BIRDS OF NORTH AND MIDDLE AMERICA. 9 
(sometimes less than half as long), the rectrices (usually 12, but 
sometimes only 10) usually rounded, never acuminate, terminally. 
The Formicariide comprise one of the larger groups of birds 
peculiar to the Neotropical region, but, like several others of the same 
class, wholly absent from the Antillean Subregion. The group is 
well ‘represented in the whole of the extensive territory extending 
from Costa Rica to southern Brazil, but reaches its greatest develop- 
ment in number of species and genera in the great valley of the 
Amazon and the Guianas. Dr. Sclater? recognizes 250 species 
belonging to 33 genera—numbers considerably less than those that 
are actually known at the present time.? 
Among so great a number of species and genera there is, naturally, 
great variation in size and form. The smallest are no larger than a 
Gnatcatcher (Polioptila), while the largest are fully equal to an 
average-sized Jay; some resemble Shrikes, others Thrushes, Wrens, 
Dippers, or other oscinine groups in their general appearance. They 
are nearly all birds of plain plumage, none baving any brilliant colors, 
and most of the species are terrestrial, or nearly so, feeding, as their 
name implies, largely upon ants, though perhaps less extensively so 
than has been supposed; it having been stated by competent observers 
that some species subsist more upon various forms of insect life which 
the immense armies of ants, as they march across the forest floor, 
startle from their hiding place among the dead leaves, etc., over 
which they pass. 
Notwithstanding their mesomyodian larynx, many of the Formi- 
cariide are good songsters, some of them being conspicuous for their 
vocal powers. 
The classification of this group is very difficult, more so probably 
than in the case of any other American family of birds. Indeed it 
may be truly said that even the most recent attempts to present an 
orderly and natural sequence of the genera are very far from satis- 
factory. My effort to bring order out of chaos can be considered as 
only partially successful, but it is hoped that some improvement at 
least has been made in that direction. Undoubtedly better results 
would have been reached had a better representation of the genera 
and species been available; but unfortunately there are very many 
species and several recognized genera which I have not been able to 
examine in this connection. As in the case of many other groups, 
the segregation of the species into genera has been based far too 
much on general resemblance, and too little attention paid to struc- 
tural characters. This may be said of nearly every group of birds; 
but in the present instance it is difficult to understand how certain 
associations could have been made. 
@ Catalogue of the Birds in the British Museum, vol. xv, 1890, pp. 177-328. 
bIn his Hand-List of the Genera and Species of Birds (vol. iii, 1901, pp. 9-45) 
Dr. Sharpe enumerates 347 species and 38 genera. 
