[ 290 ]

The inconvenience may be prevented also, as effectually,and with more advantage in other respects, by makingthe air vessel of a very considerable depth below the sur-face of the stone : it may be sunk several feet into theground below the level of where the outward water runsoff, so as to have always a column of water in the vessel,of any height required, or of a height which shall secureagainst any air passing down to the bottom. This struc-ture would free the workman from any care about in-creasing or diminishing the aperture, or regulating theheight of the water. For if the deep vessel has an aper-ture in its lower part, large enough to discharge all thewater that can fall into it through the pipe in the top,or, for the greater security, a good deal larger, its mag-nitude being of no inconvenience ; if this vessel is funkin a pit of water up to the level of the stone, or to a cer-tain height above it; and if the pit has a drain sufficientto carry off what more water it may receive : we may besure that the water will be always high enough in thevessel, because the pressure of the water on the outsidewill keep it so ; and that the pressure of the air withinthe vessel will always keep it below the surface of thestone.

The air extricated from the water is always moist:when let off at a little way above the stone, I have

often observed it to leave drops like dew on any solid

body opposed to it. A small degree of moisture may

perhaps be of no disadvantage but such a degree as this

must doubtless be injurious, and render the air of less effi-cacy for animating the fire.

In the water machines of Dauphiny, inclined platesare said to be placed at the entrance of the pipe whichcarries off the air, to keep back the watery drops. M.Barthes proposes letting the air off into another vessel, in

which