REVIEWS 515 



Campbell Rand Series, beds of unmistakably glacial character are met with which 

 may be provisionally correlated with the prae-Cambrian glaciation in the Gaisa 

 Series of the Varanger Fiord in Northern Norway, while similar deposits have 

 been described from China and South Australia, which would seem from their 

 position to be also of prse-Cambrian age. The Lower Huronian glaciation of 

 Cobalt and other localties in Northern Ontario would, if its stratigraphical 

 position be correctly determined, be in all probability of still greater antiquity 

 than any of these. 



The Matsap Series, the latest of the old unfossiliferous formations of northern 

 Cape Colony, is identified with the Waterberg Sandstone of the Transvaal. It 

 appears to resemble in some respects the Torridonian, and was probably, like it, a 

 continental formation formed to a large extent of sands and conglomerates laid 

 down by streams from mountain slopes. The authors do not accept the correlation 

 of the Waterberg and Matsap Beds with the Table Mountain Sandstone in the 

 south. They believe them to be much older, and thus throw back the unfossili- 

 ferous formations of the northern area to a still higher antiquity than that with 

 which they are usually credited. 



The Table Mountain Sandstone Series itself is usually considered to be 

 Devonian, since it is overlaid conformably by the Bokkeveld Series, which may be 

 placed at or near the base of the Middle Devonian. The " sudden and exact " 

 (to use the authors' expression) line of division between the sandstones of the 

 former and shales of the latter may, however, represent a considerable lapse of 

 time. It is unfortunate that the only fossils yet discovered in the Table Mountain 

 Sandstone Series, some lamellibranchs in micaceous sandy shales at the base, 

 should be so badly preserved as to be undeterminable. The most remarkable 

 episode in the deposition of these beds, which the authors regard as of fluviatile 

 origin, is represented by the intercalation of some 300 ft. of shales, the lower third 

 of which consists of greenish blue or reddish mudstone without lamination contain- 

 ing scattered pebbles and boulders with flattened and striated surfaces, indicating 

 a second period of intense cold in South Africa. It may be suggested that they 

 represent a period when the waters of the rivers were dammed back by the accumu- 

 lation of glacier ice to the southward, so that great lakes were formed in much the 

 same way as happened in North America in the Pleistocene glaciation. 



The Lower Bokkeveld Beds afford the first unmistakable evidence of the 

 presence of the sea in South Africa, but the marine forms of life that characterised 

 them soon disappeared, and it is probable that for a prolonged period, represented 

 by the Upper Bokkeveld, Witteberg, and the greater part of the Karoo rocks, 

 characterised by the frequent occurrence of remains of terrestrial plants, the 

 geographical conditions in Cape Colony were akin to those that now prevail on 

 the northern shores of Asia, where the gradual recession of the ocean has laid bare 

 a wide tract of low-lying country on which alluvial desposits containing a great 

 quantity of vegetable material are slowly accumulating. 1 The special character of 

 these deposits is usually attributed to the melting of the upper course of the rivers, 

 while the lower portions are still frozen. This may also have been the case in 

 South Africa during a large portion of the period above referred to, but the 

 incoming of the Karroo is marked by evidence of the prevalence of much more 



1 The frequent occurrence of reptile remains in the Karroo Beds may be com- 

 pared to the numerous entombments of mammoths in the swamps of Northern 

 Siberia. 



