Olive. Mitotic division of the nuclei of the Cyanopliyceae. 29 
contortecl sliapes by tlie surrounding cyanophycin granules; and 
also to tlie instance given by Raciborski, in whicli tlie nuclei 
of certaiii seeds assume irregulär sliapes, due to the pressure of 
tlie granules of food substances. 
The writer liiay here be allowecl, before leaving tliis subject, 
to give an opinion which has not yet beeil at all definitely 
established in his investigations. It is possible tliat, in the spore 
of Cylindrospermurn shown in hg. 80, the nucleus, which is 
extreniely large in comparison with tliose of the vegetative cells, 
represents in reality several nuclei. Tliere is sonie evidenee tliat 
nuclear division continues in the young, developing spores of 
these plant s, until ab out four nuclei are formed; no wall, liowever. 
separates them. In another species, Cylindrospermum catenatum , 
walls are at once formed, and we have, as a result, several 
spores borne in a chain, instead of one. The writer hopes to 
establish these interesting points more definitely by further 
research. If it be true, however, that there are several nuclei 
in the spores of this alga, then we can readily understand how 
such an abundance of chromatm comes to be present. 
Mitosis in Gloeocapsci polydermatica. 
Gloeocapsa presents a peculiar type of cell division which 
lias beeil, so far as I am aware, nowhere eise observed in tlie 
organic kingdom. This plant seems to have beeil employed for 
study by but few investigators sihce the time when Schmitz 
first, in 1879, thought tliat the granules in the center of the cell 
represented the nucleus, and. later (80), conctuded that he had 
beeil mistaken. Sections of young cells are shown in figs. 62, 
61, 69 and 76, all stained witli iron haematoxvhn. Usuallv about 
eight dark granules can be counted, which, for reasons which 
will be explained later, are called chromosomes in the following 
account, and which are surrounded by an acliromatic substance. 
Minute granules of food material — the ,.slime globules" of Schmitz* 
are abmidant in the peripheral protoplasm. Tlie cytoplasm, in 
the living cell, also contains the diffused pale green coloring 
matter. With in or close beside tlie central bodv mav further 
*/ fj 
occur globules of a substance stainable with haematoxylin; these 
are, in all probability, the ,,slime globules" of Palla (figs. 62, 
61, 66 etc.). In examining a preparation stained witli iron hae- 
matoxylin (which gives much better results in this instance than 
PI emmin g‘s triple stain on account of the staining by the latter 
of the thick, gelatinous wall in which the cells lie), we often 
lind such a spireme-hke arrangement of the chromatin as is 
shown in figs. 63, 68 and 72. It is possible even to distinguish, 
in fig. 72, the individual granules, bound togetlier by the Imin 
substance; wliereas in figs. 63 and 68, the granules of the thread 
can not be easily made out, perhaps because of overstaining. 
In figs. 71 and 75, it is plainly evident that the spireme thread 
has begun to segment at both ends; and in figs. 67 and 70, a 
complete longitudinal Splitting has taken place, for we can now 
