124 3 BOTANICAL GAZETTE : [FEBRUARY 
mature spindle has very broad poles and its formation does not 
seem to have been controlled by a centrosome or a centrosphere, 
as a comparison of the spindles of the spore mother-cells with those 
of the cells of the antheridium makes clear. 
The minute chromosomes separate, four going to each pole, after 
which a cell plate is formed in the usual way (figs. 40, 41). The 
daughter nuclei do not come to a true resting stage. The chromatin 
is scattered in almost spherical bodies in the hyaline cavity of the 
nucleus, which do not represent the individual chromosomes, as 
their number and size vary considerably (jig. 42). 
The second division takes place in much the same manner as 
the first. The spindles are arranged with their long axes parallel 
to the first cell plate, so that the cell plates formed in these spindles 
are almost perpendicular to that formed in the first division (figs. 
43-47). The latter does not disappear during the second division 
but remains and the walls separating the spores are laid down here 
(fig. 47). The walls separating the cells of the young tetrad are 
thin and delicate, but in the mature spore the outer layer of the 
wall becomes thickened and folded. ‘The mature spore is almost 
black, and its contents are largely oil. When carried through chloro- 
form into paraffin and sectioned, the spores seem to have only scanty 
granular contents, due to the fact that the oil has been removed in 
the process. The nucleus is very small. 
During the winter and spring following their development, unsuc- 
cessful attempts were made to germinate the spores. It may be 
that they had been allowed to remain dry too long before they were 
moistened, for in nature they would not be dry very long even in 
tiding over a dry season. : 
The spore mother-cells of Riccia natans do not furnish such , 
satisfactory material for study as do those of Riccia crystallina, 
because the large air cavities of the thallus prevent the penetration 
of the fixing fluid and so the spore mother-cells often shrunk. A 
sufficient number of good preparations was secured, however, to 
show that the process of division does not differ from that of Rzccza 
crystallina. 
