79 



are heavier than atmospheric air. If they have not passed through a 

 sufficient, quantity of coals to be saturated with carbon, this may 

 be the case in the air-furnace, and when flaming fuel is burned is 

 so in every case. Were the mixture nitrogen and carbonic acid, 

 s would be 1.09 and 0, 45*: If on the other hand tlie acid is 

 changed into oxide the chimney cannot be too long, and the 



draught must be augmented because I. (-) is greater than f.( '^t' ') 



therefore the difference of their logarithms is negative and the 

 less its coefficient, the greater is the whole. It might however 

 be questioned how far the absorption of caloric by the carbonic 

 oxide in its formation might operate by diminishing 0, but as this 

 is much more diminished than the logarithms are, there seems 

 reason to concUide that on the whole its presence is advanta- 

 geous except from the increased consumption of fuel. The up- 

 per part of the furnace must therefore be carefully closed as the 

 admission of air there is injurious, by lessening the quantity which 

 passes through the coals, by burning the carbonic oxide, and by 

 lowering the temperature of the ascensional column. The force 

 of draught is never very considerable, I observed it in a rever- 

 beratory, where of course the chimney was nmch hotter than 

 it can ever be in the class(o) of furnaces which we are at pre- 

 sent considering, a glass tube, like the syphon gage of an air- 

 pump with water in its bend was luted to an aperture con- 

 nected with the bottom of the chimney, the difference of the 

 heights oscillated between ± and f of an inch. We may therefore 

 infer that bellows working with a load equivalent to -\- of a pound 



(o) From the current of flame ; I liave seen it so hot as to proJuc;; iostant combus 

 tion in fragments of wood introduced through apertures several feet above the fireplace 



