314 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



number. As they now increase still more in bulk, and at the same time 

 diminish in number, it is probable that adjacent karyosomes coalesce 

 (Figs. 45, 46), while the strands of the network show similar changes, 

 becoming thicker and less numerous and staining more deeply. This 

 process of coalescence on the part of the karyosomes, and thickening on 

 the part of the strands, continues until finally there results, after the 

 several successive stages shown in Figures 44-46, a continuous and 

 homogeneous net composed of a comparatively small number of thick, 

 smooth, and very strongly staining strands, often with thickenings at 

 the nodes, as is shown in Figure 47. Fortunately, these changes do not 

 always take place at precisely the same time, so that several stages are 

 often to be seen in different parts of a single nucleus (Fig. 46). As the 

 cells are exceedingly abundant and, from the clearness of their nuclei, 

 easily studied on isolation preparations, it has been possible to follow 

 the process in detail. This stage, as shown by its earlier history 

 and fate, corresponds exactly to the " chromatin-segment " stage seen in 

 the spermatogonia, and hence both are to be regarded as the homologs 

 of the " spireme" of other forms. Cells of this stage show great varia- 

 tion in the number and arrangement of their meshes ; in some cases, as 

 in Figure 47, these are comparatively simple, while in others, as in Fig- 

 ure 49, they are so complicated and interlocked that it is impossible to 

 resolve them. But they all agree in showing no trace of a "beaded" 

 appearance, and we may therefore assume the union of the chromatin 

 granules to be extremely intimate. The time necessary for these changes 

 must be considerable, since a great majority of the cells in all the prepa- 

 rations examined exhibit some of the above described stages, and since 

 these changes are accompanied by a decided diminution in the bulk of 

 the cytoplasm (compare Figs. 40 and 47). 



We now come to a consideration of one of the most difficult, and at 

 the same time one of the most important, stages in the whole spermato- 

 genesis of Gonionemus, that of the "pseudosynapsis." The unravelling 

 of the processes leading from the spireme to the formation of the chro- 

 mosomes in the metaphase of the first maturation division has proved 

 exceedingly difficult, and even now, after careful study, T am only too 

 well aware that my interpretation may fairly be questioned. As is well 

 known, the majority of students of spermatogenesis have described a stage 

 occurring at some period between the last spermatogonial and the first 

 maturation divisions, in which the chromatin strands become more or 

 less densely massed at one side of the nucleus. This was first observed 

 by Moore ('95), and, following him has usually been termed the stage 



