BIGELOW: NUCLEAR CYCLE OF GONIONEMUS MURBAOHII. 371 



twelve. When we come to compare this process with the corresponding 

 one in the spermatogonia, it is clear that the earlier prophase, and the 

 occurrence of chromomeres from the fusion of which are formed half as 

 many chromosomes, are practically identical in both cases. The forma- 

 tion of chromosomes by the pairing of distinct pre-existing bodies is in 

 Gonionemus prubably peculiar to these two cell generations ; certainly 

 it occurs neither in somatic cells nor in oogonia. Yet, since in the sper- 

 matogonia, in which the chromosomes appear in the full somatic number, 

 the pairing can, of course, have nothing to do with synapsis or reduction, 

 it seems too unreasonable to give it such significance in the spermatocyte. 



Without entering here upon the theoretic side of the question, we 

 may briefly consider two or three possible explanations of this reduction. 



In the first place, if the primary spermatocytes were considered by 

 themselves, apart from the preceding generation, it might naturally be 

 supposed that, as in so many other metazoan cells, the chromatin struc- 

 tures resulting from the segmentation of the chromatin net were ordinary 

 chromosomes in the usual somatic number (twenty-four), and that these 

 then paired by a synaptic process to form the twelve bivalent chromo- 

 somes of the first maturation spindle. I repeat, did this cell generation 

 stand by itself, or were it taken only in conjunction with the processes 

 in somatic cells, such an explanation would seem altogether the most 

 reasonable and natural one. But when we compare it with the corre- 

 sponding stages of the spermatogonia, this solution does not seem to fit 

 the case so well, for it gives to the process of pairing in the spermato- 

 cytes a significance which it cannot possibly bear in the earlier genera- 

 tion, although in both the actual course of events is, so far as can be 

 seen, the same. 



There is another explanation which seems to me to fit more fully the 

 actual internal evidence afforded by the chromatin structures, although 

 it is perhaps less easy to reconcile with the stages in the germ cells of 

 other animals. This is, that a pairing of individual chromosomes does 

 not occur at all in Gonionemus, but that synapsis occurs between the 

 chromatic microsomes ; and takes place while these are intimately asso- 

 ciated in the homogeneous net. If we accept this solution, and to me, 

 although the evidence is not altogether convincing, it seems the more 

 probable one, it not only accounts for the occurrence of the chromosomes 

 in the reduced number, but also offers a possible explanation for the 

 formation in the spermatocytes of a continuous chromatin net, instead of 

 the irregular and independent chromatin segments which occur in sper- 

 matogonia, oogonia, and somatic cells. Finally, there is still one more 



