218 



N. YATSU : 



difference in their chromosomal characters. This seems rather remark- 

 able in view of the fact that the larvae arc so divergent in exter- 

 nal characters. It may, therefore, be concluded that among varieties 

 of the silkworm there is no morphological correlation between external 

 features and chromosomes. Thus the primary object of the present 

 study was frustrated. Nevertheless, I may mention the two following 

 facts as worth noting. 



1. The haploid number of chromosomes in the domesticated 

 silkworm is 28, contrary to ToYAMA's statement. This was determined 

 by counting the chromosomes at the metaphase of the first and the 

 second spermatocyte division (figs. 5, 6 and 7). The unreduced 

 number, therefore, must be 56. During spcrmatogonial divisions the 

 chromosomes cannot be exactly counted owing to their crowded state, 

 but they number unquestionably between 50 and 60, and decidedly 

 not 28 as To YAM A has stated ('94b, p. 132). There occurs appa- 

 rently the same number of chromosomes in the equatorial plate during 

 the division of oogonia and of the follicle cells of both the testis and 

 ovary. It need hardly be mentioned that the second spermatocyte 

 division is of the ordinary type, contrary to ToYAMA's observation 

 ('94b p. 137). That author thought that each of the 28 bivalent 

 chromosomes does not divide, but half of them go to one pole and the 

 remaining to the other, thus reducing the number to 14. It is a rather 

 singular coincidence that MUNSON ('06) also observed a similar mode 

 of division in Papilio rittulus. As a matter of fact, each of the 

 chromosomes becomes constricted into two, as is seen in the side-view 

 shown in fig. 8, each spermatid receiving 28 chromosomes. I have no 

 evidence to decide which of the two spermatocyte divisions is reducing. 



2. In the wild silkworm known to systematists as Thcophila 

 anandriana or Bombyx mandriano, the haploid number of chromosomes 

 is 27 (figs. 2, 3 and 4). The unreduced number should then be 54, 

 though for the same reason as mentioned above, the exact number 

 of the chromosome in the diploid group cannot be made out. 



