418 ME. A. J. C. MOLYNET/X ON THE [Aug. I909, 



have not reached the constant angle of a natural batter. River- 

 erosion, the prime cause, is thus still at work and undermines the 

 base of the canon, while subaerial agencies attack the verge and 

 slopes. The formation is of gneiss, striking E. 7° N., dipping 

 southwards 45° to. 65°, and so jointed that the removal of titanic 

 blocks by rains, river, and floods is greatly facilitated. These 

 blocks, abraded to spherical shape, strew the path of the river in 

 the canon. 



At a bend of the gorge a sand-river comes in from the east- 

 south-east ; it also has a wide, though stunted chasm. The junction 

 is about 500 feet above, and 6 miles from, the outlet of the 

 Molongushi to the Luano plain. That is, the gradual reduction 

 of the latter basin has led to the erosion of this canon exceeding 

 6 miles in length, and 1500 feet in depth, in order that the Molon- 

 gushi River should accommodate itself to a new change of level. 

 The causes that brought this about will lead us to discuss (pp. 430- 

 35) the influence that structure, folding, and denudation have had 

 on the surface-features or physiography of these wide regions. 



The Danda Plats. 



It has already been mentioned that the Zambesi river, between 

 its junction with the Kafue and Feira, runs in a similar trough, 

 or ' straits ', between the abrupt plateau-edges on each side. That 

 on the southern side is well known to travellers on the path from 

 Salisbury by way of Sipolilo's to Feira, when a descent of 2000 feet 

 is made in about three times that distance horizontally. 



At the back of this escarpment the plateau is made up of granite 

 and hornblende-gneiss with east-to-west cleavage. Then the path 

 rises rapidly over a pass of 3550 feet altitude, the rocky summits 

 around being somewhat higher. This range is of pre-Karroo, 

 fine, fissile sandstone, very micaceous and showing patches of 

 specular iron — similar to and probably possessing affinities with 

 the Chasonsa Series in North- Western Rhodesia. Great blocks of 

 rock from the summits lie round about, and from here one enjoys 

 a marvellous view of the river-flats far 'below, stretching away to 

 the left and right, and across the Zambesi towards Feira, behind 

 which the corresponding plateau of North-Eastern Rhodesia makes 

 a hazy background. 



In making the descent, it is seen that these ancient sedimentary 

 rocks strike east and west, and gradually increase in dip from ] 5° 

 at the summit to 60° and 70° at the base. At nearly two-thirds of 

 the drop, the sandstone changes to a micaceous schist with augen- 

 structure around the quartz. At the base the rock contains quartz 

 in layers of an inch, and resembles the saccharoidal quartzites, 

 which, on disintegration, give rise to screes of rubble at the hill-foot. 

 The section in fig. 5 (p. 417) illustrates the arrangement of strata, 

 but not of the contact with the Karroo beds that lie in the flats. 



From the base the country comprises the remarkable Danda 



