372 BOTANICAL GAZETTE [NOVEMBER 
another, but at other times more or less parallel, resulting in differ- 
ent arrangements of megaspores (pl. figs. 23-26). The micro- 
pylar megaspores degenerate in connection with the maturation of 
the large innermost one in forming the embryo sac (pl. fig. 27). 
In my material I observed all of the megaspores disorganizing 
(fig. 11), and it is evident that this disorganization of the megaspore 
which ordinarily gives rise to the embryo sac is one reason for seed- 
less fruit. In such a country as Japan, where many different kinds 
of persimmons are cultivated together, the difficulty of pollination 
is out of the question. 
Summary 
1. Diospyros Kaki is not a dioecious plant, but a monoecious 
one whose staminate flowers are disappearing under cultivation. 
2. The monoecious habit might have been derived from a con- 
dition of perfect flowers; therefore this habit is not a primitive 
character in this species. 
3. Perfect flowers do not indicate the primitive character of the 
variety in which they occur; they appear among other varieties only 
through restoration of lost parts. 
4. The primitive character of Ebenaceae among Sympetalae is 
indicated by the spiral arrangement of petals; the stamen situation, 
although the number of stamens is not definite; and the two 
integuments. 
5. Megaspore formation is also of a primitive character, and 
suggests, along with other characters of the family, that it may 
have some relation to the Myrtiflorae. 
6. There is no parietal tissue in the microsporangium, which 
indicates that Ebenaceae come from some higher family of Archi- 
chlamydeae, because in the lower families parietal tissue usually 
occurs in the megasporangium. 
7. Embryo sac formation occurs in the usual way, and in general 
furnishes no evidence for the evolution of dicotyledons. It is true 
in general, of course, that the gametophytes of angiosperms are 
of less value for evidence concerning evolution than the sporo- 
phytes. 
