Solar Light on Combustion. 191 



effect must be equivalent to a diminution of bai'ometric pressure, 

 and consequently must tend to retard the process of combustion. 

 Assuming the temperature of the flame to be constant, it is like- 

 wise plain that the draught created by it — depending, as it is 

 known, on the difference of temperature between the flame and 

 that of the surrounding air — must be diminished in a warm 

 atmosphere, and therefore also tend to retard the rate of com- 

 bustion, cateris paribus, during hot seasons. But inasmuch as 

 the variations in the velocity of the draught ai'e proportional to 

 the square roots of these differences of temperature, it is obvious 

 that its effects must be insignificant under ordinary fluctuations 

 of atmospheric temperature. For example, supposing the tem- 

 perature of the flame to be 1500° F., then the fluctuation of the 

 draught between the temperatures of 80° and 60° F. would be 

 in the ratio of 



-v/lSOO-SO: v/1 500-60= 'v/1420 : '/1440=1 : 1-0070. 



When, however, the comparison is made between bodies burning 

 in summer and in winter, the influence from this cause will be 

 more sensible, and ought not to be entirely ovei'looked*. 



On the contrary, it is possible that an augmentation of tem- 

 perature might tend to accelerate the process of combustion by 

 favouring the liquefaction of the wax, and perliaps facilitating 

 the oxidation of the combustible matter. If, however, any such 

 influence is exercised, it is probable that its effect must be inap- 

 preciable under ordinary circumstances. Under this view of the 

 subject, the only obvious influence which atmospheric heat exer- 

 cises on the rapidity of combustion is connected with its effects 

 on the density of the air ; and consequently an increase of tem- 

 perature should, ccsteris paribus, retard combustion, and vice 

 versa. 



3. Amount of Aqueous Vapour present. 



Sir Humphry Davy found that " a very large quantity " of 

 steam was required to prevent sulphur from burning; that an 

 explosive mixture of oxygen and hydrogen, when mixed with five 

 times its volume of steam, still exploded by the electric spark ; and 

 that a mixture of air and carburetted hydrogen gas required " a 

 third oi steam to prevent its explosion, whereas one-fifth of azote 

 produced the eftectf." Under any point of view, it is obvious 

 that the presence of aqueous vapour can only tend to retard the 



* I endeavoured to test the influence of temperature on the rate of com- 

 bustion by placing the burning caudle over a large heated plate; but, as 

 might have been expected, the unsteadiness of the flame rendered the ex- 

 j)erimcnt unsatisfactorv. 



t Phil. Tiiins. for 1S17, p. G5. 



