STUDIES ON GERM CELLS 423 



by an envelope of epithelial cells it seems certain that, as Wie- 

 man ('10b) maintains, the spermatozoa in a single cyst are de- 

 rived from a single spermatogonium. 



A cyst containing four spermatogonia is represented in figure 

 72. Here again appear the strongly basic-staining spindle re- 

 mains connecting the nuclei. These black strands persist until 

 the succeeding mitotic division occurs as figure 73, which was 

 drawn from a section of a cyst containing eight spermatogonia, 

 shows. Spindle remains are still evident in later stages, as in 

 figure 74, which represents a cyst containing thirty-two sper- 

 matogonia, but were not observed in cysts containing more than 

 sixty-four cells. 



Many investigators have figured spermatogojiial divisions 

 which result in rosette-like groups of cells similar to that rep- 

 resented in figure 73. Apparently, however, the remains of 

 the spindle, if present, did not possess such a strong affinity for 

 basic stains. Furthermore, only those of my preparations which 

 were fixed in Carnoy's fluid and stained in iron hematoxylin 

 exhibited these black strands. Similar spindle remains have 

 been observed in Dytiscus (especially by Giinthert, '10) and 

 Carabus (Govaerts, '13) during the differentiation of nurse 

 cells and oocytes from oogonia, and there can be little doubt 

 but that the process of cyst formation in the male, as described 

 above, is similar to the differential divisions in the female and 

 may also be compared with the differentiation of spermatocytes 

 and Sertoli cells in mammals (Montgomery, '11; von Wini- 

 warter '12). , 



In Dytiscus Glinthert found that the chromatic mass elimi- 

 nated from the nucleus always passed to the pole of the oogo- 

 nium containing the ^Spindelreste.' Govaerts ('13) has pointed 

 out that, although in Carabus and Cicindela no such chromatic 

 mass is demonstrable, still a distinct 'residu fusorial' exists, and 

 that during the differential divisions one cell differs from its 

 sister in the possession of these spindle remains. Polarity is 

 held responsible for the localization of the 'residu fusorial' and 

 the cause of the differential divisions is therefore considered to 

 be a 'polarite pre-differentielle.' 



JOURNAL OF MORPHOLOGY, VOL. 25, NO. 3 



