DAA STUDY OF NATURAL HISTORY. 
stated, are seen to put on so many different forms. 
Yet, although we are unable to employ these variable 
peculiarities in a primary sense, they afford ad- 
mirable distinctions of a secondary nature; and 
these, when coupled with the peculiar formation of 
the wing common to the whole group, give us a com- 
pound of characters by which all titmice warblers 
may be distinguished, almost at a single glance, 
from the hundreds of species composing the family 
Sylviade. 
(168.) Now if, in proceeding to the investigation 
of another genus, we find thaé also characterised, as a 
whole, by some one peculiarity of structure; and that 
it also comprehends subordinate forms, more or 
less agreeing with those in the last; we have every 
solid reason to suppose these subordinate forms, in 
both groups, to be analogous; or, in other words, 
to represent each other. To give them, therefore, 
discriminating characters, we unite that of the 
entire genus to that of the sub-genus, as before 
intimated ; and these, collectively, give us a distin- 
guishing formula, by adhering to which we cannot 
possibly err. 
(169.) Of natural groups, Linnzus certainly had 
a very sound theoretical idea, when he said, that 
every genus would furnish its own characters, and 
not that the characters should form the genus; 
thereby implying, that we were first to place objects 
together which appeared closely related, and then 
to discern what were the peculiar and tangible 
characters which made them so. The truth is, 
that, generally speaking, an unscientific person, but 
with a discriminating eye, is much more likely to 
