October 31, 190/J 



NA TURE 



665 



The next step was the tracing bv Mr. K. G. Holmes 

 in iqo4 from the Transvaal into Bechuanaland of the 

 Black Reef, Dolomite, and Pretoria Series (Trans. 

 Geol. Soc. S.A., vol. vii., p. 130). He followed 

 these well-known Transvaal formations near enough 

 to the original localities of the Keis, Campbell 

 Rand, and Griquatown Series described by Stow to 

 suggest the correlation which was proposed in the 

 following year by Hatch and Corstorphine in their 

 "Geology of South Africa" (p. 311), namely, in the 

 following manner : — 



Griqualand West Transvaal 



Matsap Series =Waterberg Series 



(unconformity) (unconformity) 



Griquatown Series = Pretoria Series 

 Campbell Rand Series = Dolomite Series 

 Keis Series = Blacl< Reef Series 



Mr. Rogers appears to doubt whether Stow, bv 

 his Keis Series, referred to the quartzites which 

 succeed the Campbell Rand Series in downward suc- 

 cession. He therefore substitutes the Transvaal 

 name — Black Reef Series — and wdth this modification 

 accepts the correlation ; but he points out that the 

 Matsap and Waterberg Series present marked points 

 of difference, especially in regard to their volcanic 

 rocks, which are of a more acid character in the latter 

 than in the former. 



The Matsap Series is subdivided by Mr. Rogers as 

 follows : an upper group of quartzites and sandstones, 

 developed in the Langebergen proper; a middle group 

 of quartzites, lavas, and fragmental rocks of volcanic 

 origin ; and a lower group of quartzites, slates, and 

 conglomerates forming the foothills east of the Lange- 

 bergen, north of Pad Kloof. The basal bed con- 

 tains many boulders and pebbles of quartzite, quartz, 

 iind red jasper in a quartzite matrix which is some- 

 times highly ferruginous. 



In the Transvaal, Mr. Mellor has recently shown 

 (Trans. Geol. Soc. S.A., vol. x., p. 44) that 

 \vhile the base of the Waterberg System is often 

 marked by the presence of coarse conglomerates, in 

 the Middelburg district a series of acid and inter- 

 mediate lavas, interbedded with tuffs, agglomerates, 

 shales and sandstones, underlies the usual sandy types 

 of sedimentation. The whole volcanic series attains 

 a thickness of approximately Sooo feet. 



The Griquatown Series is divided by Mr. Rogers into 

 an upper group, consisting largely of slaty rocks, 

 together with some brown and red jaspcroid rocks and 

 thin beds of chert and limestone; a middle group, con- 

 sisting for the main part of the Ongeluk volcanic 

 beds, together with some banded jasper beds ; and a 

 lower group, comprising banded jaspers, quartzites 

 and mudstones. The conglomeratic rocks that occur 

 at or near the top of the lower group in the Hay 

 district contain " striated and flattened pebbles and 

 boulders," which, according to Mr. Rogers, " certainly 

 owe their characteristic shape and scratches to glacial 

 action " (Rep. for 1905, p. 162). Seeing that 

 these beds belong to a geological period con- 

 siderably older than the Dwyka Glacial Series (they 

 occur at least 15,000 feet lower down in the succession 

 — even more if the Matsap Beds are considered to be 

 older than the Table Mountain Sandstone), this is an 

 interesting contribution to the fast-growing history of 

 glacial action in the remotest periods of geological 

 time. 



The Campbell Rand Series occupies practically the 

 whole of the Knap plateau. As in the Transvaal, this 

 formation consists mainly of dolomitic limestone with 

 which cherts are often associated. The shales which 

 occur near the bottom of the series were searched by 



NO. 1983, VOL. 76] 



Mr. Rogers without success for fossils. Conformably 

 underlying it is the Black Reef Series, consisting of 

 quartzites, grits, felspathic quartzites, conglomerates, 

 shales, &c. An exact divisional line between the two 

 formations cannot be drawn owing to the alterna- 

 tion of beds of quartzite and limestone near the 

 junction. 



The three series — Black Reef, Campbell Rand, and 

 Griquatown — are grouped together by Mr. Rogers 

 as the Transvaal System, a term introduced by 

 Molengraaff for the corresponding series in the Trans- 

 vaal. The name is scarcely a happy one; but since 

 it has also been adopted by the Transvaal Geological 

 .Survey, it will probably remain. Below the Black Reef 

 .Series is a basic series of amygdaloidal lavas — the 

 Pniel series of Stow — and this is succeeded by an 

 acid volcanic series to which the Cape geologists have 

 given the name Zoetlief Beds. There is an uncon- 

 formity between the two series ; but Mr. Rogers has 

 decided to include them in one group under the name 

 of the Ventersdorp System, which name is used for the 

 equivalent formation in the Transvaal. 



The Ventersdorp, Transvaal, and Matsap Systems 

 tie on a floor of granite and schists, and have been 

 folded during pre-Dwyka times into arches and 

 troughs, the axes of the folds embracing a wide area of 

 low mountains, the principal range of which is the 

 Langebergen. The underlying schists are grouped as 

 the Kraaipan Series ; they are equivalent to the Swazi- 

 land Series of the Transvaal, and consist largely of 

 magnetitic schists and thin quartzites. The Cape geolo- 

 gists have found no evidence of the granite being 

 intrusive into the schistose formation, although this 

 has been clearly proved in the Transvaal (Abelskop, 

 Monte Mar6, &c.). 'In this connection it is interest- 

 ing to note tliat the Geological Survey of the Trans- 

 vaal has recently recognised that the granite mass 

 which lies north of Johannesburg is younger than 

 the schists of the Swaziland Series (Moodies Series), 

 while it also admits the correctness of the view that 

 it constituted an older floor on which the basement 

 beds of the Witwatersrand Series were deposited 

 (Trans. Geol. Soc. S.A., vol. x., pp. 5^ and 57; cp. 

 also " Geology of South Africa," p. 93). 



Summarising, there is in South Africa, lying on an 

 old floor of schists and granite (which is almost cer- 

 tainly of Archjean age), a succession of sedimentary 

 rocks, older than the Devonian, and including the 

 Waterberg or Matsap, the Transvaal, the Ventersdorp 

 and the Witwatersrand Systems, all separated by 

 strong unconformities. In this accumulation of sedi- 

 ments, which is estimated to have a thickness of 

 between 50,000 and 60,000 feet, not a single fossil has 

 as yet been found. 



With regard to the Karroo rocks of the Transvaal, 

 a further step in their elucidation is marked by the 

 conclusion arrived at by the Survey, that the Bushveld 

 Sandstone Series (a succession of red marls, shales, and 

 fine-grained sandstones), together with the overlying 

 amygdaloids, are the equivalent of the Red Beds, 

 Cave Sandstone, and Volcanic Group of the Stormberg 

 Series of the Cape, the Orange River Colony, and 

 Natal (Kvnaston, Trans. Geol. Soc. S.A., vol. x., 

 p. 34). The recent discovery by Mr. J. Mitford 

 Bowker, in the Zoutpansberg district near the Limpopo 

 River, of fossil bones, \vhich Dr. Broom has deter- 

 mined to be of Stormberg age, sliows that the Bushveld 

 .Sandstone Series extended far to the north ; for it is 

 found to cross the Limpopo River into the Victoria 

 district of Mashonaland, where coal-seams have been 

 reported to occur in it (Proceedings of the meeting 

 of the Geol. Soc. of S.A. held on June 24, 1907). 



F. H. Hatch. 



