Vol. G^.'] GEOLOGY OF TflE ZAMBEZI EASIX. 187 



are much more strongly and closely developed in the massive basalts 

 than in the breccias and slaggy amygbaloids. The columnar 

 appearance which is often conspicuous in the walls of the gorge is 

 usually due to the emergence of close-set parallel joints on the 

 vertical faces of the more massive beds. The only place at which 

 I saw really well-formed columnar structure was iu a low cliff 

 overlooking the ']Nongu, an eastern tributary of the Kalomo River ; 

 although in a few other places, as, for example, in the main gorge at 

 the confluence of the Karamba, the basalt showed a tendency towards 

 wide-spaced polygonal jointing, without system or regularity. 



Besides possessing this joint-system, the basalts are also sliced 

 at intervals in the same approximately east-to-west direction by 

 pronounced fractures, along which there are sometimes signs of 

 actual displacement; the opposing rock-faces showing slickensides 

 and brecciation, with the occasional intercalation of vein-stuff, 

 principally calcite and chalcedony. That vertical movement has 

 occurred along these fractures is certain, but the monotonous 

 character of the basalts generally renders its estimation impracticable. 

 In one case, however, just above Kalonga's Cleft on the Karamba 

 (see fig. 8, p. 191, and PI. XV) where the throw of a very conspicuous 

 fracture of this kind could be measured, it was only 3 or 4 feet. 



These fractures tend to run in belts, within which they recur at 

 short intervals, forming veritable ' shatter-belts/ They probably 

 represent the minor re-adjustments of the great subsidence that finds 

 its main expression in the Deka Fault. 1 As we approached this fault 

 in descending the Deka valley, veins of calcite were particularly 

 conspicuous in the basalts, all striking approximately parallel to the 

 fault. 



Effect of the Joints and Fractures on the 

 Drainage-Channels. 



Both joints and ' shatter-belts,' but especially the latter, form 

 vertical planes of rapid erosion, which, though invisible on the surface 

 of the plateau where the stream-gradients are low, are picked out 

 very rapidly into deep troughs by the forceful torrents rejuvenated 

 by the recession of the main gorge (see p. 168). And as these 

 structural planes frequently lie athwart the water-courses of the 

 plateau, they constantly give rise to sharply-angular diversions 

 from the previous direction where the streams plunge suddenly down- 

 ward on leaving the upland. This effect is produced the more 

 readily owing to the enormous seasonal variation in the volume of 

 drainage. Except the Zambezi, almost all the stream-beds of the 

 country traversed are mere flood-channels, filled to the brim for short 

 periods during the rains, but during much of the year either quite 



1 One of these veins, over 2 feet wide, which we crossed when about 12 miles 

 south-west of Bumbusi, was entirely filled with calcite in very large crystals, and 

 was seen to the east and west of our track for at least 40 or 50 yards. In a country 

 where limestone is scarce, it is possible that this material may eventually become 

 of some economic value. 



o2 



