Page:The American Indian.djvu/169

Rh stone tool models, suggesting that the copper-working art had begun a development of its own. The most curious of these is a socketed ax reminding one of certain bronze axes found in Europe (Fig. 83).

No satisfactory evidence of casting or even beating in dies has been found for regions north of Mexico. Ingenious experiments by Cushing and others have demonstrated the possibility of making all the objects so far found here with the most primitive tools. Stefánsson collected specimens from Coronation Gulf, showing long, slender rods fashioned by beating together thin sheets of copper.

A very important problem is that of tracing out the distribution of Lake Superior copper. It is generally believed that most of the copper objects found in Ohio and eastward came from this source, but the source of supply for the Gulf states is not so clear, since some copper was to be found in the Appalachian chain.

Now turning to the region of higher culture, we have the best data from Peru, where ore was smelted in small pottery furnaces into which the operator blew through copper tubes. However, some ores, particularly silver, required greater heat and for such, large hopper-like pottery furnaces were set up on high hills where the wind would create a draft. At the various intakes to these furnaces, fires were placed to heat the air, a mechanism employed by some Old World smelters.

Another question of great interest is the use of bronze in the New World. The fact that implements contain tin, but in varying quantities, has led to the theory that the appearance of bronze is merely accidental or due to the natural mixtures in the ore. This view has received its strongest support from analytic studies of Peruvian copper tools, which reveal a lack of any consistent correlation between the amount of tin contained and the use for which the object was designed. Still, the investigations of Mead have shown that the percentage of tin is often too great to be attributed to natural mixtures in the ore, and those of Mathewson suggest that there is a correlation between the difficulty of casting tools of different forms and the amount of tin they contain.