THE SPERMATOGENESIS OF UMBRA LIMI. 123 



the close of spermatogonial multiplication and the first matu- 

 ration division. Certainly if there is any increase in size in the 

 primary spermatocyte in such forms as Umbra it is very slight 

 as compared with the increase found in the primary spermatocytes 

 of most animals. The spermatogonial growth period is ac- 

 companied by very little mitotic activity of the cell. It is 

 initiated at the time the migrating cells assume a resting position 

 in the lobule and persists up to the first week in August or there- 

 abouts, although small clusters of isolated cells in a growing 

 condition may, as already indicated, be found throughout the 

 entire cycle of spermatogenesis. The large cells of the spermato- 

 gonial growth period and the cells resulting from their first 

 division are the most favorable for chromosome study. Such 

 cells (Fig. 2) in a resting condition may show one to three 

 chromatin nucleoli. The chromatin granules are more cen- 

 trally arranged than in the migrating cells (Fig. i) and 

 scattered along the limn network. The reticulum (Fig. 2) 

 appears more compact than in the early migrating cells. These 

 early growing or resting cells show (Fig. 2) from eighteen to 

 twenty-one small compact bodies which react deeply to nuclear 

 stains and are scattered throughout the reticulum. Their 

 relative constancy in number and the fact that they correspond 

 very closely to the number of chromosomes found in the spermato- 

 gonial cells, suggests the possibility of their being prochromo- 

 somes. They appear in the early resting or growing stages of 

 the cell as deeply staining bodies scattered throughout the 

 reticulum. In a very early prophase, as the chromatin granules 

 begin to collect along the linin threads, these bodies seem to be 

 the centers (Fig. 3) toward which the migrations of the chromatin 

 granules progress. A somewhat later prophase shows the chro- 

 matin aggregated along the linin threads in a beaded condition. 

 Although larger chromatin centers may still be seen, the beaded 

 structure (Fig. 4) of the forming spireme in the middle prophase, 

 makes further identification of these bodies as possible pro- 

 chromosomes uncertain. The evidence goes to show, however, 

 that these chromatin bodies are relatively constant as to number 

 and are aggregation centers for the future chromosomes. No 

 cells were observed in which such bodies could be seen in the 



