24 BUREAU OF AMEEICAN ETHNOLOGY [bull. 52 



in the Pampean terrane up to the more recent formations, where they 

 may be identified by their topographic relations on the present sur- 

 face as having formed in ponds or lakes. The reduction of the iron 

 oxide to which the green color is due is a natural effect of the presence 

 of organic matter that gathers in ponds. Ameghino was the first 

 to recognize their lacustrine origin. In the Barrancas del Norte 

 they are sharply contrasted in color with the dark-brown Ensena- 

 dean, and the eroded surface is direct evidence of an interval between 

 the episodes of deposition. 



The eroded surface of the lower stratum worked out by wind and 

 not by running water no doubt presented more or less extensive 

 hollows in which pools or ponds would form and water-laid sediments 

 would accumulate, provided there was sufficient rain. The deposits 

 are water-laid and appear to constitute evidence of a return to con- 

 ditions of greater humidity than had existed during the episode of 

 erosion. 



The formations which succeeded the lacustrine deposits where the 

 latter occur, or which rest on the lower brown earth where the lacus- 

 trines are wanting, are of two kinds. There is a fine Vvhitish or light- 

 gray stratum which exhibits the vertical structure characteristic 

 of eolian loess and which also possesses the fineness and uniformity 

 of wind-blown dust. The light color is peculiar and is not explained 

 by any field observation. This white or gray loess is locally con- 

 formable to the brown sandy earth and again is separated from it by 

 pockets of pink pebbly loess described below. 



The formation which may be called pink pebbly loess, according to 

 its color and constitution, is a fine-grained, light reddish-brown 

 deposit, Avliich includes small pebbles of the same material. The 

 texture, homogeneity, and structure are eolian. The pebbles also 

 could have been formed only by wind action, since the loess of which 

 they consist would readily melt down in water and lose its form. The 

 formation thus suggests arid conditions. It occurs characteristically 

 in wind-eroded hollows which are more or less undercut, and thus it 

 indicates the activity of the wind as it erodes and fills. Were it not 

 for the intervening lacustrine formation there would be no reason 

 for separating the episode of erosion during which the basal stratum 

 was sculptured from the episode of erosion and filling which is marked 

 by the pebbly loess, and the climatic variation would be simply 

 from more humid to more arid. But the occurrence of the lacustrine 

 deposits in the hollows cut in the basal stratum and the fact that 

 they are themselves sometimes cut out b}^- the hollows filled by the 

 pebbly loess indicate that between two episodes when wind erosion 

 was favored, presumably by aridity, there was an interval of pre- 

 cipitation adequate to produce ponds. 



