316 bulletin: museum of comparative zoology. 



there is in addition other evidence nearly as strong. Although these 

 appearances are commonest in the primary spermatocytes, they are not 

 entirely restricted to them, hut occur occasionally in the " chromatin- 

 segment " stage of spermatogonia as well, and likewise very commonly in 

 the secondary spermatocytes, both before and after the disappearance of 

 the nucleolus (Plate 4, Figs. 74, 78, and 79) ; moreover they often ap- 

 pear in their most extreme type among the oogonia in the prophase 

 (Plate 5, Fig. 111). But whenever they occur iu female germ cells it is 

 always previous to the last oogonial division, where, of course, it can not 

 be supposed that they have anything to do with synapsis, since in that 

 mitosis no reduction in the number of the chromosomes occurs. Finally, 

 perhaps the strongest evidence of all is the fact that, while in cer- 

 tain preparations and in certain individual medusae they occur in great 

 numbers, in others they are comparatively rare ; this latter condition 

 was particularly the case in sections of a certain gonad in which all the 

 normal prophases of the primary spermatocytes were exceedingly nu- 

 merous. It is true, however, that they are present to a greater or less 

 degree in every preparation, either section or isolation, examined. 



In consideration of the foregoing evidence, I think we can not avoid 

 the conclusion that, at least iu Gonionemus murbachii, the " pseudosyn- 

 apsis" phases are something entirely apart from the normal stages of the 

 prophase of the first maturation division. There is one feature, and only 

 one, possessed in common by all cells in Gonionemus which exhibit this 

 phenomenon; that is the absence of karyoplasm — the " empty " con- 

 dition of the nucleus. (By this I of course do not mean that it is actually 

 empty, for no doubt it is occupied by transparent, probably inert, karyo- 

 lymph). This, I think, gives us the key to the situation. Instead of 

 being supported in a comparatively dense substance, as in the resting 

 stages, the now much thickened chromatic reticulum is supported, if at 

 all, only at its points of contact with the nuclear membrane, from which 

 it easily breaks away. I think we must consider the segments of the 

 spireme as being under tension, and thus particularly prone to shorten 

 under slight influences. The fact that the massing is usually towards 

 one side of the nucleus is readily explained by the reasonable assump- 

 tion that the connection of the strands with the. nuclear membrane is 

 not likely to be equally strong on all sides, and that they would break 

 away at the weakest points. Furthermore, when we recall how very 

 common is the phenomenon of the shrinkage of nuclear structures under 

 the action of reagents, we need seek no further, I think, for an explana- 

 tion of their occurrence in the present case. It is, of course, only in the 



