116 E. T. Mellor — Glacial Conglomerate of South Africa. 



from erosion it fills preexisting valleys, and is usually espe- 

 cially abundant below ancient escarpments of the older rocks, 

 and in such places bowlders often of very large size, attaining 

 in some cases eight or ten feet in diameter, are exceptionally 

 numerous. After the complete weathering away of the matrix 

 the bowlders remain abundantly scattered over areas previ- 

 ously occupied by the conglomerate. (See fig. 2.) 



Glaciated Surfaces below the Conglomerate — Direction of 

 Ice-Movement. — The progressive removal by denudation of 

 the Glacial Conglomerate around the margins of the areas now 

 occupied by the Karroo System and its outliers continually lays 

 bare fresh portions of the underlying old land surface. Where 

 these include outcrops of hard and moderately fine-grained rocks, 

 the latter frequently present excellent examples of glacially 

 striated surfaces,* some of which are represented in the photo- 

 graphs reproduced in figures 3, 4, 5. Striated surfaces of 

 this kind were long ago described by Sutherland in Natal, by 

 Griesbach in the same colony, by Dunn and Schenck in the 

 neighbourhood of the Yaal River, and more recently by Molen- 

 graaff in the South-Eastern Transvaal, and by Rogers and 

 Schwartz in the Prieska district in the north of Cape Colony. 

 While working on an area lying about 25 miles east of Pretoria 

 in 1903, I found the surface shown in figure 3, and later those 

 in figures 4 and 5. These latter occur on the edge of an out- 

 lier of Karroo rocks some 25 miles further east, which includes 

 the coal seam worked at the Douglas colliery. I have since 

 met with many similar surfaces distributed over an area of 

 some 300 square miles. The striation in most cases is exceed- 

 ingly clear, and the direction of ice-movement easily deter- 

 mined and remarkably consistent. In all the examples found 

 it only varies within a few degrees from magnetic north and 

 south, the direction of movement being in a southerly direction, 

 which is also true in general for the other districts in South 

 Africa where striated surfaces have been found. This con- 

 sistency of direction over so considerable an area and in the 

 case of surfaces lying 25 miles apart, points to the existence of 

 an ice-sheet of considerable magnitude, rather than to that of 

 a number of more or less isolated glaciers, a conclusion which 

 is supported by the nature of the land surface laid bare by 

 the disappearance of the Karroo deposits. 



Where the Water berg Sandstone Formation, which occupies 

 much of the district here referred to, has been long exposed to 

 ordinary denudation, the rivers cut deep valleys and gorges in 

 the sandstone, giving rise to very varied and occasionally rug- 



* E. T. Mellor, On Some Glaciated Land Surfaces occurring in the Dis- 

 trict between Pretoria and Balmoral. Trans. Geol. Soc. S. A., vol. vii, part 1, 

 1904. 



