1910.] NATURAL SCIENCES OF PHILADELPHIA. 307 



at first was irregularly arranged, Is now collected into a spherical 

 mass at the proximal pole of the nucleus to form the beginning of the 

 nebenkern (fig. 56). The centrosomes in an archoplasmic mass still 

 retain their position at the distal pole of the nucleus, and from the more 

 distal one is seen to grow out a flagellum which corresponds to similar 

 structures to be described for both first and second spermatocytes and 

 interpreted as a precocious axial filament (figs. 55 and 56). As the 

 spermatid develops the centrosomes may be seen to migrate until 

 they assume their ultimate position at the posterior side of the 

 spermhead and the archoplasm or idiozome which surrounds the cen- 

 trosomes migrates with them (figs 57-59). After the centrosomes 

 have moved around to their final position at the proximal pole of the 

 nucleus, the idiozome is seen as a small body, clearer than the neben- 

 kern and lying beside it (fig. 60), and at this time, in the nucleus of 

 each spermatid, a round darkly staining body is to be seen (figs 56-61). 

 Since there is no evidence of an unequal division in either first or second 

 maturations and since this body is found in most of the spermatids 

 (its absence in some being explained by oblique cutting), there seems 

 no likelihood of its bearing any relation to the accessory chromosome 

 described by so many workers on insect spermatogenesis. I conclude 

 that it is a new formation, and that it is comparable to a similar body 

 described as " chromatin nucleolus" by Stevens (1906) for Coleoptera 

 and by Boring (1907) for Hemiptera. From the distal centrosome, 

 which now lies just beside the nucleus, the axial fibre continues to 

 grow; and as this grows through the center of the nebenkern and cell 

 cytoplasm both elongate, the former to form the inner, and the latter 

 the outer tail envelope. 



7. Centrosomes. — In the earlier "rest stages" of the spermatogonia 

 the centrosomes, which could be followed through every subsequent 

 step of development, were not visible; but the amount of cytoplasm 

 is so small and the cells so crowded that they might easily be over- 

 looked, and since they are to be seen in all spermatogonia divisions 

 of both the primary and secondary spermatogonia they may be assumed 

 to be present even though not seen except when the cells are actively 

 dividing. Just before the formation of the last spermatogonia! spireme 

 (figs. 4 and 5), when the chromatin is scattered through the nucleus 

 as separate granules, two minute centrosomes situated near the 

 nucleus are seen to divide and move toward opposite poles : the nuclear 

 membrane disappears and the mitotic figure is formed as usual (figs. 

 7 and 8). After the division of the chromosomes the centrosomes 

 divide and may be seen surrounded by a mass of archoplasm (fig. 10). 



