Flames of Carbonic Oxide and Hydrogen. 499 



and the hydrogen and oxygen mixture at 2024° C, 



Oxygen . . ; 1 vol. 

 Hydrogen . . 2 vols. 

 Water-vapour . 2 vols. 



The graphic representation A (fig. 5) derived from columns 

 III. and IV. (in which the abscissas represent the temperatures, 

 and the ordinates the volumes of combustible gas obtained at 

 these temperatures from 10 vols, of hydrogen or carbonic oxide) 

 shows that the portion of the gas rendered incombustible by 

 the gradually increasing temperature does not vary proportionally 

 with the increase of temperature, but is represented by a broken 

 line. The curve B (fig. 5) represents in the abscissae-lines the 

 temperature of the flame of the carbonic oxide detonating mix- 

 ture, calculated upon the volumes of excess of gas added to 

 1 vol. of the mixture corresponding to the vertical ordinates. 



Where two experiments were made with the same mixture of 

 gases, the mean value is given. In curve B a turning-point is 

 shown corresponding to the broken part of curve A ; and this 

 represents the temperature at which the portion of the gas 

 undergoing combustion changes from one-third to one-half. 

 When the result of these experiments is considered, it is clear 

 that the phenomena of the combustion of gases depend upon 

 circumstances totally different from those which were formerly 

 supposed to regulate these actions. 



When carbonic oxide and oxygen are exploded, and the tem- 

 perature of the mixture in consequence raised from 0° to 3033° C, 

 two-thirds of the carbonic oxide present remains in an uncon- 

 sumed and incombustible condition, and the temperature 3033°C. 

 is lowered by radiation and conduction to 2558° C. without any 

 of the two-thirds carbonic oxide being burnt. If, however, the 

 temperature sinks below this, a second burning begins, which 

 prevents a further loss of heat by radiation and conduction, and 

 restores the temperature to 2558° C, beyond which point 

 it does not rise. There follows then a constant abstraction of 

 heat from the temperature of 3033° C, which produces a steady 

 temperature of 2558° C, until exactly the half of the carbonic 

 oxide is burned ; a third change now takes place by the cooling 

 of the inflamed mixture to 1146° C; and after this no combustion 

 occurs. Since the gaseous mixture on cooling consists entirely of 

 carbonic acid, these changes must be repeated at temperatures 

 under 1146° C. until the last portion of the gas is burned. 

 If carbonic acid is decomposed by slowly increased temperature 

 into carbonic oxide and oxygen, in the same way as it is formed 

 from a mixture of carbonic oxide and oxygen when losing tempe- 

 rature, a series of specific-gravity-determinations must give the 



