of the Nuclei in Peronospora parasitica . 139 
by a process of karyokinesis, into a large number of very 
small nuclei, and some of these are probably used up in 
the formation of the inner wall of the oosphere. 
At about the time that the nuclei are arranged regularly 
in the peripheral protoplasm of the oogonium the antheridium 
begins to develop a fertilising tube which penetrates the 
oogonium, and increases in length until it comes in contact 
with the oosphere just at the time when the cell-wall of the 
latter is formed (Figs. 9 and 13 ,a'). The antheridial tube is 
much elongated, and does not pass directly to the oosphere, 
but takes a somewhat oblique course to one side of it. The 
end of the tube is swollen in a characteristic manner (Figs. 14 
and 15, a'). The swelling is cylindrical and takes up about 
one-third of the length of the tube. It is this portion of the 
tube which comes into close contact with the oosphere. In 
two cases, at least, I have been able to observe an opening 
between the terminal portion of this swelling and the oosphere, 
(Fig. 15, f) ; this opening, however, is extremely small and 
easily overlooked. 
The antheridium contains at this period a large number of 
small nuclei which have been derived from the division of the 
pre-existing larger nuclei. One or more of these nuclei pass 
over into the antheridial tube (Figs. 13 and 14, n), together 
with a small quantity of protoplasm. The oosphere at this 
stage contains one nucleus which exists in place of the two 
pre-existing nuclei, these having probably fused together to 
form the single central nucleus ; at a later stage two nuclei 
are to be observed, and at a still later stage a single nucleus 
only is again seen. The most natural interpretation of this is, 
that a nucleus passes over from the antheridial tube into the 
oosphere, and finally fuses with the central nucleus. I have 
not been able to observe these phenomena directly however, 
and therefore I think the question as to whether fertilisation 
does or does not take place is at present an open one. I may 
mention here that Zalewski 1 has already pointed out that in 
Peronospora one nuclear body only is present at a late stage. 
1 Loc. cit. 
