Dr. F. H. Hatch — Geological Historij of South Africa. 165 



these northern regions ; and in the Transvaal it is most likely that 

 even the Bokkeveld Series was never represented. During the 

 accumulation of the Dwyka Series the northern part of South Africa 

 was covered with ice, for it is now generally admitted that the 

 Dwyka Conglomerate has been formed of rock fragments, boulders, 

 and mud, which have been carried towards their present site by 

 glaciers moving from a northern mountainous country southward. 

 That the highest portions of these mountains were probably situated 

 in the Northern Transvaal somewhere about the present Waterberg 

 district, is indicated by the distribution of the glacial conglomerate. 



The Dwyka Series as well as the rest of the Karroo rocks appears 

 to have accumulated in a great inland sea which occupied practically 

 the whole of Central South Africa as at present constituted. The 

 southern shore-line of this vast lake extended east and west along 

 the northern margin of the present coast ranges of Southern Cape 

 Colony. To the east it passed into the present Indian Ocean some- 

 where about the Gualana River, returning at Port St. John, whence 

 it stretched north-eastward parallel to the present coast of Natal, 

 the north-western boundary extending roughly along the present 

 course of the Vaal Eiver as far as Vereeniging. The fossil evidence 

 points to fresh-water, or at least to brackish conditions, and the 

 frequent occurrence of false bedding, ripple - marks, sun - cracks, 

 worm burrows, etc., indicates that the water could not have been 

 deep. The Dwyka Conglomerate was deposited partly in this lake 

 by the agency of floating ice ; partly it consists of ancient moraines 

 the accumulation of which slowly followed the retreating ice north- 

 wards. In conformity with this twofold mode of origin, there are 

 two facies of the Dwyka Series : a northern, lying unconformably 

 on an uneven surface (often grooved and polished) of the older 

 rocks, and a southern, resting conformably on the uppermost member 

 of the Cape System. In the Transvaal the Dwyka covering has 

 preserved interesting features of the pre-Karroo land surface. Thus 

 Mr. Mellor^ considers that the valleys of the Elands Eiver, 

 Bronkhorstspruit, and the Wilge River are of pre-Karroo origin. 

 At the beginning of Karroo times they became filled with the 

 conglomerate, and have in recent times been re-excavated. The 

 ■coal deposits, which at Vereeniging and elsewhere in the Southern 

 Transvaal and at Dundee and Newcastle in Natal are found in the 

 Ecca Beds immediately above the Dwyka Conglomerate, show that 

 a luxuriant tropical vegetation flourished in the Ecca epoch ; while 

 the fact, first pointed out by E. J. Dunn,^ that seams of breccia of 

 almost identical character with the Dwyka are interbedded with 

 the coal at Vereeniging, seems to indicate that glaciers were still 

 in existence while the coal beds were being deposited. Professor 

 Edgeworth David ^ has described a similar occurrence in New South 



1 E. T. Mellor : Transvaal Geol. Surv. Eep., 1903, p. 20. 



2 E. J. Dunn, "Notes on the Dwyka Coal-measures" : Trans. S. Afr. Phil. Soc, 

 vol. xi (1900), p. 67. 



^ Edgeworth David, ' ' Evidences of Glacial Action in Australia in Permo- 

 "Carbonil'erous Time" : Q.J G.S., vol. lii (1896), p. 289. 



