56 ON THE CLASSIFICATION OP BIRDS. 
Accentor, since the links connecting all three are nearly 
filled up. Our suspicion, therefore, is merely excited 
by our total ignorance of the manner in which nature 
passes from Trichas into the AmpelidcB, or fruit-eaters ; 
since, if this is the next family, its junction with that 
of the Sylviadce may be supposed to take place at this 
point. On the other, hand it must be remembered that 
many of tlie most indisputable passages or transitions 
in the ornithological system take place through groups 
which are terrestrial ; and, as Trichas is strictly of this 
description, it is contrary to sound logic to doubt that 
which is known, from our ignorance of that which may 
be unknown. 
(65.) The intimate connection between Setophaga 
and Sylvicola enables us, in this place, to illustrate our 
views on one of the properties of typical or external 
groups. We have already had occasion to remark, 
that Mr. MacLeay long ago pointed out the resemblance 
which such groups as these, in two or more different but 
contiguous circles, bear to each other ; and this resem- 
blance he has termed analogy. Now, this term appears 
strictly correct, when the groups in question are very 
comprehensive, anil the intervening forms very dissi- 
milar. If, for instance, we look to the two typical 
divisions of vertebrated animals — as quadrupeds and 
birds — and compare them with the corresponding groups 
of the Annulosa, or insects, we immediately perceive that 
there may be an analogy, but that there cannot exist 
any possible affinity. But how stands the case when 
the groups to be compared are smaller .i" Let us look to 
the circle of the Pariana to illustrate this question. 
(66.) The typical genera, as seen in the annexed dia- 
gram, are Paras and SyMcola ; the subgenera Paras and 
Egithalus in one, and Vermivora and SyMcola in the 
other, arc consequently the external or typical groups. 
Now, in support of our opinion, that what are to be 
termed analogies in larger groups, insensibly change 
their character, and blend into close affinities in smalle 
groups, we appeal to this circle. Is not the resem- 
