49, THE JOURNAL OF BOTANY 
and acute, and not usually more than three times as long as 
broad, often much less. 
The other species of the section with conical acute ultimate 
cells, een the exception of N. Sec opts — be roughly differen- 
tiated from N. Dizonii as follows:—N. plumosa by the fruits being 
produced in the base of the ea rls, by the smaller oospores 
showing less numerous ridges, by the branchlets Bane less divided 
and having the more uneven Tolypella-like development of the 
rays; N. diffusa by the less frequently divided branchlets with 
less numerous rays at each gree oy the shorter end-segments 
with more graduated cells, and the oospore showing less 
numerous ridges; N. hu lieste hy the larger pear-shaped 
antheridia, by the shorter not whip-like end-segments, and by the 
antheridia and fruits being produced in small mucous heads; 
N. cristata by the very olga ie ridges of the oospore, the 
by the amnail fruiting heads enveloped in mucus, on 
branchlets, and many other characters; N. Hookeri pay being 
moncecious, having fewer rays at each forking and greatly 
eaep itr lower os ee the end- segments, the other cells forming 
ro; N. bon 
8 
mianidselo us, by ving roe subglobose fruits showing fewer 
on 
antheridia, fewer ridges owns on tis 6 oospore, aa the aati 
cells of the end-segments are often longer. Were N. Dixonit 
panels: the differences would perhaps be considered hardly 
sufficient to separate it as a species, and in ment natural arrange- 
ment it feet be put alongside N. ornithopo 
Bra o doubt rather as a matter of convenience and only 
jdovintoally, divided his sections of the larger genera of the 
haracee into moneecious and diwcious species. This is clearly 
an artificial arrangement, as there is often closer relationship 
character. In the genus Chara, for instance, nd 
plants as C. Rabenhorstit with fruits and antheridia enbddasd 
at different nodes of the same branchlet, and with the larger 
antheridia characteristic of dicecious species. A very inter- 
step in this direction. In it the fruits and antheridia are some- 
times produced on separate Sipe of the same whorl, while 
the antheridia are very large, the diameter being twice that ns 
normal C. vulgaris. In Tolypella a still further stage appears 
be reached, as Dr. Prosper, in his admirable work on the Beanish 
