20 Mr. W. Sutherland on the Fundamental 



The first conclusion to be drawn from inspecting this table 

 is one unfavourable to the view of those who hold that each 

 atom in passing from the elementary to the combined state 

 evolves a fixed amount of heat irrespective of the atoms with 

 which it combines ; or, in other words, that the heats of com- 

 bination can be separated into thermochemical moduluses 

 or parameters characteristic of each element : for if this 

 were so, then the columns headed Br — CI and I — CI in this 

 table ought to be constant, whereas the large variations 

 in these columns establish the fundamental principle that 

 the amount of heat evolved on the combination of two 

 atoms depends on certain mutual relations of the atoms, a 

 point that BeketofF appears to have been most strenuous in 

 insisting on, and also Berthelot. Now in the case of com- 

 pounds formed in dilute solution, it has been proved that in 

 a large number of cases the heat of formation is independent 

 of mutual relations between the reacting atoms, and may be 

 expressed as the sum of moduluses characteristic of each; and 

 thus it will be seen that our first result is calculated to throw 

 new light on the nature of solution, for it shows that in the 

 process of solution certain mutual relations of the atoms of 

 the dissolved molecule cease to be operative. 



The second conclusion which can be drawn from this table 

 is that the ratio of CI— I to CI — Br is nearly constant, 

 because the large departures in the case of Ag and Hg do not 

 count for anything, seeing that with them CI — Br is so small 

 that its value is relatively greatly altered by relatively small 

 unavoidable errors in the values for CI and Br respectively. 

 Indeed with the existing uncertainties in our calculations of 

 the latent heats of vaporization of solids, it is obvious that 

 we can hardly ascertain whether the departures of the ratio 

 from constancy are real or due to errors, but under the 

 circumstances the approximation to constancy is close enough 

 to warrant our assuming it to be probably a law of nature. 

 This, then, we take to be our second result, that such ratios 

 as those tabulated under the heading (CI — I)/(C1— Br) are 

 constant, and it must be our first business to inquire what 

 may be the meaning of such a result. The simplest way of 

 expressing a dependence of the heat of combination on mutual 

 relations of the reacting atoms is to suppose that H (RS) is 

 of the form (R) +/(RS) + (S) where (R) depends on R 

 alone, (S) on S alone, and/(RS) on both R and S. Suppose 

 Si to stand for CI, S 2 for Br, and S 3 for I, then 



HtHSJ-HCB S.) (BQ-(8d+/(BB 1 )-/(RS,) 

 HCRSJ-HIHSJ - (8 l )-(8 8 )+/(BS 1 )-/(BS 1 ) ; 



