of Peronospora parasitica. 271 
The changes just described differ from those observed in 
Cystopus. In C. candidus the protoplasm first becomes 
separated into a peripheral vacuolar portion and a central 
denser protoplasm containing all the nuclei and smaller 
vacuoles. Then the central mass separates into a vacuolar 
ooplasm and a non-vacuolate granular layer of periplasm, 
into which all the nuclei pass. Stevens describes a somewhat 
similar phenomenon in C. Bliti 1 i where the protoplasm of the 
young oogonium is irregularly vacuolate. In the process of 
differentiation of the oosphere accumulations of denser proto¬ 
plasm are found, separated from one another and the wall 
of the oogonium by vacuoles of varying sizes. These denser 
protoplasmic masses coalesce and the vacuoles become 
restricted to the periphery, so that we get, as in C. candidus , 
a central denser mass of protoplasm surrounded by a vacuolar 
layer. The nuclei pass to the periphery of the central mass, 
and then the periplasm becomes distinctly differentiated as 
a dense granular layer around and between the nuclei. The 
condition in which there is a distinct line of demarcation 
between periplasm and ooplasm, but no definite wall between 
the two, is called by Stevens ‘the stage of zonation’ (p. 157). 
As the nuclei pass into the periplasm they undergo changes 
which lead to division. They increase in size and the chro¬ 
matin-granules fuse together to form the chromosomes, deeply 
staining granules, which accumulate in the equatorial plane. 
A spindle is formed, the nuclear membrane disappears, and 
the chromosomes become divided into two groups which 
separate to become the daughter-nuclei. No nucleolus is 
visible during the mitosis and no definite centrosomes, although 
occasionally granules at the poles of the spindle were observed 
which might have been taken for such. The nuclear division 
described by Stevens in C. Bliti agrees in its essential 
details with this, except that the nucleolus persists until 
a late stage in the division and the nuclear membrane does 
not disappear so soon. The nuclei of the antheridium appear 
to divide simultaneously with those in the oogonium. 
1 loc. cit. 
