290 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. 



SUPPLEMENTARY NOTES ON GLACIATION BEFORE THE PERMIC PERIOD. 

 PRE-DEVONIC GLACIAL DEPOSITS OF SOUTH AFRICA. 



In Cape Colony, South Africa, there is a thick marine series of shales and sandstones known as 

 the Bokkeveld series, which appears to l)e of late Lower Devonic age. In western and southern 

 Cape Colony it is everywhere seen to lie upon the Table Mountain series of wide distribution 

 and with evidence of a glacial climate. The actual contact of these two series of strata shows 

 "no signs of unconformity" between them, but in the places "where a clean-cut section of the 

 junction can be seen," as "that on the left bank of the Gamka River immediately above its 

 great Poort through the Zwartebergen, * * * 'the end of the white sandstones [of the Table 

 Mountain series] and the beginning of the blue-black shales of the Bokkeveld is so sudden and 

 exact that one can place a knife between them and say confidently that on one side are the rocks 

 of the Table Mountain series and on the other those of the Bokkeveld series' " (Rogers, 1905: 121). 

 This sharp differentiation of the Lower Devonic black shales from the white sandstones of the 

 Table Mountain series seems to indicate clearlj' that the contact is a disconformable one and 

 that the sea invaded a land of sandstones. If this is true, the section is broken and we can not 

 therefore positively state the age of the Table Mountain series — it has as yet yielded no fossils of 

 stratigraphic value — other than that it is older than Devonic, but how much older is still to be 

 determined. The facts, however, that the rocks older than the Table Mountain series are far 

 more deformed and greatly intruded by igneous materials, that the Table Mountain and Bokke- 

 veld series do not give evidence of a long erosion interval between them, and that both were 

 deformed together subsequent to Bokkeveld time, seem to indicate that the age of the former is 

 rather late Siluric than early Devonic. The further fact that the fossils are bivalves and gastro- 

 pods indicates clearly that the Table Mountain series is of post-Cambric deposition. 



Rogers also says: 



"The Table Mountain series is remarkably constant in lithological characters throughout 

 its extent. The maximum thickness is about 5,000 feet, and of this more than 4,000 feet are 

 sandstones or quartzites" (107). 



" The whitish-grey colour of so much of the sandstone belonging to this series is due to weather- 

 ing. At a distance of 1 or 2 feet from the outside the rock is usually blue, owing to a small quantity 

 of iron in the state of ferrous compounds" (108). 



"A very frequent characteristic of the sandstones of this group is the occurrence of round 

 pebbles of white quartz up to 3 inches in diameter. They usually occur singly, more rarely in 

 thin layers a few feet long and about an inch thick. The pebbles themselves are rarely more 

 than an inch in diameter. It is rather difficult to explain the frequence of isolated pebbles in the 

 sandstone without recourse to some agency that lifted pebbles from the shore and dropped them 

 in deeper waters" (109-10). 



"In the western mountains a second shale band is found about 1,000 feet below the top of the 

 series. * * * The most interesting point about the Pakhuis section is the occurrence of 

 pebbles up to 5 inches in diameter scattered irregularly through the shale and mudstone, without 

 any tendency to form beds of conglomerate. Several of the pebbles have been found to be 

 flattened on one or more sides and deeply striated in the manner characteristic of pebbles that 

 have come from a glaciated region" (111-12). 



"The occurrence of flattened and striated pebbles scattered at intervals through a fine-grained 

 laminated rock is very strong evidence that glacial conditions prevailed on the land whence the peb- 

 bles came, and that these pebbles were carried away from the land by floating ice and dropped 

 by the melting of the ice on to the mud being deposited at the bottom of the water" (113). 



"If the Table Mountain sandstone is regarded as an ordinary coarse deposit formed in either 

 a fresh-water basin or the sea, the land from which the material was washed can not have lain far 

 from the present outcrops of the rock. The only evidence of the closer proximity to land of one 

 part of the sandstone than another is the greater development of conglomerates on the west, in 

 the Piquetberg Division and the Ohfant's River Mountains, than elsewhere. There is no such 

 evidence known from the Bokkeveld Mountain, or along the Zwartebergen, or the south coast. 

 At present, then, we must conclude that while the nature of the rock renders it probable that the 

 Table Mountain series, so far as exposed in the Colony, was formed not far from land, and that 

 consequently the land lay more or less parallel to the present distribution of the series, the only 

 definite clue to the position of any part of that land is to be found in the conglomerates of the 

 west" (116-17). 



