228 SOLVENT POWER OF WATER ON SILICEOUS MINERALS. 
all manufacturers are familiar. Ci For the sake of experi- 
ment, I pulled down the four furnaces and rebuilt them, after 
having made between each and the kiln a deep pit as wide as 
each furnace, and only nine inches from front to back. 
About three feet of water was put into each pit, and was 
renewable from without. Some felspathic and siliceous mine- 
rals were placed in the way of the current, just inside of the 
kiln, and upon some of the arches a few articles-of ware were 
placed, that any action upon them might be observed. Be- 
low a full red heat, little effect was perceived, but at a heat 
above that of fused cast iron, a rapid solution of mineral mat- 
ter took place. The heat was continued ten hours. When 
the kiln was opened, more than a hundred weight of mineral 
matter, though in a very dense and refractory form, had been 
dissolved, and carried away in the vapor. The wall was 
eaten away, and presented a rough and quite unglazed surface, 
like loaf sugar partially melted by water, or as if eroded by 
some animal ; and nothing of the smooth glazed surface, 
which invariably attends the action of alkali on a siliceous 
surface. Some articles of ware, in the hottest situation, were 
partially eaten through ; but upon the uppermost arch, where 
the heat was only a full red, a curious phenomenon appeared. 
The articles there had received exterior to their own brown 
glass, and loosely encrusting it, a complete frosted coat of 
silica, having the appearance of a candied surface. It was 
manifestly a precipitation from the mineral vapor, and in fact 
a hoar frost of silica. There was probably from half an 
ounce to an ounce on each vessel, and several pounds, alto- 
gether, were thus precipitated ; but by far the greatest portion 
of the mineral vaporised was, as might be supposed, carried 
away in the current. Hence, this powerful action was ap- 
parently entirely due to the presence of water, there being 
at all times the same quantity of alkali present in the fuel, 
whatever that might have amounted to, producing no such 
effect; the experiment seems to establish, at very high tem- 
peratures, a powerful action of water on siliceous matter. To 
attribute the action to alkali, would not lessen the difficulty, 
