Vol. 65.] NEIGHBOURHOOD OP THE VICTORIA FALLS. 403 



It will be seen that worked stones are found on the floor of the 

 old Zambesi Valley above the Falls, not many feet above the level 

 of the annual flood ; and below the Falls, on both sides of the 

 gorge in which the river flows, 400 feet below ; further, that in the 

 tributary valley they are found embedded in the gravel of the 

 Maramba, in the dry bed and on the banks of that river, and on 

 the floor of the valley, and also on the ridge rising in the middle 

 of the tributary valley to a height of more than 100 feet. 



When the specimens from these localities are compared, the 

 difference between them becomes apparent. Those from the ridge, 

 retaining the rough coating, with the angles sharp and the flaked 

 surface weather-stained by long exposure, appear to be the refuse 

 left by fabricators of implements who were dealing with an 

 intractable material where there were no watercourses or sand by 

 which the waste material left behind could be rolled or polished. 



Many of the worked stones from the right bank of the Zambesi 

 have much the same character. Of the 10 specimens collected on 

 that side above the Falls, where the spruits are few and short, the 

 majority retain portions of the rough outer crust of the quartzite, 

 and all are sharp and unworn, with little or no surface-polish, the 

 majority having none. On the tongue of land below the Falls to 

 the south of the Hotel, out of 44 specimens collected and examined 

 35 retain their sharp angles, part of the rough outer crust remaining 

 on many of them, while others are polished more or less. 



The large proportion of sharp unworn specimens on the right 

 bank of the river is noticeable, especially on the tongue of land 

 below the Falls. From their position on this long narrow pro- 

 montory the artificially -worked stones must either belong to the 

 gravel of the Zambesi when flowing upwards of 400 feet higher 

 than at present, or must have been fashioned from quartzite- 

 fragments found on the surface at some time since the gorge was 

 cut out. The condition of the flaking is all in favour of the latter 

 alternative. 



Passing to the other side of the Zambesi, on the flat south of 

 the Maramba and draining to it, the majority of worked stones 

 retain their sharpness, but in a smaller proportion. Of 38 speci- 

 mens, 22 are unworn and sharp and 16 waterworn, but none are 

 polished. 



On that part of the left bank which drains directly to the 

 Zambesi, comprising ground between 3 and 4 miles long above and 

 below the Falls, 130 artificially-worked stones were collected, all, 

 with one exception, of the local quartzite. Of these, 59 are sharp 

 in the angles of the chipping, some retaining the rough outer coat 

 of the stone, and 71 are more or less waterworn and polished. 

 The polish of some is remarkable, others are less polished, and 

 some not polished at all. In one or two specimens the angles of 

 the chipping have been blunted after the polishing. The proportion 

 of worn to sharp specimens is here reversed. While on the right 

 bank 80 per cent, retain their sharpness, on this part of the left 



