Sixty years ago the French traveler, Alcide D’Orbigny, published his important work devoted to South American Ethnography, but confined to that portion of the continent he had visited, south of the parallel of 12° south latitude.[221] His classification was based partly on language, partly on physical traits, and as it seemed simple and clear, it has retained its popularity quite to the present day. He subsumes all the tribes in the area referred to under three “races,” subdivided into “branches” and “nations” as follows:-- 1. _Ando-Peruvian Race._ BRANCH. NATIONS. 1. PERUVIAN. Quichuas. Aymaras. Chancos. Atacamas. 2. ANDEAN. (Antisian.) Yuracares. Mocetenes. Tacanas. Maropas. Apolistas. 3. ARAUCANIAN. Aucas. Fuegians. 2. _Pampean Race._ BRANCH. NATIONS. 1. PAMPEAN. Tehuelches. Puelches. Charruas. Mbocobis. Mataguayos. Abipones. Lenguas. 2. CHIQUITEAN. Samucus. Chiquitos. Saravecas. Otuquis. Curuminacas. Covarecas. Curaves. Tapiis. Curucanecas. Paiconecas. Corabecas. 3. MOXEAN. Moxos. Chapacuras. Itonamas. Canichanas. Mobimas. Cayuvavas. Pacaguaras. Itenes. 3. _Brasilio-Guaranian Race._ NATIONS. Guaranis. Botocudos. In this classification, the distinctions of “races” and “branches” are based exclusively on physical characteristics, and are at times in conflict with a linguistic arrangement. The Botocudos and Guaranis, for instance, are wholly dissimilar and should no more be classed together than the Peruvians and the Tupis; the Saravecas and Paiconecas speak Arawak dialects; and other examples could be cited. When D’Orbigny confined himself to the identification of related tribes by a close scrutiny of their idioms, he rendered valuable service by introducing order into the chaotic nomenclature of earlier writers, as he forcibly points out; but his physical discriminations are of little value.