TABITATS. 239 
and like them are found on decaying vegetable substances, old 
stems of herbaceous plants, dead twigs, wood, stumps of trees, &e. 
The exceptions are in favour of such species as Torula sporendo- 
nema, which is the red mould of cheese, and also occurs on rats’ 
dung, old glue, &c., and Sporendonema Musce, which is only 
the conidia of a species of Achlya. One species of Bactridium 
is parasitic on the hymenium of Peziza, and Hchinobotryum 
atrum, on the flocci of black moulds. 
In the other section of Coniomyvetes the species are parasitic 
upon, and destructive to, living plants, very seldom being found 
on really dead substances, and even in such rare cases un- 
doubtedly developed during the life of the tissues. Mostly the 
ultimate stage of these parasites is exhibited in the ruptured cu- 
ticle, and the dispersion of the dust-like spores; but in Tvlletia 
caries, Thecaphora hyalina, and Puccinia incarcerata, they remain 
enclosed within the fruit of the foster-plant. The different 
genera exhibit in some instances a liking for plants of certain 
orders on which to develop themselves. Peridermium attacks 
the Conifere; Gyninosporangium and Podisoma the different 
species of Juniper; A/elampsora chiefly the leaves of deciduous 
trees; Restelia attaches itself to pomaceous trees, whilst Gra- 
phiola affects the Palmacee, and Endophyllum the succulent 
leaves of houscleek. In cidium a few orders seem to be more 
liable to attack than others, as the Composite, Ranunculaceae, 
Leguminose, Labiate, &c., whilst others, as the Graminacee, 
Ericacee, Malvacee, Crucifere, are exempt. There are, never- 
theless, very few natural orders of phanerogamous plants in 
which some one or more species, belonging to this section of the 
Coniomycetes, may not be found; and the same foster-plant will 
occasionally nurture several forms. Recent investigations tend 
to confirm the distinct specific characters of the species found 
on different plants, and to prove that the parasite of onc host 
will not vegetate upon another, however closely allied. This 
admission must not, hcwever, be accepted as universilly appli- 
cable, and therefore it should not be assumed, because a 
certain parasite is found developed on a special host, that it is 
distinct, unless distinctive characters, apart from habitat, can be 
