O. A. Merritt Hawkbs 145 



ment(12), p. 148 : " In the first generation of the crossing between white 

 and common marking worms, all the worms were of the common marking 

 and in the second generation two kinds appeared ; of the common 

 marking 77°/^ and of the white 23°/^. In the fourth generation the 

 worms produced by the mating of the white worms were all white." This 

 looks like a case of simple dominance, but later work of Tanaka ( 11 ), p. 2 1 1 , 

 states: "There is great variation of pigment intensity in the noi'mal... 

 type. In this we find almost every gi-adation from the lightest to the 

 darkest, apparently presenting a continuous variation.... Provisionally I 

 have divided the normal, according to the heritable characteristics into 

 four sub- types ; namely normal 1, 2,3,4." At the bottom nf the series 

 is a form '' hardly distinguishable from the pure plain " and he further 

 adds in a footnote, p. 211, "I assume the existence of different genes 

 respectively for the different sub-types of marking and colour." Now 

 this is what I also am inclined to do as regards the series of spots, 

 although it is stated already that thei'e have not been enough experi- 

 ments done to prove the point, but it is strongly suggested by the fact 

 that the spots disappear in an orderly and continuous manner. The 

 experiments in Japan by Tanaka, Toyama, etc. and those above recorded 

 all confirm the statement of Kellogg, p. 68 : " In larval colour-pattern 

 characters, the inheritance behaviour is rigorously alternative and 

 Mendelian, dominance always being consistent in relation to a given 

 colour pattern as related to another." 



Experiments have been made in crossing certain allied Saturnians ; 

 thus Joutel{6) and Pollard crossed Philosamia cynthia {advena) and 

 Callosamia prometlna, but as this ci'oss was made for the sake of the 

 imago rather than the larva, the results with regard to the latter are 

 not stated in detail, but it seems clear that there is a sex-linked inherit- 

 ance as the reciprocal crosses gave different results. Soule(lO) also 

 crossed these species, but her findings are different from those of Joutel. 

 Her results are shown in Plate LXXI of Packard(9), which shows plain 

 larvae as a result of the cross between a plain (P) and a spotted (S) 

 parent. This indicates that although P. cynthia advena and its Chinese 

 relative P. cynthia (Ning-po) appear so alike, genetically they are very 

 different — if these results are confirmed it will be an intei-esting point 

 in insect heredity. 



