348 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



strands to the surrounding and supporting karyoplasm is best shown on 

 crushed specimens, where it is seen that, though rather firmly coherent, 

 they are not sharply defined from the karyoplasm, but merge insensibly 

 into it — a condition in contrast to that seen in the late prophases of 

 the various mitoses in Gonionemus, in all of which the chromatin strands 

 and segments are entirely distiuct from the surrounding karyolymph. 

 With the further growth .of the nucleus the strands increase in length 

 and also in number ; the readiest explanation of this latter change is of 

 course that the Y-shaped figures above described are due to longitudinal 

 splitting. With this increase in number, and with the further growth 

 of the nucleus as a whole, their relations to one another become so in- 

 tricate that it is extremely difficult to determine whether they are still 

 completely separate or form a loose and irregular net ; a section of a 

 nucleus Showing this condition is represented in Figure 133, in which 

 sections of various strands are seen, none of them probably being en- 

 tire, aud in addition the chief and the accessory nucleoli. With these 

 changes the individual chromatin strands seem to diminish somewhat in 

 thickness, though still preserving their granular consistency and show- 

 ing that they are composed of series of separate block-like masses ; the 

 apparent dwindling is to be explained, I think, on the assumption that 

 the rapid growth of the nucleus allows them to elongate more readily 

 (Figs. 132 and 133). This is the most advanced stage to which I have 

 been able to follow the changes in the chromatin structures, and since the 

 germinative vesicle has now come to lie close against the periphery of 

 the egg, I have little doubt that further stages must be sought after the 

 eggs ai*e set free. 



Having thus briefly considered the genesis of the chromatin struc- 

 tures of the maturing germinative vesicle, and having seen that they are 

 formed from the preexisting nuclear reticulum without any apparent as- 

 sistance from the nucleoli, I return to trace the further histm-v of these 

 latter bodies. The chief nucleolus at the stage last described (Fig. 131) 

 consisted of a spherical mass of an apparently homogeneous granular 

 ground substance without membrane, but enclosing one or more large 

 vacuoles. With the further growth of the germinative vesicle, the 

 nucleolus increases rapidly in bulk. Different specimens now exhibit a 

 variable number of vacuoles; but since Pflucke ('95) has demonstrated 

 in the living nucleus in gasteropods that one lar*ge vacuole may break 

 up into several small ones, and then these may again unite into a single 

 large one, variation in this respect cannot be considered of any great sig- 

 nificance. Much more important, and at the same time of a mor*e ob- 



