54 



Heterostylism in Oxalis 



making eighteen possible pollinations within the species, of which six 

 are self- and twelve cross-fertilizations. The six self-fertilizations are 

 always partially or quite sterile. The six " illegitimate " cross-fertiliza- 

 tions ( ? and ^ of different tiers) are likewise infertile, and only the 

 six "legitimate" unions are fully fertile. This partial sterility is a 

 great obstacle in determining the gametic constitution of individuals, 

 as selfed seed is difficult to obtain in any quantity and difficult to 

 raise ■ nor can the " illegitimate " unions be relied on to produce 

 sufficient seed. 



Long_x_Short 



on X LoO<i 



Fig. 1. Showing the aiiaiigeiiient of tlie styles and stamens in Long-, 

 Mid-, and Short-styled plants. The connecting dotted lines 

 indicate the six "legitimate'' forms of nniou. 



Sterility must prove to be an integral part of the problem when 

 it is fully worked out. The relation to the main issue, however, still 

 remains obscure. 



It is significant that the three forms have different average 

 degrees of fertility even when producing their fullest complement of 

 seed. The relative numbers of seeds in Lythrum salicaria can be 

 given as follows: Long-styled 100; Mid-styled 130; Short-styled 81. 

 (Darwin, Forms of Flowers, p. 143.) Also the different degrees of 

 sterility in different self-fertilizations and "illegitimate unions "seem 

 constant. Again, the mid-styled form is more productive than the 

 other two; but not ecjually so tVtr all illegitimate unions. Thus Darwin 

 got as many as 69 5 seeds per capsule in the cro-ss Mid-styled x long 

 stamens of Short-styled, but on the other hand not a single seed for 

 Mid X own from short stamens. 



