PIirsiCAL GEOGKAPHY 01' THE CAPE COLONY. 243 



stay in Africa. The first was on the northern tlank of the Witte- 

 bergen, north of Willowmore. The rock had here the same general 

 character as at GIrahamstown, but at some spots it contained very 

 lew blocks ; at others they were numerous, and principally of a 

 quartzite that I could not distinguish from the quartzite of the 

 adjoining mountains (no. 4 of the table on p. 240). In a slice, cut 

 from a specimen from this neighbourhood, rock-fragments are plen- 

 tiful : among them are one showing micropegmatitic structure and a 

 bit of Chiastolite-rock. At Prince Albert, my next locality, the 

 conglomerate contained a large number of good-sized boulders, many 

 angular and subangular, rounded pebbles, many small fragments, 

 and the usual quartz- chips. Many of the large boulders were of 

 various kinds of crystalline schist, and there were a fair number of 

 quartzite-blocks. I had by this time got to know the quartzites 

 of the Wittebergen fairly well, and, if it were not that all quart- 

 zites are a good deal alike, I should have said without hesitation 

 that the quartzite-blocks in the conglomerate had come from these 

 mountains. I also noticed bits of a jaspideous breccia exactly like a 

 rock that I had been shown from the Rhinoster liiver (in the Trans- 

 vaal), and amygdaloids like those occurring in the gravels of the 

 Vaal Elver. The matrix, under the microscope, was the same as 

 in the specimen from Grahamstown, but there were recognizable 

 chips of granite and quartzite. 



1 also saw the conglomerate in the valley of the Buffels River, 

 but noticed nothing new in it there. 



The matrix of the rock has all the look of mashed-up granite 

 mixed with fine mud *. Many difi'erent rocks have contributed to 

 furnish the boulders ; among these we may safely put the Witteberg 

 quartzites, granite, and rocks apparently very similar to the 

 crystalline series of the Transvaal. The size and angularity of the 

 boulders suggest the action of ice. I have a pebble from the 

 conglomerate of Prince Albert, which might pass for an ice- 

 scratched block in a country known to have been glaciated ; but such 

 scratches as the block shows are not necessarily ice-marks ; and this 

 is all the evidence I could gather on this point during my hurried 

 examinations. 



The notion I formed as to the origin of this rock was that it was 

 a coarse shingle formed along a receding coast-line (see p. 267). 

 Both Dr. A. Geikie and Dr. Hatch have suggested to me that it has 

 very markedly the aspect of a volcanic breccia, and its strongly 

 fragmental character under the microscope is in favour of the view 

 that this is its nature. But though in this respect it is scarcely 

 possible to draw any distinction between the Dwyka Conglomerate 

 and some undoubted tufi^s, its resemblance to such rocks as the 

 " Dolomitic Conglomerate " of Bristol, and the " Brockrams " of the 

 Vale of Eden, is equally close. Dr. Hatch also lays some stress on 



* Dr. Cohen has contributed to the ' Neues Jahrbuch ' (Beilage-Band v. 

 p. 198, 1887) an elaborate account of the rocks of the Cape Colony, from the 

 Dwyka Conglomerate upwards. He finds that tlie cement of this conglomerate 

 contains a considerable amount of soluble silica. 



