SPERMATOGENESIS OF AMERICAN CRAYFISH 605 



distributed throughout the interior (fig. 80). In some cases, 

 the nutritive cells closely resemble the nuclei of the sperma- 

 togonia and it appears not at all impossible that the spermato- 

 gonia may be derived from the nutritive cells by becoming 

 surrounded with a mass of cytoplasm, as Grobben ('78), and 

 Herrmann ('90) have claimed. In other cases, spermatogonia 

 are seen in which the cytoplasm appears to be disintegrating, 

 thus baring the nucleus and this might lead one to the conclusion 

 of St. George ('92) and Keppen ('06), that some of the sperma- 

 togonia do not undergo further development, but degenerate 

 again into nutritive cells. 



The investigator cannot assert definitely as to the origin of 

 the nutritive cells and spermatogonia. But he is inclined to 

 the view advocated by Sabatier ('85), and Keppen ('06), that 

 both of these cellular structures originate from the germinal 

 epithelium of the testis, and that some of the nutritive cells may 

 then be derived from a disintegration of the spermatogonia! 

 cells. 



The germinal epithelium can best be studied in tubules which 

 are filled with ripe spermatozoa. In these it is found lining the 

 walls of the cavity (fig. 78). The component cells are large and 

 stain heavily. They contain huge nuclei which possess a great 

 many irregular masses of chromatin, having strong affinity for 

 the basic dyes. The cytoplasm is more or less consistent through- 

 out, with indications of boundaries here and there. These 

 cells correspond to the 'mother cells' or 'replacement cells' of 

 Sabatier. As the spermatozoa are discharged from the tubule 

 the epithelial cells multiply rapidly, displace the discharged 

 elements and soon give origin to the spermatogonia and nutri- 

 tive cells. 



The division of the germinal epithelium cells strongly suggests 

 amitosis. It is difficult, however, to uphold this with certainty. 

 The earlier investigators on the decapods all claim that these 

 cells undergo such division and the preparations of tl^p writer 

 seem to indicate such a conclusion (fig. 80). But in studies of 

 fixed material there is always a danger of misinterpretation. 

 The only way of accurately determining that such division does 



