Different Kinds of Coal fur the purpose of Illumination. 47 



I have observed similar results in trials which I have 

 lately made. Thus three coals, submitted to experiment, 

 yielded gases, the indication of which, by chlorine, was 14 ; 

 the durability in the one being 57', in the other two 66' ; and 

 again, with other gases, in which the condensible matter was 

 as high as 19 and 22, the durability did not exceed 77' and 

 81'- In numerous trials I found that the gas from English 

 caking coal gave condensation by chlorine 4-33, and dura- 

 bility 50' 30", or nearly so. That from Wigan cannel coal 

 had condensation as 7-5, but the durability was only 57' ; the 

 Yorkshire cannel-coal gas was, condensation 7*66, and dura- 

 bility only 52' 30". Had the durability of the English can- 

 nel-coal gas kept pace with the condensation test, it ought 

 to have been at least 87' instead of 52' and 57'. 



It is evident from this that the durability is affected by 

 other circumstances than the presence of the ingredient, 

 whatever it may be, which causes the condensation by chlo- 

 rine, in other words, than by the illuminating power. 



Considering this still farther, I observed a remarkable 

 coincidence between the durability and the specific gravity, 

 as is shewn in the following table : — 



In the above table it is shewn, that as the specific gravity 

 becomes greater, the times required for the consumpt of 

 equal quantities become longer ; but the increase of the one 

 does not keep pace with that of the other. There is, how- 

 ever, some connection between them, and on farther investi- 

 gation it occurred to me, that pei'haps the consumpt of 

 gases by combustion is regulated by the same law as the 

 diffusion of gases, as pointed out by Professor Graham, viz., 

 that under equal pressures the diffusion is inverseli/ as the 

 square roots of the specific gravities. Accordingly, in equal 

 times, the consumpt should be inversely as the scjuare 

 roots of the specific gravities ; and, convei*sely, the times for 

 the consumpt of equal volumes, from similar burners, and 

 under the same circumstances, will be as the roots of the 

 jn-avitios. 



