320 SCIENCE PROGRESS. 



mesentery, figured by Flemming, the sphere is composed of 

 a centrosome, enclosed on the outside by the light space 

 {heller H of of the German authors) (Fig. III. c), from 

 which spreads a more or less extended radiation through 

 the general protoplasm of the cell. There is, however, 

 nothing comparable to an archoplasmic differentiation of 

 this substance, and in many cases (as for example in some 

 leucocytes) these simple radii extend at once from the 

 centrosomes to the cell periphery. They have in fact, 

 exactly the appearance of the radial envelope of a sphere 

 of which the archoplasmic constituent is wanting (cf. 

 Figs. I.-III.). 



The spheres in these somatic tissues therefore appear 

 to be less complex than those of the reproductive elements 

 with which we have been previously dealing ; they may 

 be referred to a simple in contradistinction to a more com- 

 pound type, and our appreciation of the mutual relationships 

 of the different modifications in their structure, will, like 

 other problems of comparative morphology, depend upon 

 the possibility of finding intermediate forms. 



So far, the compound spheres which we have examined 

 have been such as exist either in ova, or some element of 

 the ovigenetic series. It will be well therefore, lor the 

 sake of comparison, to ascertain what is known respecting 

 the nature of these bodies in the analogous spermatic series 

 of reproductive cells. In the spermatocytes of many, in 

 fact I believe of all animals, whether vertebrate or inverte- 

 brate, which have been examined with sufficient care, there 

 exists beside the nucleus a large slightly staining body, 

 generally known in the earlier literature as the "Nebenkern". 

 In the case of Salamandra and Proteus, this nebenkern was 

 shown by F. Hermann (18) to contain a central granule, 

 which, at the commencement of mitosis, divides into two 

 halves, that separate from one another through the mass, 

 and draw its substance out between them into an initial 

 spindle figure, since known as Hermann s Central Spindle. 

 The subsequent mitotic phenomena occur in relation to 

 these central granules, as if they were true centrosomes, 

 and consequently Hermann was justified in regarding the 



