where it was reduced to water, so that a vacuum was left in the upper part of 

 the cylinder. The steam from the boiler passing below the piston, pressed it 

 upward with such force, that it lifted a weight of eighteen pounds hung from 

 the end of the piston-rod. When the piston reached the top of the cylinder, 

 the cocks L and N were closed, and the cock C opened. All communication 

 between the cylinder and the boiler, as well as between the cylinder and tho 

 condenser, were now cut off, and the steam in the cylinder circulated freely 

 above and below the piston, by means of the open tube D. The piston, being 

 subject to equal forces upward and downward, would therefore descend by its 

 own weight, and would reach the bottom of the cylinder. The air-pump 

 piston meanwhile being drawn up, the air and the condensed steam in the 

 tubes F and G were drawn into the air-pump I, through the open horizontal 

 tube at the bottom. Its return was stopped by the valve M. By another 

 stroke of the air-pump, this water and air were drawn out through valves in 

 the piston, which opened upward. The cock C was now closed, and the 

 cocks L and N opened, preparatory to another stroke of the piston. The 

 steam in the upper part of the cylinder rushed, as before, into the tubes F and 

 G, and was condensed by their cold surfaces, while steam from the boiler 

 coming through the pipe S, pressed the piston upward. The piston again 

 ascended with the same force as before, and in the same manner the process 

 was continually repeated. 



The quantity of steam expended in this experimental model in the produc- 

 tion of a given number of strokes of the piston was inferred from the quantity 

 of water evaporated in the boiler ; and on comparing this with the magnitude 

 of the cylinder and the weight raised by the pressure of the steam, the contri 

 vance was proved to affect the economy of steam, as far as the imperfect con- 

 ditions of such a model could have permitted. A larger model was next con- 

 structed, having an outer cylinder, or steam case, surrounding the working 

 cylinder, and the experiments made with it fully realized Watt's expectations, 

 and left no doubt of the great advantages which would attend his invention. 

 The weights raised by the piston proved that the vacuum in the cylinder pro- 

 duced by the condensation was almost perfect ; and he found that when he 

 used water in the boiler which by long boiling had been well cleared of air, 

 the weight raised was not much less than the whole amount of the pressure 

 of the steam upon the piston. In this large model, the cylinder was placed 

 in the usual position, with a working lever and other apparatus similar to that 

 employed in the atmospheric engine. 



It was in the beginning of the year 1765, Watt being then in the twenty- 

 ninth year of his age, that he arrived at these great discoveries. The experi- 

 mental models just described, by which his invention was first reduced to a 

 rude practical test, were fitted up at a place called Delft house, in Glasgow. 

 It will doubtless at the first view, be a matter of surprise that improvements 

 of such obvious importance in the economy of steam power, and capable of 

 being verified by tests so simple, were not immediately adopted wherever at- 

 mospheric engines were used. At the time, however, referred to, Watt was 

 an obscure artisan, in a provincial town, not then arrived at the celebrity to 

 which it has since attained, and the facilities by which inventions and improve- 

 ments became public were much less than they have since become. It should 

 also be considered that all great and sudden advances in the useful arts are 

 necessarily opposed by the existing interests with which their effects are in 

 conflict. From these causes of opposition, accompanied with the usual influence 

 of prejudice and envy, Watt was not exempt, and was not therefore likely sud- 

 denly to revolutionize the arts and manufactures of the country by displacing 

 the moving powers employed in them, and substituting an engine, the efficacy 



