1912] PACE— PARXASSIA 307 



perfect anthers, containing apparently normal pollen, were found 

 on these (fig. 70) . One anther had developed on a staminodium ; 

 two anthers were normal in appearance. Chamberlain (2) has 



somewhat 



ment 



Megaspores. — A complete series in the development of the 

 ovule from the first protuberance was studied. The earliest stages 

 show no differentiation of sporogenous tissue. The first difference 

 to be noted is in ovules that are somewhat advanced; and these 

 show only a difference in the size of the cells, there being no definite 

 arrangement of these and no difference in staining reaction. Cer- 

 tain characteristic groups of these larger cells are shown in figs. 1, 

 2, and 3; the first two are from the same section. Often there are 

 only two large cells, as may be seen in fig. 1, but in this case a third 

 cell is below these two. Fig. 2 has four large cells that evidently 

 were produced from one cell by two successive divisions, thus giving 

 a rather striking resemblance to four megaspores; the wall between 

 the two upper cells is very faint. At this stage, if the section is not 

 perpendicular to the wall, one gets the impression of two nuclei 

 without a separating wall, as Chodat (3) has shown in his fig. 660; 

 but I found no case in which the wall was really lacking. Fig. 3 is 

 an ovule with one of the large cells in mitosis, showing the 20 

 chromosomes of the diploid generation. The ovule is somewhat 

 larger before the difference in staining reaction appears, and the 

 difference in the size of the cells is also more striking (fig. 4). 



The inner integument begins to develop at this time. Fig. 5 has 

 both integuments and the sporogenous cell is in synapsis. Synapsis 

 was not found in younger ovules, but apparently it continues for 

 some time, as it was often found in much older ovules, judging by 

 the development of the integument. Chodat's fig. 660, c (3), is 

 similar to fig. 5, except that he shows the nucleus without other con- 

 tents than the nucleolus. Only a few instances were found in which 

 more than one cell showed both by size and staining reaction the 

 sporogenous characteristics. Fig. 6 gives two sporogenous cells, 

 the cell with the nucleus in synapsis being much the larger. In fig. 

 7 the cells are approximately the same size, but the section was cut 

 so that one cell was unfortunately directly over the other. Here 



