156 J- Moore, 



plasm, where it takes no further part in the spermatogenesis, being 

 eventually cast off, together with a degenerating mass of protoplasm, 

 as a part of the residual corpuscle of authors. 



During the archoplasmic metamorphosis, the cytoplasm of the 

 spermatid becomes very uniform and less obviously granular, while the 

 chromatic body becomes proportionately marked and assumes a beautiful 

 rounded contour (Fig. 22). Some dismembered fragments of chromatin 

 are still however to be found in its vicinity. Sometimes, the small 

 double staining structure (which I have spoken of as appearing when 

 the granulation incident to the division has more or less subsided, 

 and which I think there is every reason to believe represents the 

 spermatid centrosome) is found at this stage also in the vicinity of 

 the chromatic body, and I have sections which seem to indicate that 

 this is invariably its primitive position. As time goes on, however, 

 the duplication of this structure becomes more marked, and the two 

 small bean-shaped bodies into which it revolves itself assume the 

 indiiferent positions represented in Figs. 22, 15. In some of my 

 sections these bodies appear in numerous consecutive ceUs, and I have 

 represented such a field in Fig. 5 c. Wherever they may lie, if seen 

 at all, they can be made out with tolerable clearness, and they have 

 precisely the same appearance as the centrosoraes of the previous 

 cellular generation. These spermatid centrosomes would therefore hardly 

 seem to correspond to the "Spermocentres" of Julin. 



I see no reason to suppose that in mammals these bodies are 

 formed afresh in the spermatids, as the Belgian author found to be the 

 case in " Sty elopsis grossularia". In the rat, the centrosomes of the 

 previous generation persist until the spermatids are quite formed, and 

 it is only after the contraction of these cells and the consequent 

 crowding of the parts, that it becomes impossible to distinguish small 

 bodies in the dense granulation visible. Later on, as we have seen, 

 bodies, answering in every particular to these lost centrosomes, re- 

 appear as the granular confusion subsides, and every consideration 

 points to the conclusion that they are the centrosomes wliich have 

 been for a time obscured. 



All these changes, the appearance of the tail, its minute basal 



