and the Economy of Fuel. 105 



Before I take my leave of this subject I would just 

 remark that the cause of the appearances observed in 

 the experiments may, I think, be traced to that prop- 

 erty of flame from which it has been denominated a 

 non-conductor of heat; for, if the different particles 

 of flame give off their heat only to bodies with which 

 they actually come into contact, the quantity of heat 

 given off by it will be not as its volume (and conse- 

 quently not as the quantity of fuel consumed), but 

 rather as its surface. And as the surface of the flame, 

 when fire-places are similar, is proportionally greater in 

 small than in large fire-places, the surfaces of simi- 

 lar bodies being as the squares of their corresponding 

 sides, while their volumes are as the cubes of those 

 sides, it is evident that, on that account, less heat in 

 proportion to the quantity generated in the combustion 

 of the fuel ought to be communicated to the boiler, 

 when the fire-place and boiler are large, than when the 

 process is carried on upon a smaller scale. 



There are, however, several other circumstances to 

 be taken into the account in determining the effects of 

 size in the machinery necessary for boiling liquids ; and 

 one of them, which has great influence, is the heat 

 absorbed by the masonry of the fire-place. This loss 

 will most undoubtedly be the smaller, as the fire-place 

 is larger ; but to determine the exact point when, the 

 saving on the one hand being just counterbalanced by 

 the loss on the other, any augmentation or diminution 

 of size in the machinery would be attended with a posi- 

 tive loss of heat is not easy to be ascertained. Provided 

 however that proper attention be paid to the manage- 

 ment of the fire, and that as much heat as possible be 

 generated in the combustion of the fuel (which may 



