106 
would at first glance Scarcely be thought to be the work of a Perono- 
spora. The leaves of the host plant are harsh and firm in texture. 
Whether from adaptation to this peculiar host or from other causes this 
species seems to have developed on different lines from most Perono- 
sporacere. It does not fit well into the present scheme of classification. 
The germination by zoospores would suggest Plasmopara, but the 
long, dark-colored conidia are much unlike the typical conidia of that 
genus, which are small, orbicular, and hyaline. The conidia of this 
species are also remarkable for the papillus at the base. The conidffi- 
phores are similar to those of Peronospora, but the nearly straight 
branches and the tendency of the first branch to come out at nearly a 
right angle gives at least a suggestion of Plasmopara. It should be 
remarked that the so-called pinnate conidiophores of Plasmopara are 
not pinnate. The lateral branches are rather short and are arranged 
along a lengthened axis after the manner of the two-fifths pliyllotaxy 
in phenerogams. In the dichotomous forms, as in Peronospora, the 
branches are arranged in the same way except that the lowest branch 
is relatively long and extends upward rather than straight out, but 
there is rarely any difficulty in deciding which is the main stem. They 
are not, then, truly dichotomous. 
The conidiophores of P. celtidis , while of the type Peronospora , may 
be regarded as a step toward Plasmopara. So far as known to me 
no one has studied the exact arrangement of the branches of the 
conidiophores of any species of this order. Outline drawings are 
scarcely satisfactory representations of these objects, because they do 
not clearly indicate whether a given branch extends up or down from 
the plane of the drawing. The figures of the conidiophores given on 
Plate xvn are open to this objection. It is not easy to determine how 
each branch extends, much less to represent it accurately in a drawing. 
The mycelium of this species looks quite unlike the typical mycelium 
of Peronospora. It is much more delicate and somewhat resembles 
that of the Uredinem. The oospore is apparently identical with that 
of the section Effusce. The oospores of the two species represented on 
the plate are strikingly similar except in size. But with all the char¬ 
acters taken into account one would scarcely wish to place P. celtidis 
in the section Effusce. Mr. W. T. Swingle lias pointed out to me that 
Peronospora cubensis , B. and C., resembles P. celtidis quite closely, 
and is its nearest ally, and that these two species form a group by 
themselves, differing considerably from either Peronospora or Plasmo¬ 
para. Both have long, very dark conidia, pointed at each end and 
germinating by zoospores, with conidiophores of the so-called dichoto¬ 
mous type, and strongly hygroscopic. For the present the form on 
Celtis is thought to be best placed in the genus Peronospora. 
