INVESTIGATION OF INSECTS. 567 
insects, may also still be consulted; and thus many 
mistakes rectified, which would otherwise greatly mis- 
lead 3 . 
Though sometimes the limits that separate good spe- 
cies appear at first very slight, and require a practised 
eye to catch them, yet it occasionally happens that con- 
siderable apparent differences may safely be disregarded. 
The colour of insects, — to which unhappily for want of 
better characters we are so generally forced to have re- 
course, — though usually constant, is in some species very 
variable 15 . This is the case sometimes with whole colours. 
Thus Carabus arvensis, Pcecilus cwpreus, &c, are some- 
times of a copper colour ; at others, resemble brass ; at 
others, they are green or blue, and even black. The 
colour of spots also often varies. In some individuals 
of Pentatoma oleracea they are pale, and in others red. 
The number and shape of spots are also often incon- 
stant. Many of the species of Coccinella so abound in 
these variations, that nothing short of the most careful 
examination can enable you to distinguish the species 
from the variety. Insects vary also in size : but as this 
is never assumed as a specific character, it will not oc- 
casion you much trouble. Where the difference in this 
respect between two specimens is very great, the pre- 
sumption is that they are specifically distinct. Diffe- 
rences in sculpture and proportion do not always indicate 
different species ; this being sometimes, as we have seen' 
a It may not be amiss to mention a few : — Sphceridium dytiscoides 
is a Ilj/drophihis related to H.fuscipcs. S. glabralum is heteromerous, 
probably one of the Helop'u Latr. Carabus retusus and Madera: both 
belong to Calosoma. Cistela angustala is a true Cholcva. See Linn. 
Trans, xi. K58.— S. b See above, p. 406. 
