[79] At Rancul, in the east, 660 feet of loess overlying red sandstone: at Telen, in the west, 2,800 feet of sand, marl, sandstone and gravel.
[80] Roth claims to have found gravel in the San Nicolas barranca on the Paraná. I have myself found small rounded flints in the clay of the Chaco at Tartagal. But these deposits probably come from the left bank of the Paraná, where the beds of river gravel are considerable.
[81] In Ales Hrdlicka, Early Man in South America (Smithsonian Instit. Bull., 52, Washington, 1912).
[82] Many attempts have been made to classify the Pampean lands, but the results cannot be regarded as final. Ameghino, who is first and foremost a palæontologist, has done a service in showing the futility of these geological divisions based upon the actual surface of the deposits (colour, fineness, etc.). But even palæontology gives rather uncertain results, as it is impossible to recognize and follow step by step the various stages of the movement of the fossils. All the classifications of the Pampean are based upon a study of two groups of sections. The first group comprises the cliff on the right bank of the Paraná from Rosario to Buenos Aires and the coastal cliff which is a continuation of it, with a break from Enseñada at Mar Chiquita to Bahía Blanca. Ameghino has recognized there a thick series of æolian deposits separated by several discordances, the oldest elements of which, at Bahía Blanca, belong to the Miocene. The second group comprises the cliffs which enclose the valley of the Rio Primero above and below Córdoba. Doering and Bodenbender in this case describe two stages of æolian loess, each covered by torrential gravel.
From the study of these sections geologists have drawn certain conclusions as to the movements which have affected the soil of the Pampa and the changes which the climate has experienced. These conclusions have in each case only a local value, and they have not yet been co-ordinated. The majority of the observers, from Doering to Bailey Willis and Rovereto, seem not to have taken into account sufficiently the fact that in the continental formations the most diverse deposits may come next to each other in the same series, according to the particular process of deposition, and that their alternation does not imply a general change in the conditions of erosion.
[83] Certain features of the hydrographic network clearly have the character of having been superimposed: that is to say, the path of the watercourses has been bequeathed to the actual plain by former erosion-surfaces, which have now disappeared, on which the valleys were originally imposed. That is why in the district of the confluence of the Colorado and the Chadi-Leuvu the valleys pass from Pampean deposits to the crystalline sierras, which were at one time entirely covered with water.
[84] In the vicinity of San Luis and Córdoba the hard strata which are called tosca are beds of eruptive ashes.
[85] The surface of the tosca tableland is further punctuated by a great number of closed depressions of various depths: long tunnels (dolines) which can only be explained, apparently, as an effect of the dissolving of the limestone by water.
[86] Outside the districts with quick and dead dunes, a frequent type of landscape on the Pampa is a plain thinly sown with very small lagoons, generally circular, between which develop a series of barely perceptible undulations. The inequality is at times so slight that one only notices it by the contrast between the vegetation of the lower and the higher ground. This type of landscape, which is especially seen in the district of Lincoln or of Nueve de Julio, is due to the action of the wind on a plain where the level of the underground water is near the surface. This level marks a limit below which æolian erosion does not take place: a sort of base-level. The periodic variations of level of the underground water reduce or enlarges the undulations of the surface.
[87] Ensayo de la historia civil del Paraguay, Buenos Aires, y Tucumán (3 vols, in 16mo) Buenos Aires, 1816, t. iii, p. 214.