November 30, 1905] 



NA TURE 



capable of carrying the flood-waters also, and the course 

 of the stream becomes fixed along them. We found 

 striking illustrations of these conditions both in the Batoka 

 Gorge and in the beds of the tributaries in many places. 

 The sudden and acute bends that are so peculiarly 

 characteristic of the Zambesi below the Falls are in this 

 way readily explicable. 



A broad ancient river-flat, with low sloping banks on 

 both sides, excavated across the edges of the gently dipping 

 lava-flows, is distinctly traceable for many miles below 

 the Falls, until obscured by the breaking up of the plateau 

 by the gradually lengthening development of the lateral 

 chasms of the rejuvenated tributaries. This flat is com- 

 parable in breadth and general aspect to the valley of the 

 Zambesi above the Falls ; and the presence of a few rounded 

 pebbles upon it above the brink of the gorge gives further 

 evidence for the former, flow of the river over its surface. 

 It is continued southward as a shallow depression in the 

 surface of the plateau for five or six miles from the Falls, 

 and then curves eastward. 



It may be mentioned here, as a matter deserving the 

 attention of archaeologists, that rudely chipped implements 

 of chalcedony, agate, and jasper are very abundant in 

 many places on this ancient river-platform, and also 

 upon the low rocky hummocks bordering the Zambesi 

 above the Falls. A few of these implements show signs 

 of wear as if by river-action, and may therefore possibly 

 be of considerable antiquity. We found them, here and 

 there, in profusion during the first 20 miles of our east- 

 ward journey, but very rarely during the later stages of 

 the trek. A collection of these implements was exhibited 

 at a meeting of the Anthropological Section. 



The erratic zigzags of the Batoka Gorge swing to and 

 fro within this broad depression, but without escaping from 

 it. Even within the gorge, the river, still possessing a 

 high gradient, tends to confine itself within narrower limits 

 as it scoops out the less resistant portions of its bed, 

 leaving many abandoned channels, rock-terraces, and spur- 

 like ridges to break the severity of its canon walls. 



Nowhere can these features be better studied than in 

 the left bank of the Gorge, about 7 miles below the Falls, 

 around the confluence of the Songwe, a little tributarv 

 which has itself carved out a narrow chasm about three- 

 quarters of a mile long and more than 400 feet deep into 

 the margin of the plateau. As well for its savage magnifi- 

 cence as for its scientific interest, this spot deserves to 

 be visited ; and one may be allowed to express the hope 

 that the responsible authorities will undertake the com- 

 paratively light work of clearing a track from the Falls, 

 to render it accessible to the tourist. 



To one whose first impressions of the Zambesi had been 

 gained from the mile-wide river above the Falls, it was 

 astonishing to find the whole river, at its present low stage, 

 confined at this place within a channel not more than 

 35 yards in width — bordered, it is true, by a rocky scar, 

 about 150 yards wide, honeycombed with deep " pot- 

 holes," which was evidently submerged during the floods. 

 After seeing it one could understand how the idea has 

 arisen — and still lingers — that part of the Zambesi is 

 swallowed up at the Falls into an underground channel. 



But even this is not the narrowest limit within which 

 the great Zambesi can confine itself at low water ; for on 

 reaching the bottom of the gorge at the Tshimamba 

 Cataracts, some 20 miles east of the Songwe, we found the 

 whole river raging tumultuously through a water-channel 

 which, at one place, was less than 25 yards in breadth 

 (Fig. 1). This place is apparently the only part of the 

 interior of the Batoka Gorge that was ever penetrated by 

 the white man until Mr. F. W. Sykes's expedition of 

 1902. His predecessor here was David Livingstone, who 

 in his second book of travels tells how he turned aside 

 on his eastward journey at the rumour of another great 

 waterfall, and was disappointed to find, not a second 

 Victoria Falls, but only a bold cataract, in which the 

 river drops about 20 feet. Nevertheless, the Tshimamba 

 also, were it rendered more accessible, would be well 

 worth visiting, if but to see the mighty river shrunk to 

 this little measure ; and one may expect, sooner or later, 

 to find it included within the " grand tour of the 

 Zambesi. " 



Although the surface of the basalt plateau falls steadily 



NO. 1883, VOL. 7$\ 



eastward, the Zambesi within its gorge sinks somewhat 

 more rapidly in the same direction, so that while imme- 

 diately below the Victoria Falls the river is barely 400 feet 

 below the lip of the gorge, this is increased to about 500 

 feet at the Songwe, to about 600 feet at the Tshimamba, 

 and to 800 feet at the place some 35 miles farther east 

 which we reached from the south bank. Aneroid observ- 

 ations showed a difference of more than 900 feet between 

 the level of the river at the foot of the Falls and Wankie's 

 Drift, which represents the descent of the water in passing 

 through the Batoka Gorge ; and until this steep gradient 

 is very much reduced the Zambesi must continue to deepen 

 its channel along the easiest lines before there is time for 

 it to straighten out the angularities of its course. 



The results attained by this selective erosion are 

 strikingly exemplified in the immediate surroundings of the 

 Victoria Falls. The wonderful Chasm, in places onlv 

 80 yards wide, into which the broad river is here precipi- 



From a photograph hy Mr. F. W. Sykes. 

 Fig. i.— The Gorge immediately below the Tshimamba Cataract. The 

 depth of the canon here is about 600 feet. The Zambesi, in the fore- 

 ground, is confined in a channel from 20 to 25 yards wide. Note how 

 the strong jointing of the b isalt governs the course of the river and 

 tends to produce zigzags in the low-water channel. 



tated, owes its chief features to the presence of an east 

 and west vein, probably a fault-plane, that cuts vertical!} 

 through the basalts. This vein, which I found to be well 

 exposed in the steep Recess or gully at the eastern end of 

 the Chasm, is partly filled with calcite and other soft vein- 

 stuff, and the rock adjacent to it is shattered and readily 

 decomposed. When the falls, in receding northward, struck 

 upon this vein, they readily hollowed out a transverse 

 trench across the whole breadth of the river, from which 

 the waters escape southward through a single narrow 

 channel. But, having passed this easy place, it is be- 

 coming increasingly difficult for the shallow river to support 

 a fall of its full width, and consequently the wearing back 

 of the lip is at present progressing most rapidly in a 

 comparatively narrow space at its western margin. Here 

 the " Leaping Water " pours a strong flood perennially 

 into the corner of the Chasm, and may eventually con- 

 centrate the whole of the river into its trough, unless, as 

 Mr. Molyneux has suggested, the deep oblique cleft that 

 is being rent across Cataract Island should gain precedence 



