( 3° 8 ]

which consequently contained about fifty seven gallons onevery inch in depth. For admitting the water, two holeswere made in two opposite sides of the pipe, about ten incheshigh, with two sliders fitted to them, for occasionally vary-ing their height and consequently the quantity of waterreceived. On the outside of each hole was fixed an ironplate, perforated with numerous small holes, to keep backsuch matters as might choak up the throat: that the holesmight be sufficient to allow water enough to pass in, thestrainer was made wider than the aperture in the pipe,and bent to a semi-cylindrical form.

To the bottom of the funnel, enlarged as above men-tioned, was fitted a pipe six feet high, and in width fourinches by seven and a half. The lower end of this pipe wasinserted into the head of a large cask without a bottom,which was set in a tub above three feet deep, with threesupports under the lower edge of the cask to procure a spacebetween it and the bottom of the tub for the water to passfreely off. About eight inches under the orifice of the pipe,a round board, for the water to fall on* was hung by threecords, which passed up through the head of the cask andwere secured by pegs. At one side, a tin-vessel full ofwater was supported in the same manner ; and through afaucet, over the middle os this vessel, was inserted a glasstube thirty four inches long. At the other side was theblast pipe, about three quarters of an inch in diameter,

The machine being thus prepared, we proceeded to thetrial of it, expecting that the two streams, from theirsloping direction towards one another, would cross and*be daffied into drops, and carry down abundance of air,But in the effect we were greatly disappointed : the blastwas weak, and the gage rose to no considerable height,whether the wedge was dropt down or drawn up, so asLo suffer the water to pass in less or greater, quantity, inr

thin