176 Dr. H. Charlton Bastian on Diatoms and the 
examination, and even then (especially with ZL. gibba) the 
examination can often only be satisfactorily carried out by 
placing one of the leaves in water on an excavated glass slip 
(taking care that its upper surface is uppermost), and gently 
compressing the leaf, if necessary, with the cover-glass. 
An examination of a very large number of these infected 
leaves has enabled me to ascertain the following facts :— 
The very active spores of the Chlorochytrium penetrate to 
some of the intercellular spaces of the leaf through the 
stomata. Single spores or such bodies after a primary 
fission may be seen just within the stomata (Pl. XV. fig. 1, A). 
Sometimes the entire spore or the segments of the once or 
twice divided spore will grow considerably before undergoing 
any further fission (as in B), though more commonly division 
goes on so as to produce eight or more cells (C), which, as 
they grow, soon become tightly packed within the now 
dilated substomatal space (D). Hxamination of the surface 
of the leaf over one of these patches will always reveal a 
stoma greatly dilated and almost circular in shape *. 
The mode of infection in L. minor and L. gibba is there- 
fore altogether different from that described by Cohn as 
occurring in L. trisulea. In that species of duckweed there 
is, curiously enough, an absence of stomata. The average 
shape and appearance of the patches of Chlorochytrium in 
L. trisulca is also rather different from that of the patches in 
the other two duckweeds, and the latter patches also lack the 
distinct and often thick bounding membrane which occurs 
round the patches in L. trisulca. 
In each of the forms the tendency is to an ultimate pro- 
duction of minute spherical or ovoidal zoospores, which, after 
exhibiting a swarming movement, may make their way out 
of the space in which they have been developed. It often 
happens, however, in each of these forms of Chlorochytrium 
that the zoospores may, either in whole or in part, not succeed 
in escaping, but come to rest within their respective cells or 
spaces (fig. 2, B, x 375). 
What I have further to say refers especially to Oh. Knya- 
num, and to this form as it occurs in L. gibba. 
In some of the smaller patches composed only of two or of 
four enlarged cells it may occasionally be seen that segmen- 
tation of the contents of one of the cells only has occurred, 
while others have remained unaltered. This same kind of 
independence in the life of the cells occurs also in larger 
* All the components of figs. 1,2, and 3 have been magnified 250 
diameters except fig. 2, B; this latter, as well as all the components of 
figs. 4, 5, and 6, have been taken at a magnification of 375 diameters. 
