The nature of atmospheric pressure is so well understood at the present day, that a minute explanation of the operations of this engine is unnecessary; but its losses in the first stages of its structure deserve observation.
The pressure of the air seldom exceeds 14i lbs. per square inch; and sup posing the area of a piston in one of these engines to have been 100 inches, it would have lifted at each stroke 1475 lbs. of water in the mine, to a height equal to the length of the cylinder in clear cavity. But an enormous deduction from this power was to be made, especially in the first engine. The mode of effecting a void in the first instance, was by throwing cold water on the cylinder when filled by steam ; at a subsequent stage, this mode was abandoned, and the cylinder was surrounded by cold water, by means of an outer cylinder being adapted to it,—the space between them becoming the condensing medium. Superior as this latter method was to the former, the result, in either case, was necessarily attended with much loss and inconvenience. From the cylinder being placed above the boiler, it was obviously impossible, in the original con trivance, to protect the boiler from being splashed by the water thrown on the cylinder, and so far cooled as that a considerable portion of the steam therein must have been itself condensed and rendered ineffective ; and in the improved means, the water which cooled the steam became likewise heated in the pro cess, and, consequently, unfit to produce the effect which it was introduced into the concentric space to perform. The formation of a perfect void, even then, could neither be certain nor instantaneous. When the void was imperfect,. the vapour that remained in the cylinder resisted the weight of the atmospheric column ; and if this resistance amounted to three or four pounds on the square inch, so much, of course, was to be deducted from the total pressure of the air on the piston. Besides, the water from which the steam was generated contained more or less air, which boiling disengaged, and this entered with the vapour into the void, when made; the steam with which it was mixed could be con densed, but not so the air, which, remaining in the cylinder, prevented the fall of the piston, by filling the space under it to a greater or less extent ; in this condition the engine was called wind-logged. These defects, in combination
with the great amount of friotion and expenditure of time, caused necessarily a vast loss of power. But nevertheless, the great superiority of the atmospheric engine, as thus contrived by Newcomen, to Savery's engine (as well as the entire difference in respect of principle,) will be at once seen. Savery's was really an engine which raised water by the elastic force of steam ; but Newcomen's effects that object by the pressure of the atmosphere alone, steam being used simply as the means of readily forming a void into which the atmospheric pressure impels the first mover. There is, besides, in this engine, none of the danger incident to the use of a highly elastic, and, in Savery's engine, not easily manageable force ; and as the heat required is considerably less, so also is the quantity and the expense of fuel. The power, too, of the atmospheric engine, is almost boundless, being restricted only to the strength of the materials of which it is composed ; and the form of it renders it applicable to almost every mechanical purpose, by converting the reciprocating motion of the working beam into a motion of any required kind; which, in Savery's engine, was not so promptly attainable.
By sheer accident (a hole in the piston of the steam cylinder letting in some of the water which was constantly kept above it, to assist its air-tightness) the mode of creating a void, by the injection of cold water into the cylinder itself, was at a subsequent stage discovered ; and this fortunate discovery suggested, also, a method of regulating the speed of the engine, when the weight on the pumps was variable, or the engine working against a resistance beneath its power,—a larger or a•tnaller quantity of injection water thrown into the cylinder, producing a less or more perfect void, corresponding with the extent of condensation.