Atomic Laws of Thermocliemistry . 47 



8. Cyanides. — The heats of formation of hydrocyanic acid, 

 acetonitrile, and propionitrile from solid amorphous carbon 

 are —27*6, —16*3, and —10*8, so that the values of p + ad 

 are 10*9, 60*5, and 104*3, the differences between which are 

 49' 6 and 43*8, the latter is practically equal to the 44*2 due 

 to CH 2 in homologous series, but the former 49' 6 is larger by 

 5*4 kcal., on the strength of which Thomsen proposes for 

 HON an entirely different constitution from that of the 

 cyanides ; thus he treats acetonitrile as H 3 C*C \ N and pro- 

 pionitrile as H 5 C 3 *C ; N, but hydrocyanic acid as H'C'N, 

 and then he assigns the thermal values of the junctions \ N 

 and C'N as 7*6 and 2*4. For cyanogen C 2 N 2 the heat of 

 formation is —65*7, and p + ad=ll'l, on account of which 

 Thomsen assigns to it the structure C : N * N : 0, and obtains 

 for the thermal value of the junction C : N 5*5 ; and he points 

 out that the values of the three junctions C * N, C : N, C ■ N 

 are as 1, 2, and 3. But it is obvious that the structures for 

 HCN and C 2 N 2 are assigned in a manner relating only to 

 thermochemical, and not to general chemical considerations, 

 whereas we should be guided in the converse manner and 

 establish the foundations of the thermochemical theory in 

 accordance with established chemical law. 



Let us then try to express in our notation the pure facts 

 relating to cyanogen and the cyanides : if there is need for 

 distinction between pentad and triad N let its thermal effect 

 be a. ; for the values of p + ad for acetonitrile and propioni- 

 trile we have the expressions 



3/(CH)+ /(C-C)+/(C;N)+a=60-5, . . (45) 



5/(CH) + 2/(C-C)+/(0;N) + «= 104-3, . . (46) 



in which using the values /(CH