360 Mr. R. H. M. Bosanquet on the Illuminating-power 



65 respectively. The line, whose equation is given, is drawn 

 through the second mixture and a mean between the two last. 



Ethylene and Marsh-Gas Mixtures. 

 y=200-6(ff-\L840). 





x. 



3/obs. 



#calc. 



Error. 



Ethylene 



•5 



•4642 

 •4227 

 •3943 

 •3392 



•2825 



68-5 



57'91 



47-88 



40-42 



33-17 



19-35 



63-39 

 56-20 

 47-88 

 42-18 

 3113 

 19-75 

 17-19 

 13-23 



-5-11 

 -1-71 





 +1-76 

 -204 

 + -40 

 - -40 



Mixture (1) 



(2) 



(3) 



(4) 



(5) 



(6) 



•2697 17-59 



Marsh-gas (theoretioal) . 



•25 





In mixture (4), having y a little over 30, there is a con- 

 siderable deviation, which must appear in any representation. 

 In the hydrogen mixtures there is a similar deviation for 

 nearly the same value of y. If these are errors they may be 

 due to some systematic source, and carry less weight if that 

 probability is taken into account. If they represent realities, 

 this part of the range will have to be covered by a larger 

 number of experiments before they can be dealt with. 



The meaning of the equation at the head of the table is — 



The C.P. is proportional to the carbon in the unit volume, 

 diminished by *1840. If, therefore, we suppose that the unit 

 volume of diluent absorbs this '1840 of carbon, while the rest 

 is all precipitated in the illuminating form, producing a C.P. 

 proportional to the amount precipitated, the law of the mix- 

 tures represented by the straight line is accounted for. 



In entering on the explanation of the curvature which leads 

 from the top of the straight line representing the mixtures to 

 the direct experimental value of ethylene, I must refer to the 

 scheme of " illuminating ratios " at the end of my last paper. 



In the light of the explanation just given, we see that the 

 physical meaning of what I called the illuminating ratio, but 

 shall now call the illuminating value, is the C.P. equivalent of 

 the carbon added in any change; also the lines considered were 

 straight, so that the absorption was all done before starting, 

 and the ratio was the C.P. equivalent of precipitated carbon. 

 The conception of the disilluminate hydrocarbon point in- 

 volved really an attempt to find the solvent capacity of hydro- 

 gen in hydrocarbons. 



Now the numbers in the table do not agree with those 

 deduced from the marsh-gas-ethylene mixtures. The solvent 

 capacity attributed to all hydrocarbons in the table is #='138. 



