548 CELL DIVISION IN EGGS OF CREPIDULA. 



or many in number, and they may be entirely distinct or may be partially 

 united, thus giving rise to nuclear forms which might be interpreted as nuclear 

 constriction or budding (figs. 175, 177, 194-196, 198, 200-202, 205-208). Indeed 

 they are so interpreted by Konopacki, but in Crepidula a careful study of the 

 karyomeres with regard to their stages in the nuclear cycle shows that the 

 unions are more complete in the later than in the earlier stages, thus proving 

 that they are fusing rather than separating. However, where the karyomeres 

 are quite distinct they may remain separate in all the cells through several cell 

 generations (figs. 205-208) ; evidently the union of the karyomeres into a single 

 vesicle can take place only when they are in intimate contact. 



In all my preparations the nuclear vesicles are much more numerous than 

 the nuclear asters. If both nuclei and asters multiplied by constriction or 

 fragmentation this condition would be difficult of explanation, but if the nuclear 

 asters and centrosomes arise by bipartition, while the karyomeres come from 

 scattered chromosomes, the explanation of this condition is obvious. The 

 centrosomes are unit structures, the nuclei are compound, being composed of 

 many chromosomes or chromosomal vesicles; while each centrosome becomes the 

 center of an aster, each chromosome does not usually form a separate nuclear 

 vesicle, though it is capable of doing so, but unites with other chromosomes to 

 form one or many nuclear vesicles. 



Considering the fact that these irregularities in the movements of chromo- 

 somes are among the most common forms of abnormal cell division whatever 

 the disturbing factor may be, the question arises whether many of the cases of 

 amitosis in normal developmental processes, which have been recently described, 

 may not be really cases of the formation of separate karyomeres, and their sub- 

 sequent fusion, such as occurs in these experiments. I refrain from commenting 

 here on this subject further than to suggest that many of the figures given by 

 recent writers on this subject bear a striking resemblance to these karyomeres 

 of Crepidula. In the latter case, however, the actual lineage of these karyomeres 

 may be determined with great certainty and it can be affirmed that they arise 

 from irregularities in the separation of chromosomes in mitotic division. 



The size of these karyomeres is, within the same cell, proportional to the 

 number of chromosomes which enter into them, as Boveri (1905) found to be 

 the case in echinoderm eggs. However in cells which differ in their quantity of 

 cytoplasm, nuclei and karyomeres may differ in size, though they contain the 

 same number of chromosomes, since the ultimate size of a nucleus depends not 

 merely upon the number of chromosomes which it contains but also upon the 

 quantity of cytoplasm in which it lies (Conklin, 1912). 



8. Segregation of Chromatin and Achromatin. 



Attention has already been called to the fact that isolated portions of 

 achromatin may assume the form of cytasters during periods of mitosis and of 

 vesicles with a definite membrane during periods of rest. The former condition 



