R. Kudo 
231 
each becoming the nucleus of the valve cell, and occupies the outermost posi¬ 
tion in the group (Figs. 38, 50-55). The latter also divides into two nuclei, 
each one being destined for the nuclei of sporoplasms and polar capsules 
respectively (Figs. 50, 51). The third division follows quickly. Finally we have 
two nuclei for the capsulogenous cells and two for sporoplasms, which become 
arranged in a space between valve cells (Figs. 53-55). These last mentioned 
two divisions seem to take place in quick succession so that before the com¬ 
plete separation of the daughter halves, the next division becomes partly 
completed. The result is that stages such as are shown in Fig. 38 (in which the 
two sporoplasms and the two capsulogenous cells are still connected at one 
point) were seen quite frequently. This circumstance, I believe, is well fitted 
for the purpose of development of normal spores, allowing time for the valve 
cells to surround and include the other four cells inside the mass. The nuclear 
changes stated above are always seen under normal conditions. 
Various abnormalities are, however, frequently noted. In the first place, 
as already stated, one of the generative nuclei in the trinucleate stage, may 
not form a spore at all, but it may produce a gemma (Fig. 38), or it may not 
divide whilst the sister nucleus divides repeatedly and develops into a spore 
(Fig. 85). In this case, the trophozoite is monosporous. Again one of the valve 
cells and one of the capsulogenous cells, may divide once more, in which case 
a spore with three shell-valves and as many polar capsules, is produced 
(Figs. 84, 101). 
Since the trinucleate trophozoites and their nuclei differ more or less 
greatly in size, the later stages show similar difference in size. But, since the 
generative nuclei in one and the same trophozoite appear to undergo divisions 
with more or less accurate mathematical ratio, the daughter nuclei of corre¬ 
sponding stages in each of the developing stages are similar in size to each 
other. Indeed in the great majority of cases two spores developed in one 
trophozoite are of approximately the same dimensions (Figs. 53-60, 81-83, 
87-89). Two trophozoites shown in Figs. 82 and 83 were seen lying side by 
side in a smear. While the size of the spores differs considerably in these 
trophozoites, the two daughter spores in each trophozoite are of similar size. 
The valve cells which are usually located laterally in relation to the other 
four cells (Fig. 53), become elongated, and surround the other four cells com¬ 
pletely in the form of a deeply staining narrow band (Figs. 54, 55, 57-74). 
During the formation, the nuclear substances break up into small fragments 
and transform with the cytoplasm of the cell into a typical substance, the 
nature of which I have studied elsewhere (Kudo, 1921a). In the formation of 
the spore membrane of L. ohlmacheri, there is no indication that glycogen 
constitutes a part of the spore membrane as was maintained by Erdmann 
(1917) in the case of Chloromyxum leydigi. 
As to the formation of the polar capsules and the polar filaments, there 
have been large numbers of contributions by authors such as Thelohan (1895), 
Doflein (1898), Auerbach (1910), Awerinzew (1909), Davis (1916), Georg6- 
