118 Mr. D. Gilbert OH the 



existed in the form of hydrogen, at the temperature 1500° above 

 the surrounding medium, it would require an opening of two 

 square feet and three-quarters, almost quadruple to the former 

 case. 



The narrowest part of the flues in some furnaces consuming 

 coal at the full rate of two bushels, or 168 pounds in an hour, 

 is about 200 square inches, just double to the aperture found 

 by theory. It would, therefore, in all probability be sufficiently 

 large, if the fuel were pure carbon or even coke, but is scarcely 

 so perhaps when fresh coal abounding in hydrogen, and gene- 

 rally not free from water, must be continually added to the 

 fire. It seems probable that the heat of furnaces would be 

 more uniformly sustained if the flues were made larger and 

 regulated by dampers, especially as no allowance has been 

 made in the investigation, Tor those retarding powers encoun- 

 tered by all fluids in passing through tubes, and in turning 

 angles. It may further be remarked, that in all flues the 

 aperture should be largest where the temperatures has attained 

 the greatest range. Having given the consumption of coal 

 and the temperature, these formulae extend to all closed fur- 

 naces ; but it does not seem possible to apply general theorems 

 to the case of heated gases ascending from open fire-places, 

 because of the uncertain quantity of common air necessarily 

 blended with them : the following observations, however, appear 

 naturally to present themselves. 



When currents of air find their way down a flue, in conse- 

 qeunce of reverberation from adjacent buildings, or in conse- 

 quence of a direction inclined to the horizon, which may have 

 been acquired by passing over more lofty obstacles, reme- 

 dies must be sought by elevating the stack, or by the use of 

 mechanical contrivances. When ample space is required 

 round a fire for culinary purposes, the flue must continue all 

 the way of a proportionate size, and so terminate, or the whole 

 of the elastic fluids will not be able to ascend, and to escape 

 into th^ air. But, when the only purpose of a fire is to dif- 

 fuse warmth through an apartment, the line of communication 

 into the chimney should be reduced to much narrower limits, 



