406 



W. M. DAVIS OBSERVATIONS IN SOUTH AFRICA 



The chief results of this day's excursion were : First, the repetition of 

 many significant features that had been noted the day before at Matjes- 

 fontein and which were thus shown to be of more than local value: second, 

 the recognition of the composite structure of the formation, indicating at 

 least two advances of an aggrading ice-sheet into this district; third, the 

 fuller evidence of induration, deformation, and erosion suffered by the 

 Dwyka and other formations of the Karroo series, from which it is mani- 

 fest that we have here to do with a glacial formation of much greater 

 antiquity than that of Pleistocene date. 



A second excursion from Laingsburg, August 22, took us down Buffels 

 river (see figure 2) to the Witteberg range, as already described. A gen- 

 eralized section some 10 miles in length is given in figure 8, from which 

 it appears that the structure of the district is characterized by many 

 strong east-west folds, causing the Dwyka to reappear in successive out- 

 crops along the east-west ridges and valleys. The southern continuation 



Figure 8. — General Section of folded Witteberg, Dwyka, and Ecca Formations, South of 



Laingsburg. 



Length of section, about 10 miles. This is the northern continuation of figure 3. 



of this section is given in figure 3. We no longer gave special attention 

 to the structure or the composition of tillite, for a passing glance now 

 sufficed to show that it repeated persistently all the essential features pre- 

 viously recognized, and a brief pause sufficed to find striated stones on 

 almost every tillite outcrop ; but we examined the topographic features of 

 the formation with some care, because they showed that the subdivisions 

 of the Dwyka and their conformable succession from the Witteberg beloAV 

 to the Ecca above, as recognized the day before, were maintained for 10 

 or more miles along and across the strike of the ridges. This encouraged 

 us to believe that further exploration may enable South African geologists 

 to establish well defined glacial and non-glacial epochs in the Dwyka 

 period — an important matter to which but little attention seems to have 

 been given, although Dunn and others have reported the occurrence of 

 alternating shales and tillite beds at various points. Some further ac- 

 count of the ridges of Witteberg sandstone which rise to the south of this 

 Dwyka area is given in an earlier section of this paper. 



The glaciated Dwyka floor near Ngotshe, Vryheit. — Three members 

 of Rogers' party, Coleman of Toronto, Penck of Vienna, and I, left the 



