[ 288 •]

without a bottom, seven or eight inches, of its lower partbeing immersed into a tub full of water. As soon as themachine began to play and the gage to rife, the waterwithin the glass funk lower than that in the tub on theoutside; and the depresiion of the water and rising ofthe gage were, as nearly as could be judged, equal, andkept pace with one another. In a little time the waterwas forced quite out of the glass, and the air followingit rose in bubbles to the top of the tub.

The bottom of the air vessel ought therefore to befunk at least as much below the level where the externalwater pastes off, as the gage is expected to rife ■, forotherwise, before the air is sufficiently compressed toraise the gage to the due height, it will force all thewater out below, and in part escape itself by the sameaperture. Hence the depth of the air vessel, in any of thesemachines where the water has a free passage at the bottom,gives a power which the force of the blast in that machinecan never be made to exceed:: thus atLead-hills, the waterbeing only of the height of two feet from the bottom of thevessel to the level of the bank where it runs off, the air cannever be compressed further, than to be able to support acolumn of two feet of water, or to raise the gage to thatheight; whereas in the machine of St. Pierre, the com-pressure is about two thirds greater.

The sinking of the water in the air vessel may indeedbe prevented, by making the aperture at the bottom,through which the water is discharged, of such a size,that the pressure of the air may be able to drive throughit no more water than is received at the tap. But such anadjustment would be apparently very difficult; and tho’ itshould be exactly hit, yet, if the quantity of water receivedwas not always the fame, it would scarcely be possible to2 ' ' avoid