1 50 Inheritance in Philosamia Hyhrids 



as those just described, but on the left side of the seventh segment there 

 was no infra-spiraeular tubercle. The ofl'sjjring from this mating had 

 all the tubercles present, but the upper lateral and the infra-spiraeular 

 were only stumps. Again there is an indication of Mendelian segrega- 

 tion, but of course further breeding is necessary. The weather conditions 

 were so severe early in 1917, that jjrivet, the food on which the larvae 

 lived, could not be obtained and the families diud out. 



TABLE 111. 



Normal larva x Abuonual laiva 



I 



Fi 86 Normal larvae i^4 



i 

 F-i 10 families which had normal tubercles, short tubercles ami no tubercles Fr, 



1 



Fz m. 96 (nearly all normal) ; m. 100 and m. 103 (short tubercles) ; m. 98 -Fu 



(reduced tubercles) 



There is no recorded direct ancestor of the normal or abnormal parents used in this parti- 

 cular mating which had reduced tubercles in the three proceeding generations reared. But 

 they both had ancestors in the live original families which produced the tubercled forms. 



The generations on the left side refer to the breeding for the absence of tul)ercles, 

 whilst those on the right side are the total generations bred. 



Not onlj' from the general heredity view-p(.)int but also in relation 

 to the question of the origin of certain unusual forms in the Saturnidae, 

 these experiments are interesting. 



Most of the Saturnidae have larvae with tubercles, but a few, Cal- 

 losamia angulifera, Brahmaea japonica, Rhodia fugax, etc., have no 

 tubercles in the last instar. Apparently the smooth form has arisen as 

 a mutation from the tubercled forms within the group. What is the 

 relation of the abnormal forms with reduced tubercles which occurred in 

 the .^2 generation to the smooth forms. in the group of Saturniids ? Did 

 they arise because there was in the parent cynthia or in the parent 

 ricini a dormant gen for smoothness which appeared when inbreeding 

 took place in F., or was the original ci/iithia or ricini from a larva which 

 had mutated and then passed on the character f Presumably the gen 

 was not present on both sides or some non-tubercled forms would have 

 appeared in the Fj generation. As soon as material can be obtained the 

 cross will be repeated, when it will be known if one is dealing with a 

 mutation or not. Among the silkworms, normally with caudal horns, 

 with which Kellogg experimented, were found a number without a horn ; 

 these he mated, but he states(7), p. 58, "that the character is not 



