1 



362 BOTANICAL GAZETTE [November 



this case none will be formed, for all of them 



signs of disintegration. 



EMBRYO 



The embr}^o is very simple. As stated above, the fertilized egg 

 may divide before the primary endosperm nucleus, which seems to 

 be the more common behavior in my material; or the primary endo- 



F 



sperm nucleus may divide first (figs. 51-55)' An inconspicuous 

 suspensor of one or two cells is formed, but no other differentiation 

 of regions could be made out in the younger stages {figs. 55^ 62). 

 The number of chromosomes in the segmentation of the egg seemed 

 to be twenty-two (figs. 51, 52) y and this number was counted several 

 times in various parts of the ovule (fig. 14). 



DISCUSSION 



name 



resulting from the division of the megaspore mother cell. "Daughter 

 cells'' has been used in this paper to designate them. It has been 

 customary to call the cell that dcA^elops the embryo sac a megaspore, 

 and this is correct in all cases when the sac is organized from the 

 chakzal cell in the "row of three" or from one of the cells in the 

 "row of four.'* But if the name is applied to one of the "daughter 

 cells'^ there is confusion, for in this case the reduction process is not 

 complete. 



Megaspore. — In angiosperms two types of embryo sac have been 

 recognized: (i) the more common type, in which the mother cell 

 gives rise to four megaspores, one of which develops the sac; (2) the 

 Lilium type, in which the mother cell, "functioning as a mega- 

 spore,'' forms the sac directly. In the former case, the occasional 

 "row of three" does not change the situation, for the functioning 



I* 



megaspore is derived from the mother cell by two successive divi- 

 sions (Coulter and Chamberlain i6). 



The production of only two "megaspores" ("daughter cells") 

 by the mother cell has been reported for a few plants that usually 

 form megaspores in the normal way. Schnie\vind-Thies (40) 

 reports two or four in Galtonia; Guignard (25) two in Agraphis, 

 Ornithogalum, and Coramelina; Mottier (31) two in Arisaema 

 (in one instance with no wall between them, which would present the 



I 



I 



