rgit] SMITH—CLINTONIA 211 
nuclei two such lumps are formed. Thus the embryo sac now 
contains an upper healthy nucleus and three lumps of chromatic 
material without spongioplasm. The commonest arrangement is 
that of figs. 6 and 7 rather than fig. 5, the healthy nucleus being 
somewhat above the middle of the cell and much smaller than the 
nucleus of the mother cell. 
These successive divisions, especially the second, must be com- 
pleted very rapidly, if one may judge from the rarity of their occur- 
rence in the material. Thus, of about 350 ovules in the stages 
represented by figs. 2-7, more than 225 were in synapsis, and sie 
5 in the second division. 
It can scarcely be doubted that the four nuclei just described 
are the four megaspore nuclei, and that megaspore formation in 
Clintonia differs from the normal type simply by omitting the 
formation of walls and by an earlier beginning of the degeneration 
of the sterile megaspores. 
It is clear that division of the imperfect nucleus. formed by the 
heterotypic mitosis can be of no importance in the subsequent 
history of the embryo sac; yet among 80 embryo sacs of the age 
of fig. 7, there was not one in which the lower nucleus had failed to 
divide. The three sterile nuclei could be found in every case. 
This fact would imply a strong hereditary tendency to a second 
division, such as we might expect to accompany megaspore forma- 
tion; and incidentally it would indicate that the impulse to nuclear 
division must originate in the cytoplasm, since so imperfect a 
nucleus cannot be regarded as capable of exercising any of the 
functions of a normal nucleus. 
A second peculiarity of the embryo sac of Clintonia is its uni- 
polarity. Only two divisions of the megaspore nucleus occur, 
and thus are produced four nuclei which, in their position, relation 
to one another, and later behavior, exhibit the characteristics of 
the four upper nuclei of a normal embryo sac. The change from 
megaspore to mature embryo sac, involving two nuclear divisions, 
requires an interval of about three weeks, and it is therefore difficult 
to obtain karyokinetic figures. That those I have been able to 
secure are in the late phases is probably due to slow infiltration of 
the fixing medium. 
