A. W. Bellamy 67 



A combination of the results given in Tables II and III gives for the 

 first two classes 389 and for the third class, composed entirely of MM, 

 142 ; the expectations being 398*25 and 132*75 respectively. 



Table IV. Combination of the results of the matings giving a " 3 : 1 " 

 ratio, viz. (3.4), (3.5), (3.6), (3.7), (5.2), (5.5), and (17). The apparent 

 dominance and recessiveness of certain of these patterns will be men- 

 tioned in another connection. 



I II III 



In Tables I, II, III, and IV, the results have been tabulated as though 

 C, D, and F were each " dominant " or epistatic, and M " recessive " or 

 hypostatic. This was due, in part, to the fact that when these records 

 were made the writer lacked experience and familiarity with the material, 

 which, because of the gi-eat similarity in the case of CC and CM, of the 

 homozygous and heterozygous forms, made it seem desirable to record 

 the two classes together ; and in part due to the fact that CC and CM 

 are not readily distinguishable until about a week after the last moult, 

 and even then there are a few individuals which cannot readily be dis- 

 tinguished. In the case of FF and FM, the pattern FF is so small that 

 it would be difficult to distinguish between FF and FM, even though 

 the hybrids were exactly intermediate between them. The case is rare, 

 if it ever occurs, where some detectable difference does not occur between 

 a homozygous form and the hybrid. 



The behaviour of H. An examination of Table I shows that the female 

 EEH in (9) gave two sorts of gametes, viz. E and EH ; that the female 

 DEH in (9.1) gave four sorts of gametes, D, E, DH, EW, and that all the 



