1902] A MORPHOLOGICAL STUDY OF ASCLEPIADACEAE ^o\ 



tionless ones. This would presuppose some unknown cause 

 operating" to suppress three, probably the same which operates 

 to reduce the number of megaspores in heterosporous pterido- 

 phytes. It has been suggested that the central position of the 

 fourth megaspore gives it an advantage over its fellows in food 

 supply. If nutrition is obtained from the resorption of the sur- 

 rounding cells and the influx of food stuffs, this claim receives 

 added weight when it is remembered that the fourth megaspore 

 is nearest the point where the later formed vessels end. The 

 retardation would then be regarded as the beginning of the 

 suppression of a useless division, The apparent variation in the 

 relative time of division of the two daughter cells adds weight to 

 this view, and cases like figs, j8 and jg suggest that there may 

 be here a transition between the row of four and of three. If 

 one were to find a case like^^. j<J alone, it might easily be mis- 

 taken for an example of the formation of only three megaspores; 

 but from the size and appearance of the nucleus of the first 

 daughter cell, and from the fact that such stages must have been 

 passed by ^^. jp, it seems clear that the first daughter cell would 

 have divided again. But if retardation were more pronounced 

 in a case like fig. jp, one can readily see that resorption might 

 occur before the first daughter cell divides, and the condition of 

 three megaspores would be established. It is possible that in 

 some such way the reduction in numbers of megaspores took 

 place in those spermatophytes in which onl}' three or fewer meg- 

 aspores have been reported. 



2. The retardation in the division of a daughter cell may cause 

 suppression of some megaspores. This would presuppose some 

 unknown cause for the retardation. The earlier division of the 

 second daughter cell would no doubt give the third and fourth 

 megaspores an advantage over the first and second, but it would 

 not account for the much more frequent predominance of the 

 fourth than of the third unless the relation to food supply were 



+ 



again introduced. 



3. Both retardation and suppression may be due to a common 

 unknowm cause. 



That retardation in division resulted from loss of function 



