( 339) 
that such can hardly be the case, inasmuch as no potential 
megaspores were observed nor any cells that could be taken 
as such, upon this assumption. The division into four mega- 
spores occurs with so great a regularity in this type that there 
is hardly a chance for such an error. 
The functional megaspore enters upon a period of enlarge- 
ment and divides to form the embryo-sac in the usual man- 
ner. The definitive embryo-sac is smaller than that found 
in some of the other genera. As the fixation of the material 
in the definitive embryo-sac stage was not very good, no de- 
scription of its minute structure will be attempted. It pre- 
sents about the same general features as are found in the 
embryo-sacs of the other genera studied. The endosperm 
nuclei fuse before fertilization. The antipodal cells are 
noticeable only for a short time. The egg-apparatus con- 
sists of two synergids, pyriform, vacuolated, with striated 
hyaline upper portions and a centrally situated nucleus. 
The odsphere as in other cases presents a much vacuolated 
appearance in its upper portion and a small nucleus em- 
bedded in the thicker protoplasm of its lower portion. 
Preparations showing the process of fertilization were not 
secured. Some preparations showing young embryos were 
obtained, however, and the embryos appear to differ a good 
deal from those of the other genera. A small mass of cells 
is to be seen lying close against the tissue of the nucellus in 
the embryo-sac. The divisions have produced an embryo of 
about ten to twelve cells, of which a part are shown in figure 
20. The first cleavage-planes could not with certainty be 
determined in this case, but they are apparently transverse. 
In the case represented by figure 2¢ the cell-formation of 
the embryo is quite regular, but in most cases the embrya 
consists of a rounded mass in which no definite system of 
cleavages is apparent (jigs. 20 and 22). In figure 20 the 
line aé appears to represent the oldest cleavage-plane, as it is. 
the thickest of the walls there apparent. But in this case no. 
suspensor appears; which is the condition in the majority of 
the Cucurbitaceae. Later embryonic stages were not ob- 
