76 



to compress a portion of air and permit it to escape through a small 

 aperture, and this is extensively practised by the nianufecturer, and 

 on particular occasions by the Chemist, as a much higher heat can be 

 procured by this means than in the air ftunace, but it is not neces- 

 sary to examine it in detail. A more refined method consists in a 

 permitting the gases which have passed through the furnace to es- 

 cape through a vertical tube of imperfectly conducting materi- 

 als : these being hot, and therefore dilated are of less specific gra- 

 vity than the external air, so that the column of them which 

 fills the tube is not in equilibrio with the external pressure, and 

 the difference of weight between it and a similar colunni of air is 

 the force of draught, the air will rush into the fireplace with 

 the velocity due to it, and being ex])anded there Mill in its turn 

 fill the chimney and continue the process. This ingenious con- 

 trivance was probably discovered by accident, it was perhaps in- 

 vented to remove the smoke, and some attentive observer perceiv- 

 ed that it augmented the combustion ; its theory is simple although 

 as I have already observed, we are not possessed of data suffi- 

 cient to determine it completely. If tlie temperature were uni- 

 form throughout the tube and the specific gravity of the gases propor- 

 tional to it, it would be easy to determine the effect, but as this 

 is not the case, we must ascertain, first the law of the diminu- 

 tion of temperature, 2dly the weight of the column of gases, and 

 lastly determine from them the draught. It is certain that if the tube 

 be of uniformly conducting materials, NcAvton's law will be ob- 

 served, according to which the quantities of heat lost are propor- 

 tional to those which remain,(A) In this case let h be the distance 



(J) This law is beautifully illustrated by some experiments of Biot related in the Journal 

 des Mines ; he connected one end of an iron rod with a constant source of heat, and ap- 



