144 



spermatozoa enter some of these pieces, as Hertwig affirmed, I can 

 amply verify. From the very nature of the problem, it is difficult to 

 tell whether 4n any particular case a non -nucleated fragment has seg- 

 mented, because in the living egg it is absolutely impossible to tell 

 whether a fragment is nucleated or not. The very fact that the protoplasm 

 has been shaken up often makes even in the stained eggs the line 

 between nucleus and protoplasm very obscure. Further, stained eggs 

 show that occasionally the nuclear membrane breaks down setting free 

 the chromatin in the protoplasm. As the chromatin particles are ex- 

 ceedingly fine and scattered in such cases one may often believe them 

 to be non-nucleated, when in reality all the chromatin is still present. 

 How then Boveri has been able to determine that monospermic fertili- 

 zation has taken place in certain cases and that only these egg-frag- 

 ments develop, I do not know. If his conclusion is mainly based on 

 the size of the nuclei in certain larvae, I think the following results 

 throw very serious doubt on the value of such evidence. 



We have an unusual opportunity in the case of the nucleated egg- 

 fragments to test the relation existing between the size of nucleus and 

 the size of the cell. The test is here more crucial than in the normal 

 cleavage processes where hereditary conditions vitiate the result. The 

 nucleated fragment contains a segmentation nucleus of full size. After 

 the first cleavage each of the daughter nuclei is again enclosed in 

 a nuclear membrane, and the new nuclei are smaller than the nuclei 

 of the normal two cell stage and in general smaller, the smaller the 

 fragment. In other words, the new cells from the segmented egg- 

 fragments, themselves smaller than the corresponding normal cells, 

 enclose their chromatin in a smaller nuclear membrane. My study of 

 these fragments has not as yet been carried sufficiently far to make 

 any statement as to the amount of the chromatin in these segmented 

 egg - fragments. 



In the light of these observations it seems to me one side of 

 Boveri's proof gives way. If Boveri had counted the number of 

 chromatin granules in the paternal larvae and found the number half 

 that in the nuclei in the normal larvae, then his conclusion would 

 have carried far greater weight. Since however the size of the nucleus 

 seems to be regulated by the size of the cell containing it, I think 

 his experiment receives no support from the smaller size of the nuclei 

 of the paternal larvae. 



The cleavage of the nucleated pieces is much slower than the 

 cleavage of the normal eggs, and the number of divisions seems never 

 to be carried so far. As late as forty-eight hours one can see that 



