272 



THE AMERICAN NATURALIST [Vol. XL 



contained in a pollen mother-cell, the pollen mother-cell wall was 

 cuticularized and expanded into two wings. No indications of 

 this condition were observed. 



A pollen grain with a very different appearance from that of any 

 yet described is represented in Fig. 10 (PI. 1), which is a drawing 

 of the pollen grain whose discovery led to the observations recorded 

 in this article. In this grain we have the typical tube cell, central 

 cell, stalk cell, and what seem to be two disintegrating prothallial 

 cells, but in addition we find lying along the dorsal wall of the pol- 

 len grain a row of four rather prominent cells, each with a nucleus. 

 These cells appear to be of unequal size and of somewhat different 

 shapes. They are separated from the rest of the pollen grain by 

 a wall which was continuous with the inner layer of the wall of the 

 grain, and which appeared just like it in structure. The presence 

 of this wall was made evident by the contracted state of the proto- 

 plasm adjacent to it. 



The structure of this grain was reported at the meeting of botan- 

 ists mentioned above, and it was suggested that possibly the four 

 cells represented a prothallium or gametophyte of a much larger 

 size than usual, and the pollen grain showed a reversion to an ear- 

 lier type of structure. At that time the writer saw no way of deter- 

 mining definitely whether that was the correct interpretation of 

 the structure discovered. In succeeding years, it was found that 

 the class material from which that pollen grain came also contained 

 numerous others similar to it as well as the variations of the differ- 

 ent kinds that are reported in this article. 



After studying a number of grains of this same type the thought 

 occurred that perhaps a clue to the meaning of the unusual structure 

 might be obtained if the same grains could be studied from the dor- 

 sal as well as from the lateral view. Since the pollen grains were 

 mounted in clove oil, this could be done, by causing them to roll 

 over while observing them under the microscope. By repeated 

 trials it was found possible to study some of these abnormal grains 

 from all possible points of view, dorsal, ventral, lateral, and end 

 views. 'I'he dorsal and lateral views were the most important. 



A total of 22 grains of this structure was studied, and many of 

 them were drawn in two or more views for purposes of compari- 

 son. Figs. 11 and 12 (PL 1) are respectively the lateral and dorsal 



