[ 3°5 ]

through a circular one of equal area: some moreholes were made round the fides, in such positions, thatthe streams issuing from the higher holes, might nowherefall upon or coincide with those from the lower ones,but that the water might be uniformly dispersed throughthe whole cavity of the pipe. By this division of thewater it was made to fill a much larger bore than other-wise, and to produce as great an effect as the full quan-tity of water which the same pipe would otherwise haverequired ; insomuch that quantities of water which hadlittle effect in the common way of application, were bythis contrivance made to yield a strong blast.

This method is accompanied with an inconvenience,which often shewed itself in the course of the experiments,and which must be more considerable in the continuedworking of the machine. After it had acted vigoroustyfor some time, its action frequently abated of a sudden :the blast from the blowing pipe grew weak, and thegage sunk : sometimes its force increased again in a littlewhile, but for the most part it continued to diminishmore and more. The cause was discovered to be bits ofleaves and other like matters which the water had carriedinto the funnel, and which had in part stopt up the smallapertures. The remedy was obvious, letting the waterpass from the reservoir through a wire sieve whose holeswere much finer than those in the nose of the funnel ;and doubtless an expedient of the same kind wouldprove effectual for the largest machines. It is in all cafesadvifeable to have the water pass through a grating beforeit enters the funnel ; even the common large aperturesbeing sometimes choaked up by matters which the streambrings along with it. Where scantiness of water, or wantof so high a fall as is commonly required, persuade to thiscontrivance for procuring a more effectual division of it,

R r and