[ 3” ]

which the water is afterwards to pass through. Thestream should enter at one iide, or be so managed, thatthe water in the reservoir or funnel may not be agitatedby it, or put into a spiral motion, which our experimentshave strewn to be very injurious.

In the bottom of the reservoir is to be made a round,hole, for admitting the upper end of what we have hi-therto called the funnel, but which may here be moreconveniently a cylindrical pipe, of copper or of cast iron,five or six inches in the bore, and seven feet long. Tothe end of this pipe is to be fitted a cullender,, about afoot long ; with the holes triangular, of half an inch eachside ; and fix or seven strips from top to bottom, at equaldistances,, preserved without holes, for admitting air topass down to the lower stteams. All the holes should bedirected downwards, that the streams may not be forciblyprojected against the sides of the pipe which is to receivethem, so as to have their velocity too much diminished.

If there are six of the perforated spaces in the cullen-der, the number of holes in each may be twenty, sothat the whole number will be one hundred and twenty..The side of each of the triangular holes being half aninch, the area of each will be the eighth part of a squareinch, and the sum of their areas will be fifteen squareinches. The quantity of water running through oneaperture of such an area, at the depth of seven feet anda.half under the surface, comes out on calculation aboutsix hundred and twenty-two gallons in a minute ; but.the real quantity will doubtless be much less than this,on account of the great friction of the water in passingthrough a number of small holes, and of the resistanceof the air, which increases in a very high ratio accord-ing to the increase of the velocity and enlargement of thessirface : it is in part to make up for these retardations.