232 BOTANICAL GAZETTE [octosex 
nucleus through all stages of mitosis. However, since Wager 
studied this question only incidentally, and as Berlese gives no 
figures and his account is very brief, a detailed comparison 
of the species is impossible. Berlese reports from twelve to 
sixteen chromosomes, and seems to have been able to count 
them during the fusion of the sexual nuclei. Neither Berlese 
nor Wager give details of the mitosis later than prophase. 
Wager reports that as near as he can estimate there are from 
twelve to sixteen chromosomes shown in the mitotic figures. In 
A. Bit six are found with certainty in some anaphase nuclei, 
and twelve appear in some metaphase nuclei with equal certainty. 
It may be that when twelve are counted the chromosomes have 
already divided, and that they really belong to two rather thas 
to one nucleus. However, this is not certain, and there is some 
evidence that makes it appear that there is a reduction in the 
number of chromosomes during the first mitosis, but this canmot 
be considered as proved. 
Centrosomes were not observed in either A. candidus ot 4. 
Portulacé, or in any other of the Phycomycetes, so far as the 
writer is aware; but they have been described in earlier a8 
for Ascomycetes (Gjurasin ’93, Harper ’95), and for Basidomy- 
cetes (Wager ’92, Juel ’98). a P 
With such fundamental differences as have been indicated, 
is useless to attempt to establish a type of mitosis for pois" : 
or to attempt to determine the relationship of the Aaa gene 
cytology. The type here described for A. Bit, while ae 
ing a few deviations from the mode of mitosis in vogue a 
higher plants, is in no way a departure from the forms ¥ 
known among the lower types of plants and animals. i 
It is impossible to generalize on the facts of mori : 
sented in this paper, because simple processes of mae : 
have been described in all forms where homology ee of 
sought. Until the behavior of the nuclei of on Lee 2 
Albugo is known, it is impossible to say which is the eee 
form of the genus, A. candidus or A. Bliti. From mee in most 
study of A. candidus it can be affirmed that A. Bua oe 
