PERMO-CARBONIFEROUS GLACIAL PERIOD 351 



a motion of the ice in general from south to north, as might have been 

 expected; but in various places the ice sheet or sheets reached the sea, 

 large boulders occurring in stratified shale, as if dropped from ice, and 

 marine fossils being found in close connection -with the beds containing 

 boulders. 



In South Africa the Dwyka tillite has been found in all of the British 

 provinces from the south of Cape Colony to the middle of the Transvaal 

 or possibly the southern boundary of Ehodesia, and from Prieska, in 

 Cape Colony, on the west to eastern N'atal. In a direction from south- 

 west to northeast the Dwyka is known to have an extension of about 800 

 miles, and it probably extends considerably farther north. To the south- 

 Avard the Dwyka reaches nearly latitude 34°, according to Mr Rogers; 

 Doctor Molengraaf puts its northern limit in latitude 26° 40V^ but Mr 

 Mellor extends it to a point 90 miles north of Johannesburg,^^ about lat- 

 itude 24° 30', and Mr Mennell speaks of "what is almost certainly Dwyka 

 conglomerates" in the Tuli district of Ehodesia in about latitude 22°.^' 



The direction of the striations so splendidly shown in the northern 

 part of Cape Colony, Natal, and the Transvaal, and of the transport of 

 boulders, is southward — something quite unexpected — and the tillite 

 grows thinner to the north, toward the glacial center, as the Pleistocene 

 boulder clay does in northern Ontario, and thickens to the so\ith, as in 

 the northeastern United States. In the southern part of Cape Colony 

 Mr Eogers estimates its thickness at 1,000 feet, while at the Kimberley 

 diamond mines it is only a few feet thick. Toward the south no striated 

 floor is found under the Dwyka, but the tillite passes downward and up- 

 ward into stratified shale or slate, suggesting that the ice reached the sea 

 and dropped its load of englacial till on the shallow bottom. Marine fos- 

 sils have not been foimd in the shales, however, so that the deposits may 

 have been formed in a huge fresh-water lake. 



The Dwyka now occurs at elevations of 3,000 to 6,000 feet at the sea, 

 probably much higher than the level at which much or all of it was de- 

 posited, since the southern part was laid down in the sea or great mar- 

 ginal lakes which could not have been much above sealevel.^* 



Judging from my own observations, the northern part of the area cov- 

 ered by the great ice-sheet was not mountainous, but rather a peneplain, 

 like the area covered by our Labrador ice-sheet in Pleistocene times, with 

 hills and valleys, but no high ridges. 



" Geology of the Transvaal, p. 74. 



"Quarterly journal of the Geological Society, vol. 61, 1905, p. 689. 



" Science In South Africa, p. 302. 



" See Professor Davis's discussion of the subject, Journal of Geology, vol. xvl, p. 81. 



