ of the reverting 
individuals. This is true only in a very limited number of broods 
and only toward the end of the series. 
On the contrary, this case seems to add to the evidence adduced 
