No. 472] POLLEN GRAIN VARIATION 



269 



nuclei, and apparently either the stalk cell or the central cell of the 

 typical pollen grain has in this case divided again. There was no 

 indication as to which of the two furnished the third cell found here. 

 Another interpretation of this pollen grain is possible. The outer- 

 most of the three almost equal cells might be regarded as the second 

 prothallial cell. The writer does not accept this interpretation of 

 that cell because of its large size, its prominent and well preserved 

 nucleus, and its close connection with and general resemblance to 

 the adjacent cell. 



A very different type of variation is illustrated in Fig. 8 (PI. 1). 

 Here the pollen grain, while having the typical form externally, is 

 divided into two equal portions by a wall perpendicular to the long 

 axis of the grain. This wall was very plain where it joined the 

 outer wall of the grain, but where the protoplasm lay against it 

 on both sides it could not be recognized certainly, and it was 

 represented in the drawing only where is was unmistakable. Its 

 position is probably indicated by the cleft in the protoplasm which 

 throws five cells on one side and three on the other. This gives 

 the typical form and arrangement of cells with two prothallial cells 

 on one side, while on the other side of the dividing wall no pro- 

 thallial cells are formed, and the three cells present apparendy 

 represent a stalk cell, a central cell, and a tube cell. The writer 

 believes (his dillVreiur in the two halves to be of some importance 

 in relation to the variation in number of prothallial cells. If the 

 munber of those that disintegrate is invariably two for each pollen 

 grain, which ordinarily has one antheridium, in this double grain 

 where there are two antheridia, there ought to be two disintegrating 

 cells for each antheridium. If it be supposed that one set of two 

 prothallial cells serves as the basis of two antheridia. variation 

 from this arrangement and number of prothallial cells is fonii<l m 

 other grains of this same tvpe, containing two antheridia. Fig. ^) 

 shows a grain with the two\antheri(lia in different p<.sitions in their 



thaUial cel/visible. Thirtv-two grains of this .louble type were 

 studied in more or less (h^tail. In most of them the two anther- 

 idia were place<l as in Fig. s. There wa^ nmch variation in the 

 number of the prothallial cells <'onne. ted ^to the various anther- 

 idia. Put in the simplest form and nsing the single antheridium 



