THE WATT AND THK TRKVITIIICK ENGINKS. 179 



to the Cornish mines, in coal alone, of 80,000. yearly. 

 In addition to this, the increased power of the engine 

 lessened the first cost by at least one-half. 



The national importance of such weighty facts calls 

 for further corroborative proof, for we can scarcely 

 believe that two atmospheric low-pressure steam-engines, 

 made before the time of Watt, could be altered so as to 

 perform more work, and at a less cost than the Watt 

 engine, by an ingenious supply of higher steam pressure 

 from Trevi thick's boilers, together with the Watt air- 

 pump and condenser. The following words from Watt 

 are descriptive of his practice, though contrary to his 

 patent claim : 



" At a very early period, while experimenting at Kinneil, he 

 had formed the idea of working steam expansively, and altered 

 his model from time to time with that object. Boulton had 

 taken up and continued the experiments at Soho, believing the 

 principle to be sound, and that great economy would attend its 

 adoption. 



" The early engines were accordingly made so that the steam 

 might be cut off before the piston had made its full stroke, and 

 expand within the cylinder, the heat outside it being maintained 

 by the expedient of the steam-case. But it was shortly found 

 that this method of working was beyond the capacity of the 

 average enginemen of that day, and it was consequently given 

 up for a time. 



" ' We used to send out,' said Watt to Eobert Hart, * a cylinder 

 of double the size wanted, and cut off the steam at half-stroke.' 



" This was a great saving of steam, so long as the valves 

 remained as at first ; but when our men left her to the charge of 

 the person who was to keep her, he began to make, or try to 

 make, improvements, often by giving more steam. The engine 

 did more work while the steam lasted, but the boiler could not 

 keep up the demand. Then complaints came of want of steam, 

 and we had to send a man down to see what was wrong. 



" This was so expensive, that we resolved to give up the 



N 2 



