376 Kauffman . — Contribution to the 
walls as a character to distinguish, along with the size of the oospore, one 
of his varieties. 
The most potent solutions for inducing antheridial formation are the 
solutions containing potassium phosphate and disodium phosphate (Figs, 
ii, 13, 14? I 5 , 16). In solutions with potassium phosphate the antheridial 
filaments develop completely, i. e. the antheridium is differentiated at the 
apex of the side-branch, and is applied in the normal manner to the 
oogonium, and fertilizing tubes are put forth and touch the oospheres. 
Such a case is seen in Figs. 16 and 13. We have here, then, a combination 
of conditions much more favourable than in the preceding cases. It is seen, 
too, that there is considerable variation under these conditions with reference 
to the source of the antheridial side-branch. If the hypogynous cell is con- 
sidered as a potential antheridium, we would expect that the side-branch 
would naturally arise from it. This is not always the case ; instead, the 
side-branch may arise from below the hypogynous cell, as seen in Figs. 11 
and 13 ; or it may arise from the oogonial branch immediately below the 
oogonium in the absence of the hypogynous cell, as in Fig. 14 a. And, 
finally, as shown in the same figure, b, it may arise from a different filament 
in the usual manner of antheridia of diclinous origin. No cases were seen 
where the antheridial filament or branch grew out of the oogonial wall as in 
Adilya racemosa . 
The question of whether the tendency of antheridial filaments to arise 
from the oogonial stalk or from some other branch of the fungus is of 
sufficient constancy to warrant its use as a specific character seems to 
be answered here (Fig. 14). There is no doubt that the tendency of the 
male filament to arise at a definite point is to a large extent a matter 
of conditions. In N. mixta (H), which I studied, the oogonia were located 
either apically on short branches, or on the ends of long filaments, or 
intercalary. On short-stalked oogonia, the antheridial filaments grew from 
both the same and other filaments, i. e. were both androgynous or diclinal ; 
the oogonia on the ends of long filaments usually were supplied by diclinal 
antheridia, and the intercalary entirely by the latter. Evidently the 
position of the oogonia on the plant conditioned here in some way the 
origin of the antheridial stalk. In N. hypogyna this was brought out strik- 
ingly in the cultures with K 3 P 0 4 , and less so in those with KN 0 3 and 
Na 2 HP 0 4 . Here numerous oogonia occur in which the usual hypogynous 
cell is present, but there are no antheridial filaments developed from any of 
them, nor from the oogonial stalk anywhere ; on the other hand, the oogonia 
are surrounded often very profusely by antheridial filaments with well 
differentiated antheridia, as shown in Fig. 18. Here, then, all the antheridia 
are of diclinous origin, and the fact that my cultures were made from 
a single zoospore goes far to establish the notion that the place of origin of 
the antheridial filament cannot be retained as a good specific character, at 
