SCIE feed-water to about hal: that proportion, for ratios of expansion approximating six; from one-third to about one-tenth, a t a ratio of five; and apparently from twenty to ten per cent at 4.4. In this first case, also, the jacket gives best results, with 110 pounds of steam, when the ration of expansion approximates six. When the stearn pressure falls to approximately eighty pounds, the best work of the jacket occurs at a ratio not far Prom 4.75; while, at a pressure of fifty pounds, the

more »

... ue of the jacket increases through the whole range of the experiments, and not only so, but the indications are of probable improrerrlent indefinitely in the direction of increasing expansion. The highest efficiencies, however, either wit11 or without the jacket, are found, in this case, at the lowest ratios adopted, and indicate a maximum value at about 3 25. The ratios of expansion for rnaximum efficiency of fluid, in the other cases, are for 110 pounds, about fire, and for eighty pounds, about 3 5 . Similarly studying the performance of the condensing engine, we find that the best work is done, whether jacketed or not, a t about a ratio of expansion of ten (at a steam pressure of 110 pounds), but that the jacketed engine reduces the internal wastes from fifty per cent at highest ratios, and from one-fourth a t the loulest ratios, in the case of the unjaclreted engine, to five per cent, and, in some cases, probably to within the magnitude of the errors