ajl X DefcriftUn if a New Steam Engine. 



Plate X. exhibits a projedtion of part of this engine, the boiler and the working parts 

 being left out. RSUT reprefent a frame of wood. A is the largel- cylinder, of which 

 the pifton-rod is kept vertical by a wheel I, which runs between parallel plates with a very 

 fmal! degree of fiiake. The rife and fa'l of this wheel moves the arm N O of a crank, 

 which drives the fly P Q_, and gives motion to the pumps or other working parts. B is the 

 fecond cylinder open at top, the pifton-rod of which is kept vertical in lii<e manner by a 

 wheel K. A connexion is formed between the two piftons by the lever L M, which caufes 

 them to make their (trokes together. D is a valve in the pifton of A, which, by means of its 

 tail beneath, is opened when that pifton is down, and continues fo during the whole afcent, 

 until it becomes again clofed by the preflure of its crown againft the top of the cylinder. 



Suppofe the pift:ons to be near the tops of their refpeftive cylinders, and the valve D to be 

 open : if C be opened, the fteam will pafs through the cylinder A, and enter B by the pafTagc 

 E, driving the air before it through the valve G. If then the valve D be clofed by raifmg 

 the pifton or otherwife, and an injedion of cold water be made through F into B, the fteam 

 in that cylinder and beneath the pifton of A will be condenfed, and both piftons will de- 

 fcend ; that in A by the a£lion of the fteam, and that in B by the preffure of the atmofphere. 

 Near the termination of this defccnt the valve D will be opened, and the pifton in A will be 

 left at liberty to rife, in confequence of the equal action of the fteam on both fides, at the fame 

 time that, the valve C being ftiut, there will be no farther fupply from the boiler. The fteam 

 ■which occupied the fpace above the pifton in A will expand through E, and affift the rife of 

 the pifton in B. If we overlook the effe£l of refrigeration in the paftage E, and againft the 

 pifton atB, the fteam will begin to zOl beneath this pifton with a force not greatly different 

 from its laft a£lion upon that in A. It would not be eafy to calculate the whole effect of 

 the fteam in its fecond application, not only for want of an accurate ftatement of theelafti- 

 city of this fluid during its expanfion ; but becaufe the pradical effefl of the temperature of 

 the fecond cylinder in diminifliing that force is ftill more difficult to be appretiated. I fhall 

 therefore only obferve, t^hat the pifton in B having defcended fo as to touch the furface of the 

 ■injeftion water, and the elaftic fluid reding upon it, not only caufes that air but a portion of 

 the hot water to pafs through the valve W, which it carries up and difcharges through a 

 conveyance at the upper part of the cylinder to feed the boiler. The fteam which follows it 

 through E, drives all the elaftic fluid through the valve of that paflage, whence it remains in 

 B till the next ft.roke carries it off. The quantity of water which fhall rife above the pifton 

 through W, may be regulated by the height of the external furface of the ciftern into which 

 the pipe G difcharges itfelf. The excefs of the a£lion of the fteam beneath the pifton in B, 

 beyond the weight of that pifton and the water it carries, will be an acquifition to the power 

 •of the engine. When we coarider the fafts, we may obferve, that, if B were too fmalj, the 

 action of the fteam under its pifton would approach more nearly to equality with the laft 

 a£lion in A; but that the ftroke would be infufBcient to carry oiT the elaftic fluid, and the 

 addition to the force of the engine inconfidcrable ; and on the other hand, if it were too large, 

 the greateft part of the rihng ftroke would be performed againft the prelFure of the atmo- 

 ifphere, with very little affiftance from the expanded fteam. Whence it follows, that there is 

 a certain definite proportion under like circumftances of preffure, temperature and work, 

 between the two cylinders, which will produce a greater effect than any other. To deter- 

 mine this would be the obje£l of computation, if the data were at hand-; but it is probable 

 9 that 



