294: REPORT— 1905. 
distant from the Falls, where we were able, with some difficulty, to 
make a descent into the main gorge. Our next march carried us to the 
Mavangu streamlet, some 12 miles farther south-eastward, where we 
again reached the brink of the gorge and investigated its structure, but 
did not attempt to descend its precipitous walls. Thence, makirg a 
détour to avoid the intervening lateral gorges of the tributaries, which 
break backward farther and farther into the plateau as their distance 
below the Falls is increased, we journeyed for eight or ten miles to a 
‘camping-place on the nearly dry bed of the Mamba stream, [rom this 
‘camp we reached the main gorge and again descended into it, the walls, 
-here about 600 feet in height, being less steep than in the higher reaches, 
‘owing to the prolonged action of subaérial weathering. At this place, 
‘known to the natives as the Tshimamba, the swift Zambesi forms two 
‘cataracts, in the lower and larger plunging with a vertical drop of 20 feet 
into a narrow gully, not more than 25 yards wide, within which the whole 
‘river is confined except at flood-times. These cataracts were visited by 
David Livingstone in 1860 ; and this appears to be the only part of the 
‘Batoka Gorge (as we propose to name the cafion of the Zambesi below 
Victoria Falls) that was ever penetrated by the white man until Mr. 
'F. W. Sykes’s expedition of 1902. 
. East of the Tshimamba, although the walls of the gorge are no longer 
/so continuously precipitous and are probably scalable in most places, it 
becomes exceedingly difficult to reach its margin, owing to the wide belt 
‘of broken country seamed with impassable ravines that borders it on 
‘both sides. In our next march we swung northward for several miles to 
avoid this country before going east again to the Karamba stream, where 
‘we pitched our camp just above the spot where the little river drops 
; suddenly from its open valley on the plateau into a gloomy chasm only 
/ 15 or 20 feet wide at the bottom, with towering walls of basalt 300 feet 
in height. The Karamba, after passing through this chasm (which we 
propose to name Kalonga’s Cleft), flows in a zigzagging ravine for about 
five miles before joining the Zambesi. With considerable difficulty we 
managed to make our way to the confluence, but found, contrary to our 
expectation, that the basalts were still the only rocks exposed in the 
-Batoka Gorge. We then decided, as time was pressing and supplies for 
our carriers running short, to press eastward to the termination of the 
basalt country before again attempting to reach the gorge. In adhering 
to this plan we found that the basalts were continuous to Wankie’s 
Drift, so that we did not again touch the Zambesi until this place was 
reached, after four days of hard marching. The first of these days carried 
us some 14 miles N.E. from our Karamba camp to the banks of the 
Ungwesi River, the largest northern affluent of the Zambesi that we 
crossed ; on the second evening we encamped on the Gwemansi, another 
important tributary, about 16 miles, as the crow flies, farther eastward ;: 
thence, on the third day, our course lay south-eastward for about 13 miles 
to ’Ntoro ; and on the fourth day an easy march of 12 mies brought us 
down to the Zambesi at Wankie’s Drift, which was reached on July 20. 
Until we reached the Ungwesi not only was the country-rock composed 
entirely of the basaltic series, but the stream-beds draining from north- 
ward also showed no trace of detritus other than that derived from the 
basalts. In the bed of the Ungwesi, however, there were a few small 
pebbles of gramite and other igneous and metamorphic rocks, which 
denote that its head-waters probably reach hack northward beyond the edge 
