238 
THE SUBSTITUTE. 
disease brought on by this propen- 
sity, which speedily causes their 
death. 
All the Tceniocampa do not live 
on trees, but those which do not 
are the exception. Gothica and 
gracilis are in this case, but the 
perfect insects do not differ from 
the tree-feeding species, and we 
cannot always base the establish- 
ment of groups on this difference 
of life. 
The perfect insects appear ge- 
nerally in early spring, and fly at 
dusk around the sallows, from the 
blossoms of which they obtain an 
abundant repast. It sometimes 
happens that they remain there 
during the day, and then we can 
cause them to fall by shaking the 
branches, but more frequently they 
retire to the trunks of trees or 
amongst the herbage ; but the 
best means to obtain them in any 
quantity is to rear the larvaa, 
which, as I have said, are very 
easy to find. The sexes differ 
little : some species vary exces- 
sively, and this has caused the 
creation of a number of separate 
names. (Vol. I., pp. 346, 347.) 
Anchocelis. 
This genus, very homogenous, 
is quite as valuable as the other 
genera of the family, and even 
better defined, and easily recog- 
nised at the first glance. The 
larva) all live on low plants, and 
conceal themselves during the day 
under the leaves nearest the 
ground. They are of rather dull 
colours, marbled or dusted with 
brown or reddish on a green 
ground-colour; the anal plate and 
the plate on the second segment 
are well marked. It is in this 
genus that we find a very curious 
exception ; it is the larva of lu- 
nosu, which lives concealed under 
stones like an Agrotis, and of 
which the spots are warty and 
shiny entirely as in the Agrotes^ 
whereas the perfect insect has no- 
thing to distinguish it from the 
other species of Anchocelis. These 
exceptional manners would fur- 
nish a powerful argument to the 
adversaries of a natural method, 
if a single fact could invalidate a 
thousand others, and if we were 
not here to apply the universal 
maxim, the exception proves the 
rule. 
In a perfect state the species of 
this genus are known at once by 
the pale nervures of the wings and 
'the reniform stigma being con- 
tracted, and as it were strangled 
in the middle. They keep their 
wings inclined at a considerable 
angle, and do not differ in their 
manners from the other Orihosidce. 
It is in the autumn that they are 
met with, and often when the sea- 
son is far advanced ; some of the 
species vary excessively, and this 
has given rise to the creation of a 
number of pretended distinct spe- 
cies. (Vol. I., p. 363.) 
Cerastis. 
This is another genus recognised 
at a glance, and it is unnecessary 
to insist on its characters. The 
larvae of Cerastis are elongate, cy- 
lindrical, but attenuated in front; 
of a brown or reddish colour, and 
with the plate of the second seg- 
ment well marked. They feed on 
low plants, and concealed as in 
Anchocelis, but only when grown 
up ; when young they prefer the 
young shoots of trees, and it is not 
