58 ON THE CLASSIFICATION OF BIRDS. 
repeat our opinions on this intricate question: let u 
rather illustrate it by the testimony of others. 
(67.) We have seen that Setophaga and Sylvicola, 
notwithstanding their intimate resemblance, stand in 
two different circles, which circles are connected by the 
subgenus Dumecola: their relation, therefore, accord- 
ing to the views of Mr. MacLeay, are merely analo- 
gical ; not because this relation is less perceptible than 
that between Sylvicola and Dumecola, but because the — 
two groups stand opposite to each other in different 
circles. Now, setting aside the usual mode of dis- 
tinguishing affinities from analogies in larger groups, 7 
it must be at once admitted, that when the differences — 
between two given groups are so imperceptibly gra- 
duated and softened down by intervening species, that 
it is utterly impossible to discover one link that is — 
imperfect in the chain, then such a relation becomes - 
unquestionably one of absolute and direct affinity. 
To demonstrate, therefore, that this is true in regard to 
Sylvicola and Setophaga, we shall cite the authority of 
Wilson, a man totally destitute of theory, but who 
watched and studied these birds in their native haunts. 
Let us take the Sylvicola Americana, the blue yel- 
low-backed warbler of the American Ornithology 
(4 pl. 28. f. 3.), as the best and the most familiar ex- 
ample of the true Sylvicole ; we see the bill ( fig. 133. a) 
compressed on its oe and, although wider at the base, 
still the madi is not arene than its height: e 
bristles of the rictus are short, and do not extend 
