1905] SHATTUCK—ULMUS AMERICANA 213 
of the mother-cell gradually breaks down and by March 1 the micro- 
spores are rounding off; many of them have formed their tube and 
generative nuclei. These are at first very much alike, but the tube 
nucleus soon becomes larger and stains more deeply (jig. 10). At 
this time the two coats can be clearly distinguished, the exine having 
acquired an uneven, reticulate surface and showing five very dis- 
tinct openings (figs. ro-11) through which the intine can be seen. 
The division of the generative nucleus was observed March 23, or 
before the dehiscence of the sporangia (fig. 12). At this time the 
tube nucleus shows signs of disintegration which is completed by 
March 26. The tube nucleus was often found disintegrating when 
the pollen tube was just starting, and was never found to leave the 
pollen grain, and in my judgment it never does so in Ulmus ameri- 
cana. Perhaps this is due to the fact that the pollen tube is not 
more than 3™™ in length. 
It was definitely determined that the male structures are cells, 
and not merely nuclei, the delicate limiting membranes being clearly 
made out. During a large part of their existence in the pollen grain 
these lenticular cells are attached to each other by their adjoining 
ends in such a manner as to make them appear in longitudinal sec- 
tion as if astride of the tube nucleus (fig. 13). WYLIE (21) has 
shown that in Elodea the male cells are attached in a similar manner. 
FEMALE G TE. 
_The mother cell does not form the usual tetrad, but functions 
directly as a megaspore (fig. 23). This condition is well known in 
many angiosperms. While no instance of more than one mega- 
pi was found, the fact that there are frequently two embryo 
“acs in the older stages at once suggests the possibility that the 
Megaspore mother cell in Ulmus may yet be found like that of 
oo a closely allied form, to vary in the number of megaspores 
orms, or possibly to form occasionally two mother cells. This 
Would account for the double embryo sacs (figs. 55-56). 
ve - early part of February the megaspore shows only slight dif- 
htiation, being but little larger than the adjacent cells (fig. 23): 
ee nucleus however is quite large, deeply staining, and begins to 
W signs of preparation for division. 
