50 BUREAU OF AMERICAN ETHNOLOGY [boll. 52 



and intense fires were observed wliicli iiad been fed, as seen from the 

 unburned pieces, by large quantities of dried branches of the hard, 

 thorny brush growing in that locahty. One of these fires was on the 

 surface and represented merely the burning of a pile of brush; the 

 other, probably remade a number of times, was in and about a hole 

 in the ground, such as described above, and filled with ashes and 

 bones. Fires of this size are exceptional, and it is in the highest 

 degree improbable that they would ever have been equaled among 

 the aborigines ; yet even here there was no approach to a production 

 of tierra cocida, or scoria. The surface fire acted on the black veg- 

 etal soil; that in the hole principally on the yellowish loess. In 

 both cases there was superficial blackening, and beyond this some 

 obscure reddening of the soil to a maximum depth of approximately 

 lij inches, but no baking to cohesion. All this indicates the improba- 

 bility, if not the impossibility, of the production by the fires of prim- 

 itive man of the tierra cocida, or of scoriae, such as were seen in the 

 Argentine museums or were collected on the expedition. 



A variety of stout grass, growing in big bunches and known, from the 

 sharp, cutting edges of its blades, as paja hrava or cortadera (Gynerium 

 argenteum) occurs in many localities in Argentina. The burning of tliis 

 grass has been reported by Descalzi, Romero, and Amegliino (see 

 Bibliography) as resulting in the production of baked earth and 

 scoria. During the writer's trip to the Puerto San Bias, this grass 

 was found in spots among the dunes south of San Bias, and instances 

 were seen in wliich the dried bunches had been burned. In every 

 one of these cases, however, the heart of the bunch was still ahve 

 and there was no burned earth or scoria. Men on the ground said 

 that only rarely will one of these bunches die, in which case, if fired, 

 it burns out; but no one has observed anything resembhng the 

 tierra cocida or scoriae ascribed to them. Tliis shows that the ordi- 

 nary burning of cortadera is not (at least not commonly) asso- 

 ciated with products such as those under consideration. 



Small particles, and occasionally larger masses of tierra cocida, 

 were found by Mr. Willis or the writer in a number of locahties along 

 the coast from northeast of Miramar to Monte Hermoso, and were 

 relatively abundant in the deposits exposed in the barrancas at the 

 former locality. They occur at dift'erent depths from the surface, to 

 below the sea level at ordinary low tide. The pieces collected are 

 all compact, with the exception of two or three that show on one side 

 a transition to scoria. While there is a general resemblance, they 

 all differ in aspect and weight from the very porous, light products 

 of the burning of the esparto grass, collected by Mr. Willis on the 

 Colorado. 



Large masses of tierra cocida are c^WeAf agones (fireplaces) by the 

 local investigators. One of these, about 3 paces long and propor- 



