Append),'., 



57 



flakes. These Implements and debris bad evidently been dag out in the 

 excavation of this tank. 



Here, the ground is not altogether made up of gravel and breccia as is 

 the case to the south of Cupedoo, but consists of alluvial sandy and 

 clayey deposits of a grey and yellowish-grey color with two or three layers 

 of quartzite gravel and shingle as well as lateritic gravel. Rounded and 

 sub-angular fragments of quartzite are also irregularly scattered through 

 the most supeificial layer of alluvium. This tank,, like the first described 

 and many others in the neighbourhood, is not an ordinary bunded tank or 

 reservoir of water formed by throwing a bund across a shallow stream- 

 valley, but is merely a pond made by digging out the earth for two or 

 three feet from a natural depression, while the earth thrown out at the 

 sides helps to retain more water. It will be thus seen that the raised 

 banks of earth and gravel whence my Implements were procured are not 

 likely to be made of material brought from a distance, as is sometimes the 

 ease in an ordinary tank bund, and therefore the conclusion is that the 

 weapons in question have either been made on the shores of the tanks or 

 that they were dug out with the eanh. The latter supposition seems 

 the more likely one, for the few Implements found at both tanks were 

 scattered about very irregularly among the other stones of the gravel, and 

 required very long and careful search before they were recognized. In 

 coloring and weathering too, they were not to be distinguished from the 

 other stones lying about : the distinctive characters being the regular form 

 and, more particularly, the plane-edge all around the Implements, the latter 

 character being unknown in any naturally -formed fragment of rock. Be- 

 sides, in the case of the tank under description, the ground for long dis- 

 tances from it was thickly strewn with rounded and other fragments of 

 quartzite without, as far a3 I could see, and my research was extremelj 

 close, a single flake-like fragment, or Implement any where except 

 near excavated ground. 



A short distance further north of this tank, nearly two miles N. by E„ 

 of Cupedoo, and at the southern end of an irregularly rocky ridge, there 

 is another small and shallow pool close by which, at the time of my 

 visit,! the villagers were digging a* bowri 3 (well) ;and among the shingle 

 thrown out was a chipped fragment of quartzite. The layer of shingle 

 whence this had been dug is eighteen inches below the surface and would 

 nearly correspond to the dept h of the gravel and shingle from the bottoms 

 of the tanks already referred to. 



The next locality (3) of Stone Implements is a most interesting one 

 in so far as it seems to have been a place of manufacture. It lies 

 towards the north end and on the west side of the rocky ridge mentioned 

 above, two and three quarter miles N. by E. of Cupedoo. The flat ground 

 at the foot of the ridge and extending nearly to the stream flowing south- 

 wards to the Narnaverum river, is formed by a deposit of lateritic sand,, 

 gravel, and breccia, partly overgrown with scrub-jungle. Sub-angular 

 and rounded fragments of quartzite are frequent on the surface, but, as 

 far 1 could see, there are no Implements. This general level of the ground 

 is, how-ever, broken by a shallow depression ; the result of the scouring 

 out or denudation of an extensive patch of the lateritic deposit down to 

 the underlying gneissicor crystalline rocks. The thickness of the deposit 

 is very variable owing to the irregular surface of the rocky ground on 

 which it rests, but it may be taken on an average as between one and 

 three feet. 



i October 1862, 



