284 NATHAN FASTEN. 



The first change is a disintegration of the dense chromatin of the 

 nucleus. This is consistently reduced until there are two 

 spherical granules (Fig. 45) and then one (Figs. 40 and 41) left. 

 These single granules appear like karyosomes. While all this is 

 going on the nucleus takes on a lighter stain and at the same 

 time, a mass of dense material makes its appearance in the 

 cytoplasm which stains like chromatin (Figs. 40, 41 and 45). 

 Koltzoff ('06) and Binford ('13) regard this material as mito- 

 chondria. I think this is nuclear material which has diffused 

 out into the cytoplasm as the chromatin content has become 

 reduced. I found a similar condition of affairs in the spermatids 

 of Cancer magister (Fasten, '18). 



Soon a vacuole makes its appearance at one end of the cyto- 

 plasm, the nucleus wanders towards the opposite pole and the 

 centrosome with the mitochondria-like mass take the middle 

 position between them (Figs. 42, 43, 46 and 47). In the 

 spermatid which contains the chromatoid body, all changes go 

 on as usual, but the chromatoid body wanders to the periphery 

 of the cytoplasm and is ultimately eliminated from the cell 

 (Figs. 39-43), thus playing no further part in the transformations. 

 From now on the spermatids seem to be similar and the same 

 changes take place in them. 



The mitochondrial mass soon becomes ring shaped and the 

 centrosome occupies its center. The upper end of the nucleus 

 has penetrated the open inner space of the mitochondrial ring 

 as seen in Figs. 48 and 49. The karyosome-like granule has also 

 wandered upward until it comes to lie directly below the centro- 

 some. The spermatid now appears like Fig. 49. 



Subsequent to this stage there is a fusion between the centro- 

 some and the karyosome-like body of the nucleus to form a 

 single structure. This then elongates to look like a short rod 

 (Fig. 50). Going hand in hand with these changes, a bubble 

 makes its appearance at the upper end of the first vacuole 

 (Fig. 50), which gradually increases in size (Figs. 50-55) and 

 becomes the second vacuole. In Lophopanopeus bellus a careful 

 study of the smear preparations reveals the fact that this second 

 vacuole was formed through a diffusion of substance from the 

 distal end of the central rod. This is clearly shown in the figures 



