Notes on the Geology of the Drakensbergen. 423 



outcrops and the streams and dongas expose the intervening shales, 

 but the sandstone beds predominate. These lower sandstones differ 

 from the cave sandstone in being softer, usually micaceous, and 

 often inclined to be shaly. In fact, they are more like the sandstones 

 of the coal areas, and probably represent the upper portion of the 

 Molteno beds. 



Considering the several beds of the compact sandstone with their 

 intervening softer beds as forming the series called the Cave Sand- 

 stone, this formation may be put down as from 1,000 to 1,200 ft. 

 thick. 



I did not find the bone-breccia referred to by Mr. David Draper 

 in his paper (Quarterly Journal of the Geological Society, November, 

 1894, p. 551), though I was constantly on the look-out for it. 



The cave floors consist of earth, dust, ashes, eland and cattle 

 droppings, and rocks — fragments and boulders. In this debris 

 Mr. M. S. Evans found many rude Bushman implements made of 

 the chert, agate, and dolerite, previously described. They are 

 mostly spear and arrow heads and skin scrapers. These, except 

 some given to the Durban Museum, have been sent to Professor 

 Boyd Dawkins. 



From reading Draper's paper I expected to find that the great 

 wall of the Drakensbergen which we were approaching consisted of 

 sedimentary strata capped by a thousand feet of amygdaloid traps, 

 and the lines of escarpment running parallel and horizontal above 

 each other for miles, and the straight base line of the final cliffs 

 seemed entirely to confirm this idea. It was, therefore, quite a 

 surprise to find nothing but amygdaloidal rock from the top of the 

 mountain right down to below where one leaves one's horses to 

 begin climbing ; that is to say, right down to the dolerite flats forming 

 the plateau at 6,500 ft. above sea-level. These parallel minor cliffs 

 and escarpments, however, turned out to be quite the same as the 

 huge main walls. Thus the whole main range consists of amygda- 

 loidal lavas of a vertical thickness of 2,500 ft. at Bushman's Pass, 

 the lowest part, and 4,500 ft. thick at the highest peaks such as 

 Champagne Castle and Mont aux Sources. I may mention in 

 passing that my observations are based on three ascents of the 

 Bushman's Pass and the ascent of the peaks on either side of it, 

 one ascent to the top of Giant's Castle, and various visits to the 

 lower slopes, namely, at the foot of a large peak between the Pass and 

 Giant's Castle, the foot of Champagne Castle and the foot of Mont 

 aux Sources in two places, and the top of the ridge above Koodoo 

 Pass to Witzie's Hoek. This makes seven different places where I 



29 



