DISTRIBUTION OF EUROPEAN BIRDS. 25 
and Western Asia, they will amount only to ten ; leaving 
seventy-two as a marked and very prominent character- 
istic of European ornithology. In further illustration 
of the very limited range of these families, it appears 
that three only, of eighty-five, equally inhabit Ame- 
rica; and that even the identity of one of these (Parus 
atricapillus L.) with & European species (P. palustris 
L.) is more than questionable. 
(33.) The large omnivorous birds of Europe, compris- 
ing the crow and starling families (Corvide, Sturnide), 
appear widely dispersed. Yet, upon the whole, several 
species, and even peculiar genera, are left to characterise 
this portion of the world. We may state their number at 
twenty-one; thirteen of which, or more than one half, 
habitually reside in Europe; four occur in Northern 
and Central Africa ; one —the beautiful rose-coloured 
starling (Pastor roseus T.)— inhabits likewise the table 
land of Asia, and the deserts of Africa ; while three are 
found in America. 
(34.) These details, of the greatest importance to 
our present enquiry, yet tedious, perhaps, to the general 
reader, it becomes necessary to dwell upon, before 
a competent opinion can be formed on European orni- 
thology. The facts exhibited have never before been 
stated ; and they appear sufficiently strong to justify 
our looking to Europe as the principal seat of a pecu- 
liar geographic division of animals. In this difficult 
and somewhat laborious investigation, we have been 
much assisted by the writings of Wilson, Temminck, 
and Le Vaillant; but more particularly by that liber- 
ality which throws the magnificent Museum of the 
French nation open to the use of all scientific en- 
quirers. _ Yet, with all these sources of information, and 
perhaps greater, had such existed, it cannot be supposed 
that inaccuracies may not occur. Such calculations, 
in short, from their very nature, cannot be perfect ; 
for they are founded upon a state of knowledge which 
is ever improving. All we can do, in such cases, 
is to make as near an approximation to truth as 
