[2S 9 ]

avoid either a depression or elevation of the water in theair vessel.

Though the depth os water be sufficient to resist thepressure of the air, it will be easily conceived, that if therewas no solid body to support the fall, the great force ofthe stream, falling from such a height, would puili downor dash about great part of the water in the bottom, sothat the air would get at the hole, and in part makeits escape with the water. It may be presumed thateven the drops of water, rebounding from the stone, andfalling down again, have a like effect, though in a lowerdegree: for drops falling through, the common atmos-phere into water, carry air with them, which afterwardsrises in bubbles, as may often be observed in heavy rains;and it is not to be supposed, that the drops ffiould nothere also carry into the water some of the compressed air,which surrounds them and is entangled between them.Though part of the air, which thus passes into the water,doubtless rises again in bubbles, as appeared in using theglass air vessel above mentioned; yet part may also bepushed so low, as to escape through the hole, and discoveritself by bubbles in the water on the outside of the vessel,which I several times observed before the water was drivenentirely out of the glass.

Mr. Barthes likewise takes notice of air being thuscarried down into the water by the drops, or introducedinto the cavities which they form in falling. In orderto prevent it, he recommends making a partition acrossthe tub, at the level of the stone, with only a hole at onefide, and this in the part most remote from the pipethrough which the water falls : the rebounding drops arereceived upon the board, and run off gently through thehole into the water underneath.

PP

The