NA TU RE 



[November 30, 1905 



the main route until the termination of the basalts was 

 reached. These rocks proved unexpectedly to be con- 

 tinuous to Wankie's, although the " Batoka Gorge " (as 

 it is proposed to name this canon of the Zambesi) itself 

 ceases 6 or S miles above Wankie's, giving place to an 

 open valley with a broad shallow river sprinkled with 

 islets. 



On ferrying in a native " dug-out " across the Zambesi 

 at Wankie's we were met by Mr. H. F. Greer, of the 

 British South Africa Co., who holds charge in the dis- 

 trict south of the river. Here Mr. Sykes and Colonel 

 Rhodes struck southward to reach the railway at Wankie 

 Coal Mine, 35 miles distant, while Mr. Greer and myself 

 took a westerly course parallel to the Zambesi for about 

 bo miles, still traversing a basalt-country. We turned 

 aside twice in this westward journey in order to examine 

 the Zambesi valley at places eastward of (hose reached 

 from the north bank. One of these was at the confluence 

 of the Matetsi with the Zambesi, which is a little below 

 the termination of the narrow gorge ; and the other place 

 was about 15 miles farther west, where the strui lure of 

 the canon is not materially different from that which it 

 presents in the place where it had been last entered from 

 the northern side of the river. 



Mr. Greer having very kindly undertaken to escort me 

 to the headwaters of the Deka River, where previous 

 information had led me to expect that the base of the 

 Batoka Basalts would be found, we then took a south- 

 westerly course to Matetsi Camp. Crossing the railway 

 there, we continued our journey westward, southward, 

 and south-eastward across the upper part of the basin of 

 the Matetsi River, and after some days of hard trekking 

 struck the higher reaches of the Deka, only to find that 

 the interminable plateau-basalts over which the whole of 

 our route had hitherto lain were still the underlying rocks, 

 and that the surrounding country gave no indication of 

 structural change. It had been our intention to return 

 from Deka to the Falls by the old traders' route past 

 Pandamatenka and Gasuma ; but as the Bushmen reported 

 that, owing to the exceptionally dry season, no water 

 would be found in Gasuma Vley, this plan became im- 

 practicable, and we decided to follow a north-eastward 

 route, parallel to the Deka River for about 60 miles, to 

 the Wankie Coal Mine. Geologically, this proved to be 

 the most interesting part of my journey, and I therefore 

 spent four days at Wankie in further investigation, profit- 

 ing greatly from the guidance and kind hospitality of the 

 manager of the mine, Mr. J. M. Kearney. 



The basalts are cut off abruptly along the lower portion 

 of the Deka valley by a great fault striking approximately 

 north-east, which brings in the sandstones and shales with 

 which the Wankie coal-seams are associated. Some frag- 

 mentary plant-remains were collected from the Wankie 

 Coal-measures, and among these Mr. A. C. Seward has 

 recognised Vertebraria, which indicates that the deposits 

 are of Permo'-Carboniferous age, as indeed had been 

 previously surmised. Returning by rail from Wankie 

 Mine to Victoria Falls, I spent a few more days in ex- 

 amining the head of the gorge and its surroundings, and 

 was then compelled to leave Rhodesia in order to join 

 the Association at Johannesburg. 



The 600 miles of actual trekking that was accomplished 

 embraces a region of some 2000 square miles, of which 

 all except about 80 square miles east of the Deka is 

 underlain by the Batoka Basalts. The full extent of these 

 ancient lava-fields is still unknown, but, judging from 

 information that I obtained, it is likely to be not less 

 than 7000 square miles. Their thickness is also unknown, 

 but in the lower part of the Batoka Gorge, where the 

 original surface of the basalts must have been very con- 

 siderably lowered by denudation, the Zambesi has sunk for 

 800 feet further through these rocks without revealing 

 their base. In their prevalent characters they are remark- 

 ably uniform, consisting generally of thick bands of close- 

 grained dark-blue rock alternating with red, purple, or 

 ashy-looking amygdaloidal bands which mark off the sur- 

 faces of successive lava-flows. These less massive bands 

 frequently show a fragmental structure, and occasionally 

 pass into fine and coarse agglomerates suggestive of 

 volcanic tuffs or ashes ; but I think that this struc- 

 ture may represent the brecciation of the solid crust of 



NO. 1883, VOL. 7l\ 



the lava-flow before its onward movement had ceased, and 

 is not indicative of true ashes. In the whole course of 

 the journey I did not find any trace of an eruptive centre 

 or volcanic orifice, and the rarity of dykes was also re- 

 markable. Neither did I find any interstratified sediments 

 among the basalts in the country traversed, though there 

 appear to be some interstratified red and green beds of 

 shaly aspect in the railway cuttings of the Katuna valley 

 west of the Deka, which I had no opportunity to examine. 

