M. J. Thomsen's Thermo -chemical Researches. 415 



The difference of the two results is 32 thermal units, or 2 for 

 every thousand of the heat of neutralization; now the exactitude 

 of experiments cannot attain a closer approximation. In the 

 case in which hydrochloric acid replaces nitric acid, we have 



Thermal units. 

 15689 -13740=1949 



244- (-1682) = 1926, 



the agreement being as satisfactory as possible. 



We shall now show how to calculate the magnitude of the 

 decomposition for the reaction of an equivalent of nitric acid on 

 an equivalent of sulphate of soda. 



Let B be the soda, A the sulphuric acid, A f the nitric acid, 

 and x the proportion of decomposed sulphate of soda; we shall 

 have in the liquid after the reaction, 



(1— *)BA+*BA'+>A + (1— *)A'. 



We may regard the total reaction, and the thermic effect 

 which accompanies it, as the result of a series of simple reactions, 

 of which the thermic effects have been previously established ; 

 that is to say : — 



1. Decomposition of x equivalents of the salt BA, —x 15689 

 thermal units. 



2. Formation of # equivalents of the salt BA', +#13617. 



3. Action of (1 — x) equivalents of the sulphate of soda on x 

 equivalents of sulphuric acid, say 



(1 -a?) (NaO SO 3 , — - SO 3 ). 



4. Action of x equivalents of nitrate of soda on 1— x equiva- 

 lents of nitric acid : we have seen that the thermic effect produced 

 in this reaction may be neglected. 



5. Action of x equivalents of sulphuric acid on (1— x) equi- 

 valents of nitric acid : experiment proves that the thermic effect 

 produced by this reaction is not measurable. 



The sum of these effects must be equal to the total thermic 

 effect shown in the reaction ; hence 



-#2072 + (l-#)(NaO SO 3 , y£- SO 3 ) = -1752 ther. units. 



We may deduce the value of x from this empirically, by utilizing 

 the previously cited experiments which give the thermic effect 

 resulting from the action of sulphuric acid in various proportions 

 on sulphate of soda. We thus find that the preceding formula is 

 very exactly satisfied by the value # = §, because we have then, for 



