22 Abstract of recent Researches on the quantity of 



Hess states, in his letter to Arago, that he had taken notes 

 immediately after a conversation with Dulong, in which the 

 latter said he had arrived at the following general laws : — 



1st. That the quantities of heat evolved are (nearly) the 

 same for the same substance, no matter at what temperature 

 it burns. 



2nd. Equal volumes of all gases, in combining with oxygen, 

 evolve the same quantity of heat. 



3rd. The same quantity of heat is disengaged with the same 

 quantity of oxygen, whether it forms a compound R + O, or 

 R + 2 O. 



4th. The quantity of heat disengaged by different substances 

 is very different. 



These conclusions, however, are by no means completely 

 borne out by the experimental results, so far as they have 

 been recovered, except the third, which is certainly supported 

 by the results with tin, protoxide of tin, carbon and carbonic 

 oxide. 



In commenting on these numbers, Berzelius has pointed out 

 a remarkable consequence deducible from them ; that is, that 

 the heat evolved in the combustion of a compound gas is the 

 same as that evolved in the combustion of its constituents. 

 Thus, 



One volume of pond gas consists of two volumes of hydro- 

 gen and half a volume of carbon (vapour). Mean. 



Two volumes of hydrogen produce heat = 6204 



Half a volume of carbon (vapour) = 3929 



10133 



The highest experimental number is . . . = 9948 



One volume of olefiant gas consists of two volumes of hy- 

 drogen and one of carbon (vapour). Mean. 



Two volumes of hydrogen evolve heat . . = 6204 



One volume of carbon (vapour) = 7858 



14062 



The mean experimental result is == 15338 



One volume of oil of turpentine vapour consists of eight 

 volumes of hydrogen and five of carbon (vapour). 



Mean. 



Eight volumes of hydrogen evolve heat . = 24816 



Five volumes of carbon (vapour) = 39290 



64106 



The experimental number is = 70607 



The differences here found are not probably greater than 

 the limits of experimental error; their direction also is not 

 constant. 



The laws of phenomena are very frequently more easily 



