232 
Journal of Agricultural Research 
Vol. II, No. 3 
gathered in the spring of 1913 were found to be somewhat different, one 
of the principal differences being the absence of the finely mottled appear¬ 
ance which is taken on at one stage of the disease by leaves affected with 
P. koleroga . This mottling is caused by aggregations of hold-fast cells 
(fig. 1). In the South American specimens studied such cells are much 
more evenly distributed and when gathered form in smaller, closer groups. 
A further difference is that in¬ 
stead of consisting of numerous 
crowded and, in the older ones, 
overlapping branches originating 
from various parts of the sur¬ 
rounding mycelium, the hold¬ 
fast cells of the Venezuelan can- 
delillo are made up of rather 
small expansions from isolated 
short side branches (fig. 2). 
These fasten the hyphae to the 
leaf in much the same way as 
the appressoria of the powdery 
mildews, to which, however, they 
bear but slight resemblance, ex¬ 
cept in arrangement along the 
hyphae and in function. Still, the 
earlier classification of the can- 
delillo by Dr. Adolf Ernst (1878, 
p. 16) as one of the Erysiphaceae seems less ill-founded if this and not 
P. koleroga , as has been sometimes assumed, were the fungus in question. 
The manner of branching and size of the hyphae are the same, but the 
Venezuelan fungus possesses somewhat thinner or, at least, less con¬ 
spicuous cell walls, and in places masses of unbranched threads running 
in all directions are to be found, 
which are rarely found in the Porto 
Rican fungus. The differences are 
such as might be found in closely 
related species of the same genus. 
Although the specimens studied were 
gathered during the dry season and 
for that reason were not in the best 
condition, they serve to show that 
some small but real differences exist 
between the candelillo of Venezuela 
and the leaf-blight of India and Porto 
Rico. In the original description 
Pellicularia koleroga is described as 
possessing spores, hyaline, echinulate, 
of about the same diameter as the 
hyphae, in which they lie without ap¬ 
parent connection. The fact that spores are lacking in the Porto Rican 
fungus has been taken as additional evidence that it is not Pellicularia 
koleroga. This, however, is a point of little importance. On one leaf of 
the Indian specimens examined, spores were found which agree with the 
original description (fig. 3). They had no connection with the larger hy¬ 
phae, but were seen to be attached to very fine hyphae belonging, appar- 
