400 BOTANICAL GAZETTE [decemher 



The resorption of the three arrested megaspores takes place 

 rapidly, so that one might easily overlook the stage completely. 

 In A, tuherosa the megaspores farthest from the functioning one 

 seems to be resorbed first {figs* 33, 36). One would suppose 

 that the nearest would be the first to disappear, for in the disor- 

 ganization of the surrounding cells those nearest are the first to 

 change. In A. Sullivantii^ from the cases of disorganization of 

 the third cell wliile the first and second are still in good condi- 

 tion, one gets the suggestion that before the first daughter cell 

 divides it offers greater resistance to disorganization than it does 



fig^ 



e that the first daughter 



cell before its division either aids in the disintegration of the 

 third megaspore or utilizes its space as the fourth absorbs it. 



Among the Asclepiadaccae, therefore, we find one of the 

 strongest arguments for homologizing the row of four cells in 

 the ovule with the tetrad of microspores. It is known (7) that 

 in the formation of microspores each cell of a hypodermal 

 archesporial plate divides, forming a parietal and a primary spo- 

 rogenous cell. The latter without division functions as a mother 

 cell, which gives rise, with reduction of chromosomes, to a row 

 of four microspores, each of which develops a male gametophyte. 

 In the formation of megaspores usually a single hypodermal 

 archesporial cell does not divide to form a parietal cell, but 

 functions directly as a spore mother cell, giving rise, with reduc- 

 tion of chromosomes, to a row of four megaspores, any one of 

 which may develop a female gametophyte. 



But what is the significance of the retardation in the division 

 of the first daughter cell ? No cases of the first preceding the 

 second are reported, and the question as to the origin of this 

 retardation becomes pertinent. In tetrad formation among the 

 pteridophytes the daughter cells divide simultaneously. The 

 possibility of a relation between the retarded division and the 

 suppression of functional activity In three of the four mega- 

 spores is suggested. If such a relation exists, there are three 

 possibilities : 



I. The suppression of three of the four megaspores may 

 cause the retardation in division of the cell forming two func- 



