ON PNEUMATIC MACHINES. 349 



the cylinder is regulated by the action of a double revolving pendulum. 

 The piston is preserved in a situation very nearly vertical by means of a 

 moveable parallelogram, fixed on the beam, which corrects its curvilinear 

 motion by a contrary curvature. In the old engines, a chain working on an 

 arch was sufficient, because there Avas'no thrust upwards. When a rotatory 

 motion is required, it may be obtained either by means of a crank, or of a sun 

 and planet wheel,with.the assistance of a flywheel; this machinery is generally 

 applied to the opposite end of the beam; but it is sometimes immediately con- 

 nected with the piston, and the beam is not employed. The cylinder is usually 

 inclosed within a case, and the interval is filled with steam, which serves to 

 confine the heat very effectually. (Plate XXIV. Fig. 337.) 



The steam engines of Messrs. Boulton and Watt are said to save three fourths 

 of the fuel formerly used ; and it appears that only one fourth of the whole 

 force of the steam is wasted. Such a machine, with a thirty inch cylinder, 

 performs the work of 120 horses, working 8 hours each in the day. 



When the water producing the condensation is to be raised from a great 

 depth, a considerable force is sometimes lost in pumping it up. Hence, 

 Mr. Trevithick has attempted, as Mr. Watt had indeed long before pro- 

 posed, to avoid entirely the necessity of condensation, by employing steam 

 at a very high temperature, and allowing it to escape, when its elasticity is 

 so reduced by expansion, as only to equal that of the atmosphere: the air 

 pump is also unnecessary i;i this construction, and for a small machine, it may 

 perhaps succeed tolerably well. But there must always be a very consi- 

 derable loss of steam, and although the expense of fuel may not be increased 

 quite in the same proportion as the elasticity of the steam, yet the differ, 

 ence is probably inconsiderable. A great number of less essential alterations 

 have also been made in Mr. Watt's arrangements by various engineers, but 

 they have generally been calculated either for obtaining some subordinate 

 purpose of convenience, or for imposing on the public by a fallacious appear- 

 ance of novelty. (Plate XXIV. Fig. 338.) 



The force of steam, or of heated vapour, is probably also the immediate 

 agent in the astonishing effects produced by the explosion of gunpowder. 

 The initial elasticity of the fluid by which a cannon ball is impelled, ap- 



