[ 3 00 J

the bore. With funnels of two and three feet, the gagerose more, and the jet spread, though it did not appear tofill the pipe, till it had reached about half way down tothe bottom. Funnels of five and six feet produced a strongblast, and kept the gage high, the jet filling the pipe be-fore it had fallen a foot below the throat of the funnel.

On many repetitions and variations of these experiments,I have not observed that the jet spread sufficiently with lessthan a fall of five feet. With a fall of sixty-four inches,the gage rose more than five times as much as with one ofsixteen inches, though the quantity of water which run inthe first cafe was only double to that in the latter, viz. asthe square roots of 64 and 16 : from whence it is plainthat the above differences do not depend entirely on thedifferent quantities of water which run through funnels ofdifferent heights, but in great part on its different velocity.Some other experiments seemed to confirm this point: forhaving used short funnels ib much wider than the high ones,that the quantity of water discharged by the former wasequal to or greater than that by the latter, the short neverproduced so strong a blast, or raised the gage so far, as theothers.

Being satisfied of the advantage of having the funnel ofvery considerable height, I in like manner varied the lengthof the pipe. Having made a mark at the part where thegage rose to when the funnel was-five feet, and the pipeseven, I added to the pipe about a foot more : the gagescarcely rose any further. A foot being cut off from it,the gage fell a little : two feet being cut off, it fell con-siderably ; and the retrenchment of another foot made themachine of little effect, the gage sinking almost to the bot-tom, and the blowing pipe yielding but a weak current ofair. The pipe, thus reduced to four feet, was tried witha funnel of near eight feet: in this cafe there was no blast

at