 Like similar " plateau-basalts " in other pans of the 

 world, this immense mass of lava has probably had its 

 origin in "fissure-eruptions," by which a vast tract was 

 flooded under rapidli recurrent flows of high fluidity. 



We still lack definite information as to the geological 

 age of the Batoka Basalts ; by Mr. F. P. Mennell and 

 Mr. A. J. C. Molyneux they are regarded as most prob- 

 ably Tertiary, while Dr. S. Passarge correlates them with 

 the Loale Amygdaloid, which he considers to be of 

 Secondary age, perhaps Jurassic ; but the evidence for 

 either view remains inconclusive. 



The surface-deposits of sand, sandy limestone, cavernous 

 quartzite and hematite which locally overlie the basalts 

 in this part of the Zambesi basin, though of considerable 

 interest, must be dismissed for the present with the 

 remark that their mode of occurrence in this region is 

 not favourable to Dr. Passarge 's view that they represent 

 a definite order of events. The red sand (equivalent to 

 the " Kalahari Sand " of Passarge, and probably in part 

 to the " Forest Sandstone " of Molyneux) may, indeed, 

 denote a period of conditions different from those now 

 existing ; but the limestones and quartzites appear to me 

 to be due to purely local circumstances that still prevail. 



Let us now turn to regard briefly the physiography of 

 the region ; ' in which respect that wonderful natural 

 feature, the Victoria Falls, is, of course, the main pivot of 

 interest. 



Above the Falls, the Zambesi flows sedately in a broad 

 mature valley with low sides, excavated in the upper por- 

 tion of the Batoka Basalts. The gentle slopes of this 

 valley are partly buried under ancient desert-sands— the 

 " Kalahari Sand " of Passarge — and all the features point 

 to a long continuance of relatively stable conditions during 

 which the river has done very little erosive work. On the 

 brink of the Falls its bed is still about 3000 feet above 

 sea-level ; but at this point, suddenly, with a majestic 

 plunge, the Zambesi begins its impulsive descent from the 

 central plateau, and thereafter tears its way forcefully 

 across the mountainous margin of the continent, through 

 a succession of gorges alternating with relatively placid 

 reaches according to the variable endurance of the rock- 

 masses that lie in its path. It is to this rejuvenation of the 

 river at the present margin of the plateau, and its resultant 

 influence upon certain structures of the basalts, that, as 

 Mr. Molvneux has shown, we owe the magnificent Falls, 

 and not to any catastrophic rending of the earth's crust. 



The Batoka Basalts are traversed by a regular and 

 persistent system of close-set joints striking approximately 

 east and west, and are also occasionally fractured in the 

 same direction by still bolder vertical planes, probably 

 representing lines of fault, that are sometimes accompanied 

 by veins of calcite and other minerals. At the surface 

 of the plateau the basalts are much weathered, and this 

 weathering sinks deepest along the joints and fractures, 

 whereby these become the lines of readiest erosion. 



The rivers of this country are characterised by the 

 enormous difference that obtains between their volume in 

 the dry and in the wet seasons, a difference which affects 

 the great Zambesi proportionately almost as much as its 

 tributaries. During the shrinkage of the streams, the 

 greater portion of their broad rocky beds is laid bare, 

 and the water is confined within narrow gullies along 

 the joints and lines of readiest erosion, so that for more 

 than half the year it is in these channels only that there 

 is any wearing down of the stream-bed, while in flood- 

 time it is still along these gullies that the water is deepest 

 and most forceful, and that the chief portion of the detritus 

 (astonishingly scanty in these African rivers) is swept. 

 Thus, granting a sufficient gradient, these dry-season 

 channels become deepened and enlarged until they are 



1 'this portion of the report 1 

 chief features of the Gorge. 



ated by lantern slide 



ng the 



