ACTION OF THE SPERMATOZOON UPON THE Eaa 241 
if two spermatozoa enter the egg, not only two but three or 
four astrospheres are formed. When we cause the segmentation 
of an unfertilized egg by treating it with hypertonic sea-water, 
two astrospheres are formed (unless the eggs have been too 
long exposed), and the division of the nucleus takes its regular 
course. But if the eggs are left too long in the hypertonic 
solution, they divide into more than two cells at once when they 
are put back into normal sea-water. According to Morgan, 
this is owing to the formation of more than two astrospheres. 
Such eggs do not develop into normal larvae. 
Why is it, then, that only those eggs develop into vigorous 
larvae in which the first division leads to the formation of only 
two cells? To this question Boveri gives the following answer. 
In normal nuclear division, each chromosome splits length- 
wise into two similar pieces, one of which goes into each of two 
astrospheres, and into the new nucleus; so that, therefore, after 
the division is accomplished, each of the two daughter nuclei 
contains, quantitatively and qualitatively, the same nuclear 
material. But if two spermatozoa enter an egg, then three or 
four astrospheres are formed, and, correspondingly, three or 
four daughter cells. But since each chromosome of the mother 
nucleus divides only into two parts, it is naturally impossible 
that in this case each daughter nucleus will contain a half of 
each chromosome of the mother nucleus. Boveri and many 
other authors assume, and with good reason, that the different 
chromosomes of the nucleus are physiologically dissimilar. 
It wili therefore be apparent that equivalent and fully potent 
daughter nuciei will accordingly result only from regular nuclear 
division with two astrospheres; and that when three or more 
astrospheres are present the single daughter nuclei will not con- 
tain the full number and, as a rule, not qualitatively the same 
nuclear material. 
This hypothesis, then, would also explain why, under cer- 
tain conditions, the superposition of artificial parthenogenesis 
