STUDIES ON INSECT SPERMATOGENESIS. 409 



series of stages cannot be obtained at one time in a single lobe, due 

 primarily to lack of available space. The unusual length of the sperm 

 heads is an added difficulty. The main outlines are, however, quite 

 clear enough for the present purpose, and the account given is drawn 

 largely from a single animal in which the cysts formed a fairly repre- 

 sentative series of stages. 



Stages I and m. From the formation of the spermatid to the casting 

 of of the acroblast. — These stages do not differ markedly from the 

 corresponding ones in the large cells. The nuclei are larger and appear 

 clearer, the chromatic substance seeming to undergo more complete 

 dissolution than in the normal spermatids. During the early elonga- 

 tion of the halves of the divided nebenkern, the centrioles, appearing 

 in the form of two rods in "V" formation as in the normal spermatids, 

 can be made out (Fig. 11), but clear cases are not frequent because of 

 the relation of the usual plane of section to the long axis of the 

 spermatid nuclei. It is clear, however, that the general history of the 

 centrioles is like that in the normal spermatids. The pseudoblepharo- 

 plast is formed in the typical way and the acrosome arises in connec- 

 tion with the Golgi apparatus (acroblast) which is then cast off exactly 

 as in the small sperm (Fig. 12). (Bowen '22a). 



The aspect of the peripheral chromatin layer, which gradually 

 develops on the inner surface of the nuclear membrane is, however, 

 clearly different, and serves at once to differentiate the large from the 

 small generations. This layer is never thick in the large nuclei, 

 forming a very thin layer (Fig. 12) similar in extent to that in the small 

 nuclei. This thin lining is, furthermore, not at all homogeneous but 

 appears more or less " vacuolated," so that internally the contour is 

 often noticeably rough. In spermatids of the small generation this 

 layer also exhibits some unevenness when the stain is not too heavy, 

 but it is by no means so characteristic as in the case of the large sperma- 

 tids. The comparison is interesting, however, because it shows that 

 the conditions in both generations are essentially alike. 



Stage n. From the migration of the acrosome to its definitive position 

 at the tip of the head, to the inauguration of the final steps in the condensa- 

 tion of the nuclear material. — The first part of stage w follows the same 

 general lines as in the small spermatids, the acrosome becoming 

 applied to one surface of the head and developing anteriorly a darkly 

 staining (with hematoxylin) granule; simultaneously the head begins 

 to elongate (Figs. 13 and 14). The most striking divergence from the 

 small sperms is again the appearance of the peripheral chromatic 

 layer, which is now distinctly irregular and often very uneven on its 



