MECHANICS AND USEFUL ARTS. 25 



work the engine for any considerable time by means of the com- 

 pressed-air apparatus of M. de Bergue. I soon convinced myself 

 that the heating of the cylinders went on so rapidly that this sys- 

 tem was inapplicable for any length of run. It was then that I 

 drew up a complete programme of experiments, the sum and sub- 

 stance of which was to establish a communication between the 

 boiler and the lower end of the exhaust-pipe, in order to supply 

 there a jet of steam or of water, and to force into the boiler the 

 elastic fluids, — steam or gases discharged from the cylinders by 

 the return stroke of the piston. I pointed out three combinations 

 to be experimented on in succession, according to the greater or 

 less difficulty found in completely cooling the cylinders. 



'• 1st. Injection of steam mixed with air. 



*' 2d. Injection of steam in sufficient excess to prevent the en- 

 trance of air. 



** 3d. Injection of water, instead of steam. 



** At first I supposed that the steam would carry along with it a 

 sufficient quantity of water to absorb the heat produced, and that 

 it would be condensed before reaching the cylinders. This idea 

 was incorrect. During the working with steam reversed, the 

 water ceases to be in a state of violent ebullition, and is only car- 

 ried over in small quantities ; and, besides, when the steam ex- 

 pands in issuing from the boiler, it dries, and the small quantity 

 of water brought with it is almost entirely converted into steam. 



*' The first experiment with a mixture of steam and gases drawn 

 into the cylinders did not give favorable results. With the injec- 

 tion of an excess of steam — a s}- stem which I characterized as an 

 inverted steam engine — more satisfactory results were obtained, 

 and it was found possible to work-with a moderate admission of 

 steam with light loads on moderate gradients, without burning 

 the packings, and without injuring the rubbing surfaces. We 

 have in France the example of a railway on which 200 engines 

 have only a cock for the injection of steam, and the substitution 

 of this for the gases drawn from the smoke-box has proved suffi- 

 cient to render the counter-pressure steam applicable for stopping 

 and shunting in stations, and for moderating the speed in the de- 

 scent of goods trains on gradients of one in 260. Indeed, the in- 

 jection of steam alone has been efl"ectually applied to light trains 

 on a short incline of one in 22. 



" But experience soon showed that the only general and complete 

 solution of the question is found in the injection of water. To 

 complete the absorption of the heat produced by the compression 

 in the cylinders, to force back the steam into the boiler, and to ren- 

 der the reversal of the steam an absolutely innocuous operation, 

 water is the only appliance. 



"When we speak of injecting water issuingfrom the boiler into 

 the cylinders of a locomotive engine, it must be borne in mind 

 that it is not water in the state in which it would flow from a 

 fountain ; it is at a high temperature when it issues from the 

 boiler, and rushes into space at atmospheric pressure. It enters 

 at once into ebullition, and becomes steam at 100° C, in quantity 

 corresponding to the heat employed. 



3 



