[ 27° ]

arc doubtless capable of being much improved, so as toproduce greater effects, with a less quantity, and whatis of more consequence, with a less fall of water : andprinciples may doubtless be discovered, by which theirstructure may be regulated, and their power ascertained.

The importance of procuring commodious and cheapmeans of supplying the vast quantities of air which thesmelting furnaces require, induced me to examine theseveral accounts that have been published of these simplesubstitutes to bellows; and to make sundry experimentsfor bringing them nearer to perfection, and for establish-ing their laws of action.

SECT. I.

Account of the principal machines used for blowing air intofurnaces by a fall of water .

I. A simple pipe.

T H E first account I have met with of a machinefor propelling air into furnaces, by a fall of watercarrying down air with it, is of one at the copper or brassfurnaces at Tivoli near Rome, of which a descriptionand figure are given in the third number of the Philo-sophical Transactions, and in the Journal desfavans for theyear 1666.

A square wooden pipe, of considerable width, and openat both ends, is placed upright. A stream of water runsin at the top, and is discharged at the bottom ; and aboutthe middle of the height of the pipe a smaller horizon-tal one is inserted, which reaches to the furnace, and issaid to convey to it a strong blast of air.

From so imperfect a description, we can learn little ofthe nature of the machine, or of the manner in whichthe blast is produced. It may be presumed that the wa-ter,