was as much as the bellows could support, the air was not forced thro’. I also made several experiments, by lowering the horizontal pipe two feet under the surface of the water, and raising it at different times to within four inches of the surface, and could not at any depth force the air thro’, whilst the engine worked; but on opening the steam-pipe, which is a pipe for letting the steam pass from the boiler whenever the engine stops, the bellows could then readily force the air thro’, tho’ the water boiled ever so strong, and seemingly made a surprising increase of steam. I had the leaden pipe to convey the air from the bellows, which was first put thro’ the top into the boiler, carried on the outside, and passed horizontally into it, about the height the water generally stands, that by opening a cock, fixed for the purpose close to the boiler, I could readily discharge all the steam lodged in the pipe; and by shutting the cock, and making small holes at three or four inches distance, I could almost find the point, where the air and steam met in opposition, cool air being strongly expelled thro’ one, and hot steam thro’ the other. It was also perceptible, that the air was impelled somewhat, tho’ not considerably, more forward by the addition of each hundred weight on the bellows: That the deeper the horizontal pipe was placed in the water, the less resistance was made by the steam: That in proportion as the heat of the steam was increased, by making the water boil more strongly, the resistance to the pressure of the air by the weight on the bellows became greater. It is a very doubtful matter, whether air forced thro’ boiling water would have answered the purpose intended: but I believe it was never imagined, that air could not be readily forced thro’, until proved by the foregoing experiments. The attempt, tho’ it has failed demonstrably in that point, has produced the same effect from another cause, as to saving coals, and throwing up more water. For, by the constant care, that was taken during the time of making these experiments, to measure the coals, to admit only a proper quantity of fuel to be laid on, and also to mark the time exactly it took in burning; the engine then did, and still continues to require eight bushels of coals less, in every 24 hours work, than it did before; and also, from the regularity of its stroke, to throw up more water; the same care being required from the engineer, who can have no pretence for consuming more coals now, than appeared sufficient during the time the experiments were making.