192 . MR. G. W. LAMPLUGH ON THE [May I907, 



each other, until the walls are almost vertical ; and finally the 

 gorge is narrowed to a profound cleft, 300 feet deep and only 

 15 to 20 feet wide at the bottom and about as many yards wide at 

 the top (see PI. XV). At this gloomy spot, which we have named 

 'Kalonga's Cleft,' 1 the gorge turns at right angles and holds a 

 sombre, cliff-bound water-pool. Into this pool the Karamba, when 

 shrunken with the drought, makes a sideway leap from the plateau ; 

 but when flooded it must pour over the terminal wall in a terrific 

 cascade occupying the whole breadth of the chasm. Between the 

 water-pool and the lip of the shallow upper valley, the basalts and 

 amygdaloidal breccias which form the precipice are sliced trans- 

 versely to the valley by several vertical planes of fracture and 

 crushing, with anastomosing branches, which have been gouged out 

 by the stream into a series of deep riffies, as shown in the section, 

 fig. 7 (p. 190) and bird's-eye view, fig. 8 (p. 191). At the time 

 of our visit all the water was intercepted by the first of these 

 riffles, and conveyed sideways to a lateral recess, from which it 

 recoils at a sharp angle and leaps into the pool. 



Both as a spectacle, and as a most instructive lesson in the 

 erosion of the basalts, I commend Kalonga's Cleft to the notice of 

 any traveller who may find himself within reach of its flood-rent 

 portals. 



Other Noteworthy Structures. 



At the surface, the basalt usually exhibits the familiar spheroidal 

 habit of weathering ; but while in some places the rock is deeply 

 rotted to a rusty earth or loam, in others it remains perfectly 

 hard and fresh immediately beneath a thin scaly crust which is 

 shed off almost as rapidly as it is formed. Yet there did not seem 

 to be any difference in composition accompanying this difference in 

 weathering. The same anomaly is, of course, sometimes noticeable 

 in the basalts of temperate climates, but I have never seen the 

 distinction so sharply marked as in the Batoka Basalts. 



The vesicular basalts occasionally show a kind of ' pillow- 

 structure,' in the concentric arrangement of the amygdaloidal 

 vesicles, due to the rolling-over of masses of half-cooled lava during 

 the flow. This structure, so far as I saw, was not developed at 

 the surface of the flows, but in their interior. The best example 

 that came under observation was on the smoothly-worn rock-floor 

 of the Mamba River, 200 yards below our camp, where the 

 concentric outcrop of coarsely amygdaloidal bands divided up the 

 rock into irregular oval masses, measuring from 2 to 6 feet in 

 diameter. 



The more massive basalts, when fresh, occasionally show a 

 curiously knotted aspect on stream-worn surfaces, somewhat 

 resembling a spherulitic structure. This appears to be due to the 

 segregation of the porphyritic constituents, especially the plagioclase- 



1 Named after a native guide. 



