370 Illuminating- power of Hydrocarbons. 



This is very fair, for an approximate method. 



The oxygen has been omitted. The quantity is small ; and 

 the introduction of a small quantity of oxygen into diluted 

 gas seems to produce a condition so different from that of the 

 oxygen or air dilutions of ethylene, that conclusions drawn 

 from the latter do not seem applicable to the former. 



Although mixtures of marsh-gas with higher illuminants 

 will probably be free from absorption, yet they may be 

 expected to show towards the top the curvature found in the 

 ethylene mixtures. It has therefore occurred to me that 

 ethane might probably be used with good results as the 

 foundation of mixtures, for obtaining rectilinear loci leading 

 to the C.P. values of the higher illuminants. At all events 

 the chance of getting a good line in this way seems better 

 than any other. I should suggest the employment of small 

 weighed quantities of benzene completely volatilized in the 

 substance employed. 



It appears that the analysis of the heavy hydrocarbons 

 which is given cannot be correct. The nearest representation 

 of it would be one part of benzene to three of ethylene. This 

 gives = 3, H = 4*5. The analysis has practically the same 

 value of C, but 11 = 4*32. And it can be proved that the sub- 

 stance required to mix with ethylene to satisfy the analysis 

 must have fewer hydrogen atoms than carbon atoms, which I 

 suppose cannot be*. 



The verification of the table of illuminating values seems 

 to be most important for theory. We ought to try to define 

 at least two substances of each carbon density, so as to obtain 

 direct determinations of the illuminating value in each case. 



For carbon density 2, we have three substances convenient — 

 ethane, ethylene, and acetylene (C 2 H 2 ). Mixtures of ethane 

 and acetylene would determine whether the three representa- 

 tive points are on a straight line. For instance, a mixture of 

 ethane and acetylene having the same composition as ethylene 

 should have the same illuminating-power. 



For carbon density 3, propane and propylene are wanted, 

 direct or by mixtures. Ethane mixtures look promising for 

 these purposes, and so on for the higher densities. 



It seems just possible that a lamp burning benzene with a 

 wick at a known rate by weight might be available for 

 determining the C.P. of benzene directly. 



P.S. — It seems possible that the falling away of values of 

 C.P. below 15 may be partly due to a lowering of temperature. 



* I find in the scheme of hydrocarbons in Encycl. Brit., art. " Chemistry," 

 propone C 3 H 2 , butune C 4 H 3 , but I do not suppose it is seriously meant 

 that such substances exist. 



