134 The Paleolithic Implements and Gravels of Knowle, Wilts. 
Two features noticeable on the surface of a considerable number, 
both of worked and unworked flints, call for special remark. One 
is the highly polished appearance of the stones. They are seldom 
completely polished, for bands of varying hardness in the stone 
show varying degrees of polish, and the high gloss is most frequently 
produced upon the later fractured surfaces. It should be mentioned 
that these polished stones come more particularly from the bottom 
of the pit, where the gravel is rather smaller, and the flints are 
embedded in a quantity of fine sand. 
Many suggestions have been made as to the probable cause of 
the highly glossy surfaces. The most obvious explanation is that 
they are due to the action of wind-blown sand, but while to us this 
would fully account for the phenomenon, the unusual brilliancy of 
the surface has caused other suggestions to be put forward. 
Another explanation, which has met with considerable support, 
is that a varnish of soluble hydrate of silica has been left upon the 
flints. To this solution of the problem there are, however, important 
objections. This soluble hydrate, as is well known, can be dissolved 
in heated water in the presence of an alkali. It is obvious, 
then, that the action of a warm caustic potash solution would 
effectively destroy a varnish of this nature. The experiment thus 
suggested, we have carried out, and have in our possession a small, 
extremely glossy stone, one end of which has been immersed in 
boiling potash for some time. It is still absolutely impossible to 
distinguish one end of the stone from the other, a fact which should 
conclusively prove the absence of any soluble hydrate of silica. 
A further most important objection to the silica varnish theory 
is afforded by a careful survey of many of the flints in question. 
It seems barely conceivable that a flwid of any kind could touch 
certain projecting ridges and points, without glazing also the de- 
pressions which lie between them. That the hollows have usually 
escaped, while projections have been affected, is well shown on 
many of the specimens, and is to our minds a very telling argument 
against the glazing by a secondary deposit. 
1 See specimens Nos, 21, 99, Oi, 114, in North Wilts Museum. 

