The Forming of the Drakensherg. 61 



of coal, and in some places rests directly upon the coal without the 

 latter being in any way disturbed. 



At Molteno, Molyneux has recorded pebbles and boulders embedded 

 in the actual coal itself. 



The pebbles are almost entirely of white, brownish, or glassy 

 quartzite like that of the Witteberg and Table Mountain series, but 

 hard white slatey rocks are also represented. 



The distribution of these pebbles, too, indicates their original 

 source. They are most abundant to the west of Molteno ; at Indwe, 

 Cala, and Engcobo they occur rather scattered ; south of Burghers- 

 dorp they are fairly common ; in Aliwal North and Herschel very 

 rare ; while Mr. Schwarz has not noticed them in Matatiele. 



The pebbles are found principally in the Molteno beds, but some- 

 times, though not commonly, in the lower part of the Eed beds. 



The pebbles show very fine, smooth surfaces, and resemble very 

 much the rolled material found along the shores of a lake or along 

 a coast-line. To take a near example, we have the same thing at 

 Hout's Bay, where the well-rounded boulders are derived from the 

 Table Mountain sandstone. 



The difficulty, of course, lies in accounting satisfactorily for their 

 transport, for it is evident that a current with a strength sufficient to 

 move one of these boulders would entirely prevent the deposition of 

 any finely-divided sediment. The quartzites must have been carried 

 a great distance, for the nearest outcrop of the Witteberg series at 

 the present day is about 130 miles due south, at Graham stown. 



The pebbles might have been brought across embedded in sand 

 and clay adhering to the roots of trees, but perhaps a more reasonable 

 explanation is that they have been transported by ice. 



It is well known that in cold regions the beach material becomes 

 incorporated with the ice formed along the shore line. 



In the spring such a floe of ice would be carried away by currents 

 and the foreign material included in its mass deposited during the 

 process of melting. 



In this manner the pebbles might have been dropped and so become 

 embedded in soft shale and even in the coal itself. If this is the 

 correct explanation, the old shore line at this period would have 

 been formed of Witteberg quartzites somewhere to the south of 

 Grahamstown, and the Table Mountain sandstone cropping out 

 further south still would not necessarily be a contributor to the 

 conglomerate beds. 



In Matatiele, as will be seen from the map, the nearest outcrop of 

 quartzite is in East Pondoland ; but it is probable that during Storm- 

 berg times the Dwyka conglomerate entirely covered up the Table 



