134 REPORT—1863. 
tions which appear to invalidate the generalizations commonly accepted. Every 
ractical cranioscopist is aware that Retzius’s classification of human skulls into 
Necktie and dolichocephalic was applied by that illustrious Swede to the 
arrangement of the great leading South American types. The lamented and de- 
ceased cranioscopist gave, as examples of the brachycephalic type, as exhibited in 
South America, the tribes of Ecuador, Peru, Bolivia, Chile, La Plata, Patagonia, 
and Tierra del Fuego; while the dolichocephalic or long-headed type found its re- 
presentatives in the populations of Carib, Guarani, Brazilian, Paraguay, and Uruguay 
origin. This broad generalized statement of facts still remains the accepted and 
predominant hypothesis. How far is it consonant with the extent of our know- 
ledge on the subject? Those few tribes and nations of South America of which 
any accurate and reliable information exists will be briefly recapitulated in the 
following observations, and especial attention drawn to the destderata which appear 
in our collections. The geographical order will be adhered to, apart from any 
broad generalization which may arise, based on craniometrical classification ; such 
generalizations, e. g., as that of Morton, who divided the whole American races 
into two great families; the Toltecan, comprising the extinct half-civilized tribes 
which have become extinct during a prehistoric period; and the barbarous tribes. 
The latter division was subordinated amongst the Appalachian, Brazilian, Pata- 
gonian, and Fuegian branches. Mr. Blake then proceeded to criticise these types 
in detail. In the first place, he pointed to Colombia; the characteristic type pre- 
vailing amongst the tribes of Venezuela is the Carib. The skull is here markedly 
long-headed, with the parietal diameter less than the longitudinal. The frontal 
bones are strongly flattened ; the zygomatic arches large. Accurate and reliable 
evidence respecting the cranial conformation of the natives of Ecuador is wanting. 
The Cara and the Scyri are unknown. There were several types in Peru; e.g. the 
Chincha type, short-headed; the Chimu type, long-headed, so far as known; the 
Inca or Quichua, short-headed, flattened from before to behind by compression from 
the frontal bone to the occiput. In Bolivia there were the Aymara, long-headed, 
of which few examples existed in our collections; the Titicacan, long-headed, but 
of whom the other physical characters are unknown. In Chile, at the present 
day, the type was long-headed, so far as known. The Anthropological Society of 
Paris has recently prepared a series of queries respecting the physical characters of 
the Chile races, vhs showed the utter want of information on this topic. In 
Patagonia the type was also long-headed, as in Tierra del Fuego, Paraguay, La 
Plata, and Brazil. 
On Celtic Languages. By R. 8, Carnocr. 
The author commenced by stating that, having had an opportunity of readin 
Mr. Crawfurd’s paper before attending the Meetings of the Association, he shoul 
reply to it in“detail. Mr. Crawfurd stated that when between two or more lan- 
guages there was a substantial phonetic or grammatical agreement, they might be 
pronounced cognate. In the next paragraph, however, he laid down a different 
proposition, namely, that the words which most distinctly proved languages to be 
cognate were conjunctions, &c., words, in fact, which could not be constructed. 
He would not quarrel with Mr. Crawfurd for using the term German in describing 
the origin of five-sixths of our English language, when doubtless Anglo-Saxon 
was intended. The Norman element, instead of being one-sixth, probably did not 
constitute a fiftieth part of the language. On the question of grammatical struc- 
ture, he combated the notion that the leading languages of Europe, ancient and 
modern, had all sprung out of a dead language of India, and also the proposition that 
the Siamese was a monosyllabic language, and contended that race could never 
to a certainty be determined by language. It would be considered absurd in a 
man who, having given cogent reasons for not visiting Rome, forthwith started for 
the Holy City. But Mr. Crawfurd, after going to the trouble of arguing that 
the boasted test of agreement in the mere structural form of language is inadmis- 
sible, proceeded nevertheless to compare the Gaelic and Welsh, with the view of 
showing that in point of structure they were entirely different languages. Again, 
after stating that the formation of compound words by the help of prepositions was 
a distinguishing characteristic of Indo-Germanic or Aryan languages, and, amongst: 
