152 BOTANICAL GAZETTE [FEBRUARY 
Mitosis 
There are two types of mitosis in Rhodochytrium. The first 
type occurs in the earlier divisions of the zoosporangium, while the 
second is found in the last divisions before sporulation. They are 
not, however, to be considered as distinct, for they merge into 
each other. 
No evidence of a reduction division was found. Nowhere were 
nuclei seen in fours, as would be expected after reduction; and 
while the chromosomes are difficult to count accurately, I feel sure 
that their number was approximately the same in the last divisions 
as in the primary mitosis. They are extremely difficult to count, 
however, because they are usually close together and often sur- 
rounded by starch grains. For this reason it was not possible to 
count the chromosomes of as many spindles as would have been 
desirable, nor to insure exactness in the cases counted. In all of 
the cases where counting was attempted, however, the number was 
no smaller than 8 nor larger than ro. 
The assembling of a series of stages of mitosis is an exceedingly 
tedious task. As already stated, coenocytic cysts of any sort are 
comparatively rare. Those in mitosis are of course rarer still. . 
It is doubtful if one cyst in a thousand of those observed showed 
dividing nuclei. The anaphases and telophases are particularly 
difficult to find. It was not possible, therefore, to examine a large 
number of figures of the different stages. But inasmuch as the 
spindles found form a concordant series, it is believed that the 
account given accurately describes the process. 
Mitosis OF THE FIRST TYPE.—The typical mitosis of the first 
type is the division of the primary nucleus, but the second and 
third mitoses are so similar that for purposes of description they 
may be said to be identical. Drawings from all of these have been 
used in the plates indiscriminately, but they may be identified, if 
desired, by the explanation of the plates. 
Spindle formation.—The first indication of approaching mitosis 
consists in the appearance of kinoplasmic fibers among the masses 
of chromatin in the nucleus. The change shown in fig. 48 is S° 
slight that it would hardly have been detected had not the other 
nuclei of the cyst been already far advanced in spindle formation, 
