STUDIES ON SEX IN CREPIDULA 23 



vesicles as in figure 32. If the nucleus does form a large num- 

 ber of karyomerites (fig. 30) there is usually a subsequent fusion 

 into fewer and larger elements (fig. 31). 



The division and dispersal of the centrioles accompanies the 

 disintegration of the nucleus. Previous to this time the number 

 of centrioles most frequently seen is two (figs. 25, 26), occa- 

 sionally four (fig. 27). They then usually become so dispersed 

 or so much reduced in size by repeated division, they they can- 

 not be identified at all (fig. 30) until they once more gather in a 

 group at the periphery of the cell (fig. 32). There are occasional 

 exceptions to this; one may, rarely, find the centrioles scattered 

 in the cytoplasm during the stage of the karyomerites (figs. 

 29, 31). 



In the mature testis, cells like that shown in figure 28 frequently 

 occur. They resemble none of the stages of the eupyrene sper- 

 matogenesis, and they are probably apyrene spermatoblasts just 

 previous to the disintegration of the nucleus; they appear to be 

 forming indistinct pachytene threads in the nucleus which might 

 be compared with those of a late stage of a eupyrene spermatocyte 

 of the first order (fig. 6). Certain of the spermatoblasts in which 

 the nuclear membrane has just been dissolved show their karyo- 

 merites gathered into a spherical mass as if they had been pre- 

 viously contained in a spherical nucleus (fig. 28) instead of an 

 irregular one (fig. 26). 



As stated above, where a large number of karyomerites have 

 been formed they subsequently fuse together to make nuclear 

 vesicles (figs. 31, 32). Several or many karyomerites group 

 themselves together irregularly and begin to fuse in such a man- 

 ner that their chromatic substance forms a shell, of irregular 

 thickness, around a hollow interior, in which a coarse chromatic 

 reticulum appears. There may be a few large nuclear vesicles, 

 or many small ones. 



The formation of the axial fibers now begins and the nuclear 

 vesicles undergo degenerative changes. Neither of these two 

 processes seems to be dependent on the other for either one may 

 go on with greater speed than the other. In some cases the 

 most of the chromatic matter has disappeared before the axial 



