THE GEOLOGICAL FORMATION OF GERMAN S.\V. AFRICA. 335 



vertical position. As we proceed North the dip of these Archaean 

 beds diminishes as does the elevation of the surface. 



Soon we g"et to the Dwyka again, resting partly on Clayslates 

 and partly on Table Mountain Sandstone which appears again, 

 just West of the GrcaL Karas Mountains is an extensive deposit 

 of Dwyka Conglomerate forming a low level plateau and extending 

 towards the Little Karas Mountains. 



A vast extent of Sandstone is now passed over, extending to a 

 point about twenty miles south of Keetmanshoop. An interesting 

 Dolerite intrusion is noticed in the middle of this extent of Sand- 

 stone. A small boss of Dolerite is surrounded by a thin fring'c 

 of Granite, which is in turn surrounded bv Clayslate, shewing that 

 the Dolerite has carried up both rocks in the order in which they 

 exist below the surface. 



The Sandstone has at several localities, notably at a place 

 boasting the euphonious name of " Harige Kaakebeen." a number 

 of clayey nodules. Long before I visited this locality I was told 

 about some wonderful camel tracks to be seen crossing a river. 

 These tracks were described as being imprints in solid rock and, 

 when I arrived at the locality and ascertained the geological 

 horizon to which the rocks belonged, I was naturally anxious to 

 find out what sort of camels had existed in Silurian or Devonian 

 times. The supposed tracks were found to be cavities caused by 

 the eroding action of water running over a hard rock containing 

 soft nodules. Strangely enough the "Tracks" pointed across the 

 river, but, on making investigations, it was ascertained that the 

 river had fairly recently changed its course and had previously 

 run at right angles to its present direction. 



At a point twenty miles south of Keetmanshoop, the Sandstone 

 disappears and the same Clayslates, Limestones and Quartzitc 

 that were observed at the Karas Mountains appear again, this time 

 dipping to the South and apparently forming a vast Syncline over 

 thirty miles wide. Mica Schist is not well developed at the north- 

 ern end of the Syncline, having apparently been formed only where 

 the original slates came in actual contact with the intruding 

 granite. 



From this point right up to Keetmanshoop these Archsean beds 

 are met with at frequent intervals, the numerous breaks being- due 

 to very extensive intrusions of Dolerite, which has in all proba- 

 bility formed Lava flows, smce there are outcrops miles in width. 

 A borehole put down a few miles south of Keetmanshoop shews 

 dolerite underlying the Archaean slates. 



All round Keetmanshoop similar conditions prevail. 



The Little Karas Mountains are composed mainly of rather flat 

 hills of Table Mountain Sandstone. A great deal of erosion has 

 taken place. Here are extensive intrusions of Diorite, in places 

 shewing the same Archaean slates, Ouartzites and Limestones 

 round the bosses. Half way between Groendoorn, a farm in the 

 Little Karas Mountains, and Quickachis is a great boss of Amyg- 

 daloidal Diabase, which has carried un a little Granite during the 

 intrusion. 



At numerous localities throughout the whole area described, a 



