238 Hi/bruls of Bistonine Moths 



Although, therefore, there are some appearances whicli suggest the 

 presence of a lieterotropic chi'oinosnine in the mahi, I am inclined to 

 believe that these are deceptive, and that one of the smallest chromo- 

 some jjairs constantly unites with one of the largest in the first 

 spermatocyte division, and divides normally, but that the closeness of 

 the union varies in different cases in the second division, so that some- 

 times thirteen, sometimes fourteen, appear to be present. 



2. The chroniosonies of Ni/ssia (Tthi/sid) zoiiaria. 



(a) Spermatugenesis. A glance at a spennatogonial equatorial 

 plate shows at once that the chromosomes are very unlike those of 

 L. (B.) hirtaria. Instead of 28 rather large chromosomes, there are a 

 very large number of extremely small ones. I have found no figure in 

 which they can be counted with accuracy; they tend to come into 

 contact and one can oidy make nut with confidence that there are over 

 100 (fig. 15). 



The synapsis and growtJi phases are closely similar to tliose of 

 hirtaria, except that the "bouquet" stage (fig. 16) is less typical; as 

 in that species there are sometimes four chromatin-nucleoli, consisting 

 of a larger and smaller })air, but more frequently tht'.se are united to 

 form three or two. They are very nearly of the .same size as those of 

 hirtiiria. The prophases of the first spermatocyte division are clearer 

 than in hirtaria owing to the much smaller size of the chromosomes, 

 and it can be seen that the chromatin-nucleoli approach one another 

 more and more closely, and finally unite into a single roundeil mass in 

 which no division can be seen. 



The primary spermatocyte (li\isions show equatorial pl.iti's of really 

 surprising beauty and perfection (fig. 17). The chromosomes lie 

 absolutely in one plane, and wiiJely separated from one aiiothei', so that 

 there is not the smallest difficulty in counting -56 with complete 

 certainty. Of these, two are noticeably larger than the rest, and 

 u.sually two others intermediate in size. The number 56 is so clearly 

 and certainly shown, that one may conclude with confidence that the 

 spennatogonial (diploid) number is 112, that is to .say, four times that 

 of hirtaria (28). The largest chromosomes of zonaria are of about the 

 same size as, or possibly even smaller than, the smallest of hirtaria. In 

 metaphase and anaphase it is seen that each chromosome is dividing in 

 the normal " heterotypc " manner, the diverging halves being connected 

 by double strands, and I can find no evidence of unequal division, nor 

 of the union of two pairs, such as occurs in hirtaria. 



