NEW LAW IN THERMOCHEMISTRV 



25 



These quantities do not seem to bear any simple relation to 

 the formulae developed in this paper, but it is noticeable that two 

 of them approximate very closely to exact multiples of our con- 

 stant. This agreement is probably not accidental, but its sig- 

 nificance is not clear. I shall therefore, put these substances 

 in the class of exceptions, and leave them, with other outstand- 

 ing difficulties, uninterpreted. 



So far, in our comparisons of fact with theory, we have as- 

 signed to the henotherm a provisional value of 13700 calories. 

 Let us now sum up the values actually obtained, and compute 

 their general mean. 



Aliphatic hydrocarbons 14 compounds. Mean 13873 



Aliphatic halides 19 " " 13815 



Bromination series' 9 " " 13765 



Aliphatic amines 10 " " I377I 



Cyanogen compounds 4 " " 13686 



Sulphur compounds 8 " " 13773 



Ethers 9 " " 13730 



Alcohols II " " 13686 



Aldehydes, ketones and acids 9 " " 13777 



Esters 11 " " 13946 



Nitrates, nitrites, etc 6 " " 13474 



Aromatic compounds ii " " 13795 



General mean for 121 compounds 13777 



The data for 112 of these substances are due to Thomsen, 

 and only seven of the organic compounds studied by him are 

 not covered by the formula. They are trimethylene, carbonyl 

 chloride, carbon tetrachloride, ammonia, carbonyl sulphide, 

 carbon disulphide and carbon monoxide. The evidence estab- 

 lishing the constant seems to be overwhelming.^ 



Before passing on to the discussion of other data, and of the 

 law which I have sought to define, it may be well to tabulate 

 Thomsen's measurements, side by side with the calculated values. 

 The latter are simply the average value of the henotherm multi- 

 plied by the quantities heretofore used as divisors ; but reduced 

 from the quadrupled to the single equations. From this com- 

 parison we can compute the percentage variation between 



1 Data by Louguinine and Kablukoff. 



2 Glimpses of the same constant are given by Thomsen at several places in his 

 great work. See vol. 3, pp. 222, 293; and vol. 4, p. 316. The existence of the 

 constant is noted, but its significance is not shown. 



