F. P. Mennell — The Rhodesian Banket Beds. 361 



by metamorphism into hornblendic or cliloritic schist. And crushing 

 as a mode of formation is obviously inconceivable in the cases, by no 

 means uncommon, where numerous chlorite-schist pebbles occur in 

 a sandy matrix. 



III. Stratigraphical Position. 



In discussing its origin I have already remarked on the constant 

 horizon of the conglomerate. It is the topmost member of the 

 Ehodesian Archsean group, which comprises — 



(3) Conglomerate and grits. 

 (2) Banded ironstone. 

 (1) Crystalline schists. 



The last include the oldest rocks, but also a certain number 

 of recrystallized masses which represent relatively late igneous 

 intrusions. The three groups are unconformable, though they 

 usually exhibit a concordant strike and dip. These, however, are 

 obviously more apparent than real, the dip being that of foliation 

 or cleavage. The banded ironstones, it may be as well to mention, 

 are an enormously thick series of altered sediments, sometimes, 

 though rarely, slaty, but more often highly silicified and jaspery, 

 or consisting almost entirely of limonite, haematite, or magnetite. 

 The banding, which is probably a very unsafe indication of bedding, 

 is, however, always to be seen, an alternation of red and white, 

 brown and white, or other dark and light bands being conspicuous 

 in most cases. These rocks are known as Griquatown Beds in Cape 

 Colony and Hospital Hill Beds on the Eand, in which latter locality 

 they unconformably underlie the famous banket. Our conglomerate 

 has thus precisely the same stratigraphical position : it unconformably 

 overlies the banded ironstone and is implicated in most of the later 

 movements of that rock, a corresponding foliation having been 

 impressed upon it. The great granite masses of Rhodesia are 

 intrusive in all the rocks up to and including the conglomerate, 

 and some of the epidiorites, in which the granites are themselves 

 intrusive, are almost certainly the result of the alteration of doleritic 

 intrusions also later than the conglomerate. In any case the con- 

 glomerate contains no pebbles derived from them. I think the 

 probabilities — and these are all we have to go upon in any South 

 African correlations — are extremely strong in favour of the Rhodesian 

 conglomerate being the stratigraphical equivalent of the Witwaters- 

 rand Beds. 



As regards distribution, the only districts where the conglomerate 

 has yet been continuously traced over a large area are those of 

 Bulawayo and Lomagundi. It is, however, known to occur in the 

 Umtali, Abercorn, Mazoe, Hartley, Sebakwe, Selukwe, Victoria, 

 Belingwe, Bubi, and Gwanda districts, that is to say, scattered 

 over an area of fully 100,000 square miles. The thickness of the 

 rock and the associated finer sediments is very variable. Near 

 Bulawayo it is from a few hundred to several thousand feet thick, 

 and it is probable that the average is at least 1,000 feet. 



