288 FRANZ SCHRADER. 



impending division regardless of the proximity of the cleavage 

 cell, or whether the presence of the latter will make that division 

 abortive. In the first case, 8 nuclei will commence the fusion 

 process as before; in the last named eventuality only 4 will be at 

 hand (Fig. 70). 



On the whole it may be assumed that the polar nucleus as well 

 as the fertilization nucleus undergoes a definite number of divisions. 

 As has been noted previously, the chromosomes of any single giant 

 cell are from their first appearance alike in size and shape. The 

 chromosomes of the polar nucleus derivatives however decrease 

 in size with each succeeding division (Fig. 4 to 6) and the same is 

 true of the cleavage cells. If fusion or combination of these two 

 types of cells could occur after a varying number of divisions, it 

 would be expected that the chromosomes of the combination 

 nucleus would often be of two sizes. But this, as has been said, 

 is not the case. It is of course possible that there is some regula- 

 tive mechanism capable of equalizing differing sizes of chrom- 

 somes, but for this assumption there is little or no basis. 



Regarding the phenomena of fusion which now occur, the con- 

 clusions based on the numerical data receive the full support of 

 the' purely cytological evidence. In giving this last named proof 

 I am fully aware of the ease with which in a case like the present, 

 a number of isolated figures can be seriated to fit a preconceived 

 hypothesis. Standing alone, the cytological proof would there- 

 fore be advanced with considerable caution. Nevertheless one 

 or two of the figures found are of considerable value in themselves. 



Every step in the migration of the cleavage cells to the periph- 

 ery, their approach to the polar nucleus derivatives, the flow- 

 ing together of the protoplasmic areas which surround each 

 nucleus and the final apposition of the nuclei within the single 

 protoplasmic area resulting, can be traced through closely seriated 

 stages. Similarly, what appears to be a pair of polar nucleus 

 derivatives may at times be seen in close proximity. However 

 unless the chromosomes of the apposing nuclei are close to full 

 condensation, no definite conclusion can be reached as to the 

 nature of the nuclei involved in either case. All of the figures 

 show that the process takes place in either two or three cells, 

 and a greater number has not been observed (Figs. 8 to 10). 



