By S. B. Dixon. 141 






















violent floods. The spot appears to have been the confluence of 
two or three streams, which united under, or at no great distance 
from, Knowle, and then took a south-easterly direction into the 
Kennett Valley, near Hungerford. Above these vallies was the 
piece of rising ground, where the gravel beds are situated, and on 
this there was probably a very early settlement, established from 
its contiguity to the streams which were so necessary for main- 
taining life, and were possibly used as a means of transit In this 
settlement many of the implements (which were for both domestic 
and warlike purposes) would have been made, and their fresh 
unbroken appearance and sharp edges indicate that they had had 
but little, if any, use, and had never been moved to any distance. 
‘They are almost entirely of flint, although one or two specimens 
have been found of greensand chert. These must have been 
brought from a distance, as no trace of chert is ever found in the 
neighbouring greensand of the Pewsey Vale. It is a substance 
Axminster being entirely formed of that material. No unworked 
chert is found with the gravels, but there are occasionally small 
pieces of sarsen stone, and the gravels bear evidence throughout of 
having a purely local origin. The question naturally arises, From 
what source did the water come that excavated these vallies, as 
he whole is now a waterless district? As already mentioned, 
fo the south-west are the high chalk downs which terminate in 
Martinsell Hill (nearly 1000 feet above sea level) and which form 
orthern boundary of the Pewsey Vale. The whole drainage 
se chalk hills now discharges itself into this broad open valley, 
at an earlier date it is very probable that such was not the 
The valley was then a mass of trees and jungle with no 
rainage, entirely different from its present configuration, and as 
e@ water would not have the same facilities of getting off in that 
rection it would look for outlets in other places. The chalk, too, 
vuld have been covered with a thick bed of clay-with-tlnts, of 
tich there are still remains, which would have prevented the 
ater from percolating through. The springs would then have 
orked out at a much higher level and above the district in which 
VOL XXXIL—NO. 6. 10 
