James Geikie — On Changes of Climate. 261 



traders tad they existed even in small numbers. They are rare, 

 also, or altogether wanting, in Canada ; but in the valley of the 

 Mississippi the bone-licks are well known as most extensive, and 

 furnishing the remains of a different series of extinct quadrupeds." ^ 

 In Michigan, which is fairly within the glaciated region of North 

 America, mammalian remains are only met with in what appear 

 to be interglacial deposits : at all events, the deposits referred to 

 overlie and are covered by glacial accumulations. The great " bone- 

 licks," to which Sir J. Richardson alludes occur beyond the southern 

 limits of the " Northern Drift." 



Thus it will hardly fail to strike one as remarkable, that remains 

 of the extinct mammalia are either altogether absent from, or very 

 sparingly present in, regions which give evidence of having been 

 subjected to more or less intense glaciation, or are covered by deep 

 accumulations of the later glacial drifts. In Britain, Italy, and 

 Switzerland alike the old ossiferous alluvia, when traced from the 

 low grounds to the mountains, disappear as soon as the moraines 

 and " alpine diluvium " are reached. Nowhere in morainic turbaries 

 or alluvium which can be demonstrated to be of postglacial age do 

 any traces of the extinct pachyderms appear. But these, when they 

 do occur in glaciated or drift-covered regions, are invariably 

 embedded in infraglacial or interglacial deposits. It is so in 

 Scotland, and (if the Jara-wall be one of the asar) in Scandinavia 

 also. The same rule seems to hold good with respect to Asia and 

 North Amenca. The great plains of Siberia never could have 

 nourished glaciers. We cannot conceive that even during the most 

 intense cold of the glacial epoch, conditions similar to those which 

 characterized Scandinavia and Scotland could have existed in 

 Northern Siberia : the absence of high-grounds and the comparative 

 dryness of the climate must have prevented any accumulation of 

 glacier-ice. Nor can I learn that marine deposits, similar to our 

 boulder-clays and esker-drift, cover any portion of Northern Asia, 

 If cones and mounds of sand and large erratics,^ like those of North 

 America, occurred in Siberia, travellers would hardly have failed to 

 mention them. But all this is changed when we pass into the cor- 

 responding latitudes of North America east of the Rocky Mountains. 

 There the observer encounters the marks of glaciation everywhere — 

 everywhere, too, are great deposits of clay and boulders, mounds and 

 ridges of sand and gravel, and huge erratics. And all over this 

 wide area, down to the borders of the United States, the extinct 

 mammalia never appear in any postglacial deposits. In the neigh- 

 bourhood of the great lakes they occur in freshwater clays, along 

 with abundant vegetable remains, and these clays are overlaid by 

 glacial beds. It is only when the southern limits of the " Northern 

 Drift " are approached, that the extinct mammalia begin to be found 



1 Journal of a Boat Voyage through Rupert's Land, vol. ii., p. 210. 



* Middendorf told Sir C. Lyell that he had observed erratic blocks in strata of 

 clay and sand at about fifteen feet above the sea in lat. 75" 15' N., near the river 

 Taimyr. (Principles, vol. i., p. 185, tenth edit.) But these erratics were probably 

 carried down by river-ice. 



