———# 
ee ee = er 
1901] GAMETOGENESIS AND FERTILIZATION IN ALBUGO 87 
size and much larger than the nuclei found at earlier stages in 
the developing oosphere. 
A study of oospheres after the second division shows the 
potential female nuclei in various stages of degeneration. They 
appear first to lose their chromatin network, the nucleoli and 
membrane persisting longest. Such degenerate nuclei can be 
found in abundance in the stages indicated. Frequently one or 
two rather large and apparently normal nuclei are found in an 
oosphere, which also has its one large nucleus in contact with 
the coenocentrum. In other cases, many very small nuclei are 
to be found in the ooplasm, and one large nucleus near the coe- 
nocentrum. These small nuclei range from the normal size to 
such small dimensions that only the nucleolus can be perceived 
with certainty. It would be impossible to recognize the small- 
est as nuclei, were it not that they are connected with undoubted 
nuclei by a series of complete gradations, and from the fact also 
that they lie in a cytoplasm remarkably clear and free from 
granules. Occasionally these extremely small nuclei are found 
in mitosis. Sometimes I have found groups of supernumerary 
nuclei, and often small ones are found in pairs, as though trying 
to fuse (fig. 38). Since this condition has never been found 
before the opening of the antheridial tube, it may represent the 
pairing of a supernumerary oospheric with a degenerate anther- 
idial nucleus..- 
That these inclusions described as degenerate nuclei are 
really nuclei is established by their structure and behavior. 
That they do not result from the division of a fusion nucleus is 
shown by their presence when the male and female nuclei are 
found lying together, but still unfused in the same oosphere 
(fig. 38). The degenerate nuclei, as would be expected, are 
most numerous immediately after the second mitosis, while the 
one functional nucleus which lies near the coenocentrum is yet 
small. Before the opening of the antheridial tube nearly all 
have disappeared. There is convincing evidence that the 
oosphere is at first multinucleate (figs. 30-32, 34) and eventu- 
ally uninucleate (jig. 36). There is no evidence to indicate 
