276 



THE AMERICAN NATURALIST [Vol. XL 



4. In the smaller portion of such a double pollen grain there is 

 more variation in the size, number, and arrangement of cells, but 

 generally there is a distinct resemblance to the typical antheridium. 

 In several cases the similarity was striking and unmistakable. 



5. In two cases where a cell which is like the one that divides 

 into the stalk and central cell, was plainly marked, the rest of that 

 portion of the pollen grain, which corresponds to the tube cell or 

 antheridium wall, was divided into two cells by an anticlinal wall, 

 that is, the antheridium wall was two celled (PI. 1, Fig. 15). 



(1 In the material examined the proportion of double pollen 

 grains was found to be 2.4 % in a count of 1120. 



Writers who feel justified in labeling a single prothallial cell as 

 the second one, may be inclined to reject some of the conclusions 

 reached in this article, especially the conclusion that a majority 

 of pollen grains in Picea excelsa form only one prothallial cell 

 (p. 271) on the ground that so much of the material studied was 

 abnormal. It has already been admitted that there is room for 

 argument as to the number of pollen grains in which only one 

 prothallial cell is formed, and the point at issue is really the deter- 

 mination of what is the normal number, or to state it more clearly 

 perhaps, the problem is to determine the limits of variation in the 

 number of cells in the pollen grains of Picea, and the relative 

 frequency of the different numbers. Stated in that form, it is 

 apparent that the abnormal, which is only the less frequent, is as 

 much a part of the problem as the normal. The abnormal is 

 only normal variation. Hence conclusions cannot be rejected 

 because based on observation of abnormal structures. 



After consideration of both the historical and original parts of 

 this article it is obvious that variation in the number and arrange- 

 ment of the cells found in the pollen grain and pollen tube is 

 widespread among the families, genera, and species of gymno- 

 sperms. They occur in those species which usually have no dis- 

 integrating prothallial cells, as well as in those which have one or 

 more such cells. These variations cannot be dismissed with the 

 statement that they are abnormal, since the abnormal is only the 

 less frequent modification. It seems to the writer that they may 

 have a significance beyond the mere fact that the individual 

 species vary. 



