2;62 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1916. 



the Matsap beds of Cape Colony and some of the Kuddapah sand- 

 stones of India, described as shore deposits, or the Vindhian sand- 

 stones and conglomerates should be included is uncertain. 



If these are all of the same age and have been correctly interpreted 

 as arid deposits, this was the most severe and extensive period of 

 desert conditions known. In man}'^ places on the Canadian Shield 

 the coarse red sandstones, usually with some conglomerate at the 

 base, may be seen resting on an Archean surface of granitoid gneiss 

 or Keewatin schist or Animikie slate, the original land surface of 

 gently rounded hills and shallow valleys belonging to an ancient 

 peneplain. In some outcrops the crumbling gneiss beneath, an old 

 regolith, provides most of the materials for the basal conglomerate. 

 This is true at various points on the north shore of Lake Superior 

 and apparently also in Scotland, where the Torridonian rests on the 

 I^ewisian. The Lake Superior Iveweenawan, though much the best 

 known, is on a small scale as compared with the areas of sandstone 

 of the same age farther north in Canada. The Athabasca sandstones 

 of Tyrrell, those of Great Bear Lake and of central Labrador, not 

 to speak of smaller areas, indicate a very broad surface exposed to 

 arid conditions in North America. These red sandstones still occupy 

 I ot less than 50,000 square miles, and it is certain that much greater 

 a reas of such relatively soft and easily attacked rocks have been de- 

 stroyed in the long dry-land periods of later times. 



It appears that in this desert period the arid districts were mainly 

 in the Northern Hemisphere and to the north of latitude 48° — that 

 is, very much farther north than the belt of deserts of the present 

 Northern Hemisphere. It is unknown, of course, to what extent 

 Keweenawan rocks are buried to the south of Lake Superior or of 

 Scotland. The breadth of the belt as known in North America is at 

 least 20°, since rocks of this age reach nearly to 70° north latitude 

 in the region north of Great Bear Lake. The Gaisa beds on Varan- 

 ger Fjord, in Norway, reach the same latitude, and the Scotch Torri- 

 donian about latitude 58°. 



It is hard to imagine red soils, drifting sands, and the hot winds of 

 deserts as existing in regions now tundra-covered and frigid; but this 

 seems to have been true in the more northern areas. . 



GLACIAL PERIODS. 



Thus far arid conditions only have been mentioned, but the best 

 preserved land surfaces of the past are those sealed up unchangeably 

 beneath glacial deposits. It seems absurd to couple together deserts 

 and glaciers, so opposite to one another in every respect; neverthe- 

 less in running down the column of historical geology one finds these 

 GQntr?^(Iictory pheno^l.e^a closely linked togeth^i\ In almost all the 



