VISIBLE AND INVISIBLE CHARACTERS IN SILKWORMS. 143 



Nistri, the gametic composition is represented in small letters as • 

 ffyy (Table 4). There was no colour in either of its parents for 

 it to inherit, for the gametes, from the union of which it was 

 formed, bore the factors fy and fy. Table 4, Diagram (1). 



The cocoons of the Ital.-Jap. Hyb. are a pale flesh colour as 

 they inherited the flesh colour from the Yellow Italian parent 

 only, for the gamete from the White Japanese parent bore colour 

 factors lacking in both the flesh and yellow colours. The colour- 

 factors inherited from the Yellow 7 Italian parent were Fy 

 and from the white Japanese parent fy. The Ital.-Jap. Hyb. is 

 therefore not pure as regards the colour of its cocoon, for the two 

 gametes, from the union of which it was formed, were unlike and 

 did not carry pairs of similar factors. It is a hybrid (heterozygote), 

 the gametic composition being Ffyy, and it will give gametes Fy 

 and fy. When the Nistri (YYff), which has a deep yellow 

 coloured cocoon, was crossed with the Ital.-Jap. Hyb. all the 

 cocoons of ¥ 1 of this cross were of a bright yellow 7 colour, but not 

 quite the deep yellow of the Nistri. So the bright yellow of the 

 Nistri was dominant to the pale flesh colour of the Ital.-Jap. Hyb., 

 for the deeper yellow entirely masked the pale pinkish yellow 

 of the flesh colour even when it was present (Table 4). The 

 gametic composition of F 1 of the Nistri $> and Ital.-Jap. cross 

 was YyFf and Yyff, and it therefore contained two classes, 

 both numerically equally represented. Table 4, Diagram (1). 

 All the cocoons were yellow, but not quite uniform in tin. 



The gametes given off by these two classes were : — 

 YyFf giving YF, Yf, Fy, yf. 

 Yyff „ Yf, yf. 



In F 2 there were a few 7 deep yellow cocoons like those of the 

 Nistri, a good many bright yellow like those of F L , a few pinkish 

 yellow and a very few white out of 52 cocoons. The exact 

 numbers were — 13 deep yellow, 29 yellow, 6 flesh coloured, and 4 

 white out of 52 cocoons. 



The gametes for the colour character in F 1 were YF, Yf, Fy, 

 and yf, and their union at random would give, in F , cocoons of 

 the following gametic compositions— YYFF, YYFf, YYff, YyFF, 

 YyFf, Yyff, yyFF, yyFf, and yyff, which resulted in cocoons of 

 various shades with just a few white ones. 



This is just what occurred, and it will be seen that the 

 proportions of the various colours obtained out of 52 cocoons 

 run the proportion calculated very close, for according to the 

 analysis in Table 4 there should be 12 deep yellow, 24 yellow, 

 5 flesh coloured, and 7 white out of every 48 cocoons. Table 4, 

 Diagram (2). There was a difficulty, however, in distinguishing 

 between the deep yellow and the bright yellow from the 

 intermediate forms which occurred. 



In F 3 cocoons I found that in some cases when both the parent 

 cocoons were deep yellow, all the cocoons produced by the offspring 



