ON THE CLASSIFICATION OF BIRDS. 
70 
any table of supposed analogies, however ingeniously 
contrived, where this direct relation is destroyed, must 
be radically false. 
CHAP. VI. 
THE DEKTIROSTRES CONTINUED. — THE FAMILY OF AMFELID.^ 
OR CHATTERERS. 
(81.) We now arrive at the second aberrant group of 
the order. It is composed of the Aimpelidaj, or chatterers, 
so named from the loud and monotonous notes of some 
of the species. It is a family more remarkable for 
beauty and singularity than for its extent ; hence much 
difficulty has arisen in making out the subordinate divi- 
sions and the succession of its groups, just as the disjointed 
parts of a machine can seldom be fitted together when 
many others are wanting. The chatterers are very lo- 
cally distributed, for all the principal types belong to the 
tropical parts of Ame- 
rica; one only, the Bom- 
hycilla garrula, being 
found in Europe. They 
are distinguished from 
aU others of the Den- 
tirostreg, by the enor- 
mous width of their gape, 
which in many extends 
beyond the eye, and in 
some is nearly as wide as that of a goatsucker. The 
particular use of this structure is at once explained 
by the nature of their food : they live almost entirely 
on soft berries and small fruits, which, from being 
swallowed whole, naturally requires a very wide passage 
to pass down the throat : a broad bill always indi- 
