234 rALAEOLITHIC IMPLKMEXTS. 



Tlie limaiides are mostly between six and a half and 

 eigkt inches in length, but some are smaller, down to 

 four and a half inches. Xone exceed one and three-quarter 

 inches in thickness. The outline is ovate or elliptical, and 

 the edg-e is sharj^ened all round. There is no mid-rib on either 

 surface, and Avhen the implement is viewed from the side 

 the edge is approximately straight, the inverted " S " cuiTes 

 frequently seen in European Acheulian implements not being 

 noticeable in these. They are very slightly worn, though 

 weathered, and therefore somewhat soft. It seems probable 

 that they were lying when found on their original floor, and 

 that their former covering has been removed quite recently. 



A trench was dug for a distance of about five feet into 

 the bank, in the hope of finding more implements of the same 

 type in place beneath the Adei soil. Two good implements, 

 quite unworn, and some others were found by the boy during 

 my absence near the face of the bank, but the rock floor rose 

 slightly further in, and no stony bed or other implements 

 were found in the deeper part. 



In a stretch of the spruit half a mile further uj) stream, 

 implements of many types are abundant. They are chiefly 

 pointed and of crude workmanship, but two similar to the 

 limandes of the former sjiot were found, and some flakes 

 and implements worked on one side, only comjiarable with 

 the Mousterian, but the minute scrapers of the Bushman 

 tj'pe are rare here as in the other streams mentioned above. 



In the upper portion of the Imbusine River, as also in the 

 Tegwan, which joins the Umguza below the confluence of the 

 Xoce, a very large variety of implements occur. 



Photographs of some of the implements described in this 

 paper are given on Plate XXX. 



The Rhodesian Museum contains a collection made by 

 AV. H. Kenny of about a score of implements of a remarkable 

 type, characterised by a very attenuated point with the butt's 

 edge sharpened. It is remarkable that, although I have seen 

 considerably over a hundred large palaeoliths in the field, I 

 have never encountered anything approaching this type. It is 

 probable that the specimens occurred as a single hoard, similar 

 to the one m the Imbusine just described. 



In conclusion, it may be stated that certain other streams 

 were mapped in some detail, but Avere not found to contain 

 stone implements. Of these may be mentioned the Umguza 

 below the confluence of the Xoce, where it flows in a sandy 

 bed between sandy banks, the Xoce and the Bembesi across 

 the Robert Block, besides other parts of the streams already 

 mentioned. Moreover, the larger palaeoliths are very rare 

 on the bare veld. 



In every place where implements were found to be 

 abundant, the stream in which they occurred flowed on a 

 rocky bottom between banks of black \\ei soil, at the foot 

 of which there was generally a bed of stony alluvium, from 

 which it is presumed that the implements came. 



