96 TRAXSACTIONS OF THE [dEC. 21, 



My own researches have convinced me of the impossibility of 

 using small cylinders with any degree of economy of steam unless 

 they are compounded so as to permit between 8 and 10 expan- 

 sions. 



In the case of non-condei>«ing engines, the boiler pressure 

 should be about 135 to 140 lbs. by the gauge, and in the case of 

 all engines the speed should be as high as is consistent with 

 safety. 



Much of this waste is due to initial condensation of the steam 

 on entering the steam cylinder, which was perceived by Watt, 

 and partially remedied by him. 



The quantitative law of this condensation, and its influence 

 upon the most economical expansions of steam, I have endeavored 

 to formulate in a paper in the Journal of the Franklin Insti- 

 tute for March, 1884, and showed in the same journal, August, 

 1880, how narrow were the limits of Marriotte's law for the pro- 

 fitable expansion of steam, even neglecting the initial conden- 

 sation. 



Taking matters as they are, the most economical engine used 

 for the ])urpose of driving dynamos at the late Electrical Exhibi- 

 tion of the Franklin Institute required about 30 lbs. of steam at 

 90 to 100 lbs. pressure, and the most economical boiler evapo- 

 rated about 8 lbs. of water per lb. of anthracite coal at the same 

 pressures. That is, an indicated horse-power required 3| lbs. of 

 average anthracite. 



It can be assumed, with close approximation to average correct- 

 ness, that fifteen per cent of the indicated horse-power is lost in 

 the most direct method of transmission of power from engine to 

 dynamo. 



So we can say that one utilizable electrical horse-power per 

 hour may, in good practice, be obtained 



H\^ X m = 4.tV-o- IIjs. of coal 

 (such as is sold in the open market as chestnut anthracite), and 

 neglect the loss of electrical energy in the conductors. 



The carbon equivalent of the coal used was 91 per cent by 

 weight. 



Assuming 14,500 British units as the heat j)er lb. of carbon, 

 we have 



4.90 X .91 X 14,500 = 64,055 B. units of heat. 



Assuming the mechanical equivalent of one British unit as 

 774.1 ft. lbs., we have very nearly 2,558 B. U. for one horse- 

 power per hour. Dividing the last by the first we find that 

 nearly four per cent of the power latent in the coal appears as 

 electrical power in the circuit. Ninety-six per cent of our poten- 

 tial energy is lost; principally in the steam engine. 



