of Hydrocarbons and their Mixtures. 125 



point from the actual behaviour of the mixtures than direct 

 from the analysis. The disilluminate point selected is 



#=-1321 5 y=.0. 



This is the mean of the numbers given by equations of lines 

 drawn from the ethylene point at 65 to the mixture points *. 

 Two mixtures were also made with disilluminated gas and 

 benzene vapour. 



Pis. gas. Benzene. x y 



96*91 3-09 -21210 22'92 



96-70 3-30 -21222 24-17 



Mean -21216 23*54 



The two points are too close together to draw a line through 

 with any useful result. I have calculated the position of such 

 a line, and it is nearly vertical, which does not correspond at 

 all with other indications. (The illuminating-power of ben- 

 zene would come out about 1000.) So I have taken the 

 mean of the numbers, and denoted the point by an upright 

 cross of dotted lines. 



Through this benzene cross a line is drawn from the dis- 

 illuminated-gas point, # = *1321, If the various suppositions 

 that are made are at all true, this line at its intersection with 

 x=l should give the C.P. of benzene. This number calcu- 

 lated from the above data is 255. 



It remains only to consider the complete gas of the Gas Light 

 and Coke Co. The percentage and numerical formula of the 

 heavy hydrocarbons are given, so that the value of C/H for 

 the whole gas can be obtained. Hence I find 



x y 



Gas Light and Coke Gas -198 16 



This point is laid down as a dotted oblique cross. A line 

 drawn through this from the disilluminated-gas point (% = 

 •1321) lies fairly between the lines of the benzene mixture 

 and the ethylene mixture. The influence of the nature of the 

 mixture of heavy hydrocarbons thus indicated will be dealt 

 with later. 



Let us now turn to the meaning of the lines on the figure. 

 The slope of a line represents a ratio Ay/A#, the ratio of 

 difference of candle-power to difference of C/H ratio. We 

 may regard it as the illuminating value of unit C/H difference, 

 and shall speak of it as the illuminating ratio. 



* I have not succeeded in representing this with great correctness on 

 the small figure. That conveys a general idea, but the numbers have in 

 all cases been worked out by the equations. 



