528 PROFESSOR J. W. GREGORY. 



The various conclusions as to the age of the Waterberg and Transvaal Systems 

 are matters of impression rather than of direct evidence. One current of opinion is 

 in favour of their Lower Palaeozoic age, the absence of fossils being explained by 

 " the imperfection of the geological record " and the still inadequate search. Accord- 

 ing to the opposite view, the beds are lithologically similar to the unfoliated Upper 

 Ko/oie rocks, and the fact that this vast series of sediments has not yet yielded a 

 single fossil, affords strong presumptive evidence that the beds are pre-Palseozoie. 

 The balance of opinion in South Afriea is in favour of the former view ; my own 

 tendency, I must admit, is towards the latter. 



C. The Eozoic Gneisses. 



The sehistose quartzites of the Bailundo Series are interstratified with coarse 

 gneisses which are exposed at many localities in the Bailundo and Bihe districts. 

 These rocks so elosely resemble the Laurentian and Lewisian gneisses that they are 

 probably foliated plutonic rocks of Lower Eozoic age. Similar rocks are also well 

 exposed at the Lengwe Gorge, where they are not associated with quartzites ; but the 

 gneisses are so similar that they probably belong to the same division, and as the 

 strike of their foliation is roughly east and west, they are probably connected, at least 

 below the surfaee, north of the Benguella railway from Lengwe to Bailundo. 



The Benguella gneiss series is intruded by granites which occupy most of the 

 country for 250 miles between Mount Sahoa and Candumbo to the east of the Huambo. 

 These granites have clearly been intruded into the older Archean rocks, which have 

 been altered at the contact into hornfels, while their foliation is often very disturbed. 

 The granites are older than the Oendolongo Series, which is largely composed of 

 their debris. 



The most typical of the Benguella gneisses resemble those of the Laurentian rocks 

 of Canada and the Lewisian of Scotland ; and their associated granites are connected 

 by the charnockites of the Ochilesa district with the ancient rocks of Southern India 

 and with the Ivory Coast of Western Africa, where, amongst granites and gneisses 

 similar to those of Benguella, the existence of a Charnockitic Series has been deter- 

 mined by Professor Lacroix (1910). 



D; The Volcanic Rocks and Intrusive Rocks of Undetermined Age. 



The widest sheet of volcanic rocks is that of the rhyolite tuffs of the Oendolongo 

 Hills ; and the occurrence of similar tuffs in the beds east of Lepi is a point in favour 

 of the Lepi beds belonging to the same series as the Oendolongo Quartzites. Near 

 Lepi occur some of the best exposures of the widely spread olivine-dolerite. It 

 occurs in dykes in many localities, and apparently in sills near Lepi, and in sills or 

 flows to the north-east of Bailundo. The dolerite is apparently younger than the 

 Oendolongo Series, though 1 had not time to trace a dyke of this rock west of Cruz's 

 into the sandstones. 



