ON ETHNOLOGY. 
253 
Caribbean family, including the Chaymas, Tamanace, &c., is almost equally 
extensive. But besides these great families, there are spread over the whole 
country a number that cannot be estimated of petty tribes, who are generally 
supposed to have distinct languages. 
III. The Midland nations arc also very numerous, and only a few of (heir 
languages arc known to bo common to several tribes. Thev arc distinguished 
by their physical characters from the races comprised in the other divisions, 
as are the latter from each other. Tlie herculean tribes of the great plain of 
Chaco, and the equestrian Alupones, are of much greater stature than the 
diminutive Guarani, and th^' have not the retreating fort-head and the ob¬ 
liquely-placed eyes which give to the Eastern tribes a certain resemblance to 
tie Kalmuks and the Chinese. The complexion of many of the Midland 
Rces is a dark bronze colour, while that of the Guarani and the. Caribbees 
and other eastern tribes is a light copper colour, approaching in some in¬ 
stances to white. 
Of the languages of the Midland tribes wc have little information. Those 
of the Eastern nations of South America have been carefully enumerated 
and exemplified by vocabularies by Gilii and Horva.«. 
On the whole, it must be confessed that wc have much less Information 
respecting the ethnology of South America, through the medium of re¬ 
searches into its languages, than of that of almost any other part of the 
world. I have grouped these races according to their geographical rela¬ 
tions and physical resemblances, nearly according to the classification of 
M. d’Orbigny. 
The outline which I have sketched Is merely designed to furnish an idea 
of the resources which the history of languages is capable of contributing for 
the promotion of ethnological inquiries. 
I should be led far beyond the limits which I have prescribed to this 
paper if I were to attempt to deduce any general results from the different 
lines of inquiry which it has been my principal object to point out as fur- 
Dwhing resources for ethnology. But I may venture to remark, that with 
the increase of knowledge in every direction, we find continually less and 
reason for believing that the diversified races of men are separated from 
fach other by insurmountable barriers*. 
* It is with much gratification that 1 find this to be the ultimate conviction of the irreat 
Mtbor of Kosmos. 
