210 Annals of the South African Museum. 



I was hoping, as this rock-shelter had been but httle disturbed, 

 that we might obtain there evidence that would go to show whether 

 or not the former inhabitants there inhumed had manufactured 

 implements of palaeolithic style, in contradistinction to the neolithic 

 ones obtained in the sand-dunes, the littoral middens, or the rock- 

 shelters of the inland districts. Nor was I mistaken. 



Interspersed in all the layers, as deep as we went, was found a 

 great quantity of moderately large, uncouth chips, most of them with 

 at least one cutting edge, but which had flaked in a manner some- 

 what different from that with which I am acquainted. This I have 

 found to be due to the quartzitic condition of the rounded pebbles 

 used. Among these were found three which are, undoubtedly, of the 

 palaeolithic boucher type. 



The first is an axe-like cleaving irnplement, plane on one side, 

 having evidently been detached from the matrix at one blow; the 

 reverse is pluri-faceted, the lower part being produced in a broad, 

 slanting, bevelled edge resembling that of a carpenter's chisel, and 

 bearing traces of use. It resembles Fig. 57 of Plate VIII. The 

 quartzite used breaks readily into this shape. The implement is 

 126 mm. long, 85 mm. at its widest part, and 25 mm. in thickness. 

 The second is a boucher of oblong-ovate form, not unlike that of 

 Fig. 24 in Plate III. One side is the natural surface of a very flat, 

 almost plane, water-worn pebble of the kind used for placing over 

 the bodies : the other side has been obtained by repeated flaking, 

 which has produced many short facets resembling greatly secondary 

 chipping; this style of flaking was seemingly necessitated by the 

 composition of the matrix. The implement is 170 mm. in length, 

 110 mm. in width, and 31 mm. thick. The third example is a large, 

 lanciform scraper-flake of the same shape as those represented in 

 Plate XIII., Fig. 4. 



The presence of these three examples may not be thought a priori 

 to be sufficient evidence that the two styles of stone industry are the 

 handiwork of the Strand Looper aboriginals inhumed in this Cold- 

 stream shelter. But the size and shape of the flakes interspersed in 

 the layers, which were removed to a depth of nearly 16 feet, plainly 

 indicate that they are not of the small type of South African neolithic, 

 although some of the latter were also found with them ; they are the 

 by-product of moderately large bouchers, in spite of the scarcity of 

 the latter in the cave. 



There is also no doubt whatever that together with these flakes, 

 large and small, and on the same levels, were found mullers or 

 pounders of the type of, or similar to, those figured in Plate XXIII., 



