146 Inheritance in Philosamia Hybrids 



Section II. The Inheritance of the Plain Larvae. 



The plain larvae, as stated in Section I, appeared in the F^ generation 

 in the proportion of a simple Mendelian recessive, viz. one of the reces- 

 sive to three of the other types. 



The matings of plain to plain produced only plain larvae ; of such 

 matings, there were six in the F^ generation, one in the F^, seven in the 

 F-a, and six in the F^ generation. 



These results agree with those of Kellogg and the Japanese workers 

 (see p. 144). 



Mr J. W. Watson crossed P. ricini (^ (plain larva) with P pryeri % 

 (spotted larva) and produced a hybrid which he called Rothschildi (13). 

 These F^ larvae were all spotted, there being both S and RS forms. He 

 inbred the Fi generation and gave me some of the eggs. These &g^^ 

 produced an F.2 generation which consisted of P, >S^ and RS larvae in the 

 proportion of 1 P to '6Sx RS. Only one mating was made between 

 P parents, but that produced only P larvae. In this mating with pryeri, 

 ricini behaved genetically exactly as in the mating with cynthia.- 



Section III. On Larvae with reduced Tubercles (Scoli). 



P. cynthia and P. ricini have throughout their larval life six com- 

 plete longitudinal series of tubercles, three on each side, the upper rows 

 being the best developed ; the lowest row is shorter than the others. In 

 the three thoracic segments there is also present a fourth bi-lateral series 

 of tubercles, situated at the base of the thoracic legs ; these are very 

 short, especially the first pair, which may be reduced to mere knobs. 

 The setae borne by the tubercles vary in proportionate length in the 

 various stages, being fewest and shortest in the last. (Text-fig. 1.) 



. The tubercles of the two species are alike in all anatomical features, 

 but differ in colour, those of cynthia (Ning-po) having a slightly pink tip. 



The tubercles in P, were normal in appearance, but among the twelve 

 families of Pg five matings produced a few larvae with abnormal tubercles. 

 These larvae were very variable, the tubercles being entirely absent, 

 very short or varying in length on the various segments of the body. 

 (Plate VIII.) 



Before proceeding to a statement of the breeding experiments, it 

 will be as well to define the term tubercle, which is often used very 

 vaguely, and hence denotes a variety of larval appendages. Fracker(3), 

 p. 44, states that the term " tubercle has been used to mean any cuticular 



