1 28 Mr. Bosanquet on the Illuminating -Power' 



point, we get a line which represents all mixtures of these 

 hydrocarbons with the disilluminated gas. 

 Equation to line through, 



y = -69 #"=137 hydrocarbon mixture, 



and «'='132y=0 disilluminated gas 



137 

 is y=(«-132) . 6a _. 132 . 



In this put .r=198-; and we have for the C.P. of the 

 entire gas, y = 16*2. 



So that we have determined the illuminating-power within 

 2 tenths of a candle of the real value 16, from the data given. 



Of course there are many things in this process which re- 

 quire further attention. Chiefly the influence of the diluent 

 constituents, which are present in small quantity, and the 

 assumption that the heavy hydrocarbons are entirely composed 

 of ethylene and benzene, which cannot be entirely true, though 

 it may be so very nearly, according to the analysis. 



We proceed to consider the question of the carbon density 

 of pure substances. It appears at once evident that the 

 representative points surround the disilluminate region in 

 such a way that the angles with the x axis made by lines 

 radiating from that region to the points are larger for sub- 

 stances of greater carbon density, and smaller for those of 

 less. The first question is, what point is to be assumed as 

 centre ? We have the data for determining this only in the 

 case of carbon density =2. Drawing a line through ethy- 

 lene and ethane, and taking ethylene C.P. = 65 we have 

 ^ = '138 at the intersection with the x axis. In the absence 

 of further information we assume that this point represents 

 what may be called the disilluminate hydrocarbon centre. 



It appears probable that the illuminating ratio is formed 

 by the addition of differences which correspond to successive 

 units of carbon density. 



The first unit is a difficulty, as we are uncertain as to the 

 C.P. of marsh-gas. Perhaps it is best to leave this out of 

 consideration at first, and arrange the number in accordance 

 with the rest of the series, the law of the whole being as far 

 as possible continuous. The values of the illuminating ratios 

 formed by lines drawn to '138, 0, are as follows : — 



