of the Nuclei in Peronospora parasitica . 137 
take place leading to the formation and ripening of the 
oospore. 
This description may perhaps be looked upon as a simple 
expression of the phenomena in question. It does not, 
however, represent what in reality takes place. The formation 
of the oosphere is preceded by changes which take place 
in the nuclei both of the oogonium and antheridium. They 
become much enlarged, and, at the same time, the chromatin 
becomes distinctly visible in the form of threads, and in some 
cases a loop-like structure (Fig. 2, 1 ) of the threads can 
be observed. At the same time changes take place in the 
protoplasm. Numerous vacuoles appear in it (Fig. 6 , v ) ; 
and these become much larger, so that at a later stage nearly 
the whole of the central space is taken up by them (Fig. 7). 
The protoplasm thus becomes almost entirely restricted to 
the periphery of the oogonium, a small quantity only of 
granular protoplasm remaining in the centre connected with 
the peripheral portion by a few thick strands. The nuclei 
also pass over to the periphery, the last few passing along 
the protoplasmic strands. The nuclei at this stage exhibit 
a very distinct nuclear membrane and present an appearanee 
very much like that of the ‘ dense skein ’ stage in the higher 
plants. The nuclei take up a position in the protoplasm 
so as to form a very regularly arranged layer, on the inner- 
most side of it, next to the central space (Fig. 8 ). The 
nuclei enlarge still more and become slightly elongated in 
a tangential direction. The chromatin threads become more 
distinct and of approximately the same size (Fig. 2, 2 ). 
They are scattered irregularly in the substance of the nucleus, 
which may now be described as in the £ loose skein 5 stage. 
The nuclear membrane now becomes invisible and the chro- 
matic threads appear to lie loosely in the protoplasm. The 
chromatic threads next become arranged longitudinally in 
the equatorial plane of the nucleus (Fig. 2, 4 ). In some 
nuclei a slightly stained, somewhat cone-shaped mass may be 
observed at each end of the nucleus (see Fig. 9, $). This 
may be a spindle-figure, but it is much too indistinct to be 
