1835.] ANCIENT INDIAN HOUSES. 257 
In this northern part of Chile, within the Cordillera, old 
Indian houses are said to be especially numerous: by digging 
amongst the ruins, bits of woollen articles, instruments of pre- 
cious metals, and heads of Indian corn, are not unfrequently 
discovered: an arrow-head made of agate, and of precisely the 
same form with those now used in Tierra del Fuego, was given 
me. I am aware that the Peruvian Indians now frequently 
inhabit most lofty and bleak situations; but at Copiapé I was 
assured by men who had spent their lives in travelling through the 
Andes, that there were very many (muchisimas) buildings at 
heights so great as almost to border on the perpetual snow, and 
in parts where there exist no passes, and where the land pro- 
duces absolutely nothing, and what is still more extraordinary, 
where there is no water. Nevertheless it is the opinion of the 
people of the country (although they are much puzzled by the 
circumstance), that, from the appearance of the houses, the In- 
dians must have used them as places of residence. In this 
valley, at Punta Gorda, the remains consisted of seven or eight 
square little rooms, which were of a similar form with those at 
Tambillos, but built chiefly of mud, which the present inhabit- 
ants cannot, either here or, according to Ulloa, in Peru, imitate 
in durability. They were situated in the most conspicuous and 
defenceless position, at the bottom of the flat broad valley. There 
was no water nearer than three or four leagues, and that only in 
very small quantity, and bad: the soil was absolutely sterile ; I 
looked in vain even for a lichen adhering to the rocks. At the 
present day, with the advantage of beasts of burden, a mine, 
unless it were very rich, could scarcely be worked here with 
profit. Yet the Indians formerly chose it as a place of resi- 
dence! If at the present time two or three showers of rain 
were to fall annually, instead of one, as now is the case, during 
as many years, a small rill of water would probably be formed 
in this great valley ; and then, by irrigation (which was formerly 
so well understood by the Indians), the soil would easily be ren- 
dered sufficiently productive to support a few families. 
L have convincing proofs that this part of the continent of South 
America has been elevated near the coast at least from 400 to 
500, and in some parts from 1000 to 1800 feet, since the epoch 
of existing shells; and further inland the rise possibly may have 
