[ 3 ° 3 3

These experiments, and several others I have made onthe same subject, are not so conclusive as could be wished.They seem to shew that it is more eligible, to have theentrances for the air in one horizontal plain, than in twoplains above one another ; and either above, or at somedistance below the jet, than immediately under it : Thatthey ought to be of greater magnitude than in some of themachines described in the first section, particularly in thatof Lead-hills, whose air-holes, taken all together, are notof half the area of the space in the pipe which the air hasto fill. They ought at least to be of an equal, or rather ofa double extent, that the air may enter the more freely.

Experiments of the proportional bores of the funnel and pipe.

We have already seen, that unless the throat of thefunnel is less than the pipe, the quantity of air carrieddown will be inconsiderable; and that by lessening it furtherthan to a certain point, the effect is also diminished ordestroyed. To hit this precise point is not perhaps possi-ble ; and the point which is the most perfect proportionfor one height of water, cannot be so for any other, anincrease of the pressure disposing the jet to spread more andfill a larger bore.

It appears from some experiments already mentioned,that when the whole height of the fall of water is fifteenfeet, the height of the pipe ought to be nine feet, and thatof the funnel six. This being as low a fall as these kindsof machines have been generally erected for, and as high aone as is generally to be expected in this country, I madeseveral trials for adjusting the proportions to those heights ;using for the funnel a tapering copper pipe, into the lowerend of which were occasionally inserted smaller pipes ofdifferent bores.

By