236 Hjjhr'nh of B!>itonme Motha 



themselves to be composed of smaller units which show various degrees 

 of separation. The counting of the double rods is not quite easy; 

 usually thirteen seem to be present, but in some cases I cannot find 

 more than twelve. If, as seems probable, there are thirteen, the larger 

 and smaller members of the chromosome-nucleolus cannot consist of a 

 large and small pair of united chrcimosomes, as occurs in the spermato- 

 genesis (see below). Seiler has described the " sex-chromosome " in the 

 female of the moth Phragmatohia fidiginosa as being double', and the 

 doubleness of the chromatin-nucleolus seen in hirtaria may perhaps be 

 due to the same cause. As I have not investigated the polar divisions 

 of the eggs, I cannot test the truth of this suggestion. 



(b) Spermatogenesis. I have a number of clear figures of the 

 spermatogonial divisions, and as in those of the oogonia there are clearlj' 

 28 chromosomes, of which si.K are smaller than the rest, and four of these 

 conspicuously smaller (fig. 4). In the gi-owth-stages of the spermatocytes 

 a typical sjmapsis of the spireme occurs. As in other Lepidoptera, the 

 stages are less easy to arrange in order than in the ovaries, but there is 

 no doubt that a thin spireme thread, closely withdrawn to one side of 

 the nucleus, is succeeded by a thicker and looser spireme (figs. .5, 6), 

 and this by a reticulum, from which the double chromosomes of the first 

 spermatocyte division arise by a process of concentration. 



In the younger spermatocytes there are usually two chromatin- 

 nucleoli; in the older ones commonly four, two larger and two small. 

 These may be united so as to give three or two, a large one commonly 

 being paired with a small one. I cannot find any constant dift'erence in 

 size between the two pairs ; one of the larger or of the smaller is some- 

 times of greater size than its fellow, but this does not seem to be 

 regular. The stages in which the chromatin threads contract into 

 chromosomes are so confused that I am unable to count them, or in 

 many cases even to distinguish the ciiromatin-nucleolus from the other 

 chromosomes. 



The primary spermatocyte divisions are usually extremely clear and 

 diagiamraatic, but are somewhat difficult to interpret, for in place of the 

 14 chromosomes which one would expect they constantly have only 13. 

 Careful inspection, however, shows that one of the largest is comj)ound, 

 and consists of a large chromosome to which a very small one is attached 

 (figs. 7, 8). That the large one is double (i.e. composed of an ecjually 

 matched pair) is often quite clear, but the small one attached to it 



' J. Seiler, Zool. Aiizeitjer, xli. 1913, p. 246. 



