152 J- Moore, 



masses are by no means uncommon throughout mammalian and other 

 spermatogeneses. In dogs there is not the slightest doubt that they 

 result from repeated a hinetic division of the nucleus. I have repre- 

 sented stages of this process in Fig. 30, 31; theoretically, it seems to 

 have a great deal of importance, but I reserve the discussion of the 

 subject for a later paragraph. 



The immediate effect of the foregoing process is to increase the 

 number of the spermatozoa ultimately formed (for each of the elements 

 thus produced is directly converted into a spermatozoon in the ordinary 

 way, as if nothing unusual had happened). As will be seen from the 

 figures, the cytoplasm in the multinucleate cell shows a marked con- 

 densation between and aroimd the nuclei (Figs. 30, 31, 33). From the 

 general relationships and appearance of this condensed mass, and for 

 more special reasons of which I shall speak immediately, there can be 

 no doubt that it represents the joint ar(;hoplasm of the several nuclei. 

 It is moreover curious to note that in these akinetically dividing 

 elements, the archoplasm loses its definite spheroidal form, and he- 

 comes a relatively large granular mass, applied to the nuclei, in the 

 same way as the granular archoplasm described by Meves, in relation 

 to the akinetic division witnessed in the adult salamandar's testis. 

 Since writing the above I have become convinced that the bodies 

 represented in Fig. 33 c are the centrosomes. I have unfortunately not 

 yet been able to satisfy myself of the presence of centrosomes, although 

 I think it probable that certain structures observed will eventually turn 

 out to represent them. The chromatic body is distinctly represented 

 by either a group or isolated groups of bodies, often incorporated within 

 the archoplasmic mass (Fig. 33 6 c). In the corresponding stage of 

 the rafs spermatogenesis, I have been quite unable to detect a hinesis 

 at any time, although there are certain indications of such a process 

 in the preceding generation of cells. At any rate, we find the nuclei 

 of these elements often duplicated in the same cell, and both Brown 

 and Ebner seem to have noticed this condition in the nuclei of the 

 growing cells. Of course this multinucleate condition is often produced 

 and pure kynetic division. These considerations seem to shoiu that 

 (ilcinesis can be produced apparently any ivhere during the course of 



