BRONZE IN SOUTH AMERICA—DE MORTILLET. 265 
large proportion of tin would have been particularly useful that 
the greatst percentage occurs. Thus, among the bronzes from 
Tiahuanaco, the metal of a knife (No. 3) contains only 5.83 per cent 
of tin, while that of a pin (No. 6) has 7.70 per cent. From Yura, 
a utensil, a very thin cutting ax (No. 7), had in its composition only 
2.10 per cent, while two objects purely for decorative purposes, a 
pin (No. 9) and a plate (No. 10), contained up to 9.30 and 10.72 
per cent. The contrary would be indisputably more logical. It is 
important to note that the three pieces came from the same region 
and belong to the same period. 
More curious still are the observations on this subject which can 
be made from: the Argentine bronze analyses. A knife (No. 18) 
revealed only 3.65 per cent of tin; but 13.69 per cent was found in a 
bracelet (No. 26). The bells, which should have an alloy rich in 
tin, contained only a comparatively small amount—6 per cent in 
two specimens (Nos. 21 and 31), and as little as 3.92 in another 
(No. 19). The only piece of actual white bronze, or bell metal, 
was a decorative object (No. 48) whose proportion of tin was 16.53 
per cent. As for the rest of them, in one and the same category of 
objects, the quantity of tin still varied considerably. 
In regard to the big axes with lateral shoulders (fig. 5), of which 
four examples were analyzed, two of them twice, what do we find? 
First, a specimen of pure copper with no trace of tin (No. 14), and 
the others (Nos. 29, 80, and 28) with 3.34 or 4.40, 5.73, 6.06 or 7.38 
per cent respectively. It is the same in connection with the orna- 
mented disks attributed to the Calchaqui peoples. We have a number 
of analyses of their metals. Leaving out of consideration the excep- 
tional specimen containing 16.53 per cent of tin, discussed above (No. 
48), it is seen that the rest, sixteen in number (Nos. 32 to 47), possess 
tin in quantities ranging in progressive order from 1.57 to 8.67 per 
cent. The average is 3.60. 
There has also been mentioned a brass object concerning which 
something should be said. The appearance of this piece is unique. 
Its pale yellow color and the absence of all traces of oxidation caused 
it to be taken at first glance for a small plate of gold. The analysis 
showed that it was merely a sheet of brass, probably not of ancient 
origin. The metal composing it contains in round figures sixty parts 
of copper and forty of zinc, very nearly the composition of brass from 
which present bronzes of very inferior quality are made. By chemical 
analysis we are thus able in certain cases to tell whether these ob- 
jects are, as regards their age, of a period before or after the conquest. 
Analysis likewise furnishes us with valuable evidence as to the au- 
thenticity or nonauthenticity of certain pieces, just as in the experi- . 
