I 1902] A MORPHOLOGICAL STUDY OF ASCLEPJADACEAE 399 



Both daughter cells and megaspore will be numbered beginning 

 nearest the micropyle. 



The number of chromosomes in the division of the arche- 

 sporial cell could not be determined. In the division of the 

 daughter cells of A. Stilliv a7itii five were counted in several 

 cases, but not with certainty. There are approximately half as 

 many, however, as in the vegetative cells. The two daughter 

 cells do not always divide simultaneously. In A. tiiberosa only 

 one case of this division was observed, and in that they were 

 not widely separated in time, although the micropylar daughter 

 cell was the later to divide, and the spindle in it {Jig- 33) was 

 smaller. In A. Sulliva?itii many dividing daughter cells were 

 found, and in no case were the divisions in the same ovule 

 simultaneous. In some cases, however, the equality of the 

 megaspores leads one to think that this may sometimes be the 

 case. In the cases observed, the second daughter cell was 

 always the first to divide {Jigs. 3g^ ^o). The difference in time 

 appears to vary, if one may judge from the difference in devel- 

 opment of the third and fourth megaspores at the time of 

 division of the first daughter cell. Sometimes the third and 

 fourth megaspores have undergone little change when the first 

 daughter cell divides; sometimes one of the two has considera- 

 bly enlarged, usually the fourth ; and sometimes the third is 

 already largely resorbed {fig^ 39^^ The fourth megaspore does 

 not always function as the embryo sac, but any one of the four 

 may, the relative frequency rapidly decreasing from the fourth 

 to the first. All were seen to form sacs in A. Sullivantii except 



/fT\. and verv nrnh;ihlv that rould be found. The 



{fig- 



first was seen to form the sac in A, tuberosa {fig- 36). Occa- 



sionally one finds two 



{fig- 



that one is finally crowded out is evident from the absence of 

 more than one mature embryo sac in an ovule. The functioning 

 of other than the most deeply seated megaspore has been 

 reported before in several cases, but its frequency in A. Sidli- 

 vantii is rather remarkable. In perhaps lO per cent, of the 

 cases observed the fourth megaspore did not become the embryo 

 sac. In A. tuberosa and A. Cornuti the per cent, is much smaller. 



