these may be confused with the ancient delta*deposits. Though the plateau is now so 

 dry that rain immediately sinks into the ground, and does not form surface streams, 

 it must not be forgotten that when the climate was arctic, conditions were entirely 

 altered. Then the gravel must have been frozen to a considerable depth and rendered 

 impervious, either for the whole year or until late in the spring. Under such circum* 

 stances the spring thaw and spring rains would cover the plateau with irregular water* 

 channels and pools, and in the deposits of these pools we may expect to find Arctic 

 plants, or the bones of mammoth, Rhinoceros antiquitatis, and lemming, just as they are 

 found in Germany and England. The discovery of such fossils in the surface layers must 

 not be taken as proving that the formation of the plateau itself was of Pleistocene date. 



There has been much discussion as to the age of the gravels of the high terraces 

 and high plateaus, and on the ground that these gravels were thought to be earlier than 

 the first advance of the Alpine glaciers, they have been referred to the Diluvial or 

 Pleistocene period. The question of Alpine glaciations is however of very little import* 

 ance in the lowlands. The extension of the Alpine glaciers depends mainly on two 

 factors which do not here concern us. The first factor is the varying elevation of the 

 Alps at different periods, and this at present we have no satisfactory means of estim* 

 ating. The delta*deposits show no sign of tectonic disturbance in England ; in Limburg 

 and Prussia they show numerous faults; probably in the Alps far greater movements 

 took place. The second, and probably the more important, factor is the rainfall. The 

 glaciation of a mountain chain depends entirely on the accumulation of snow, and this 

 may be quite unconnected with the climate of the plain a hundred kilometres away. 

 The advance of glaciers may be due to a general climatic change; but it depends much 

 more on the angle at which the moist winds impinge on the mountain slopes. The 

 snowline may be found at quite different levels on the two sides of a mountain*chain. 

 In the Himalaya it is far lower (about 1000 metres), and the glaciers are much larger 

 on the moistsswarm south side than on the dry*cold north side. 



A dry*cold period, with a subarctic fauna like that of the loess, will be con* 

 sidered in the Alps, if the strata are unfossiliferous, to be interglacial, not glacial, 

 for the glaciers retreated. But a much warmer moist period, provided the moist winds 

 reached the Alps, will there be considered as glacial, instead of interglacial, for the 

 glaciers advanced. The discussion of the extension of the Alpine glaciers may therefore 

 be left out of the question, and we must judge of the age of the deposits of the plain 

 from their internal evidence, and from their correlation with other strata of known 

 date in the basin of the North Sea. 



Our classification of the Tertiary deposits is founded mainly on the percentage 

 of extinct mollusca contained in the marine deposits. This, owing to the continuity of 

 the oceans, and the variety, good preservation, and great ease with which the mollusca 

 can be determined, gives a means of correlation far more satisfactory than that yielded 

 by local climatic conditions, or by the examination of the fauna and flora of isolated 

 land*areas. In the case of isolated lacustrine deposits it may be necessary to adopt 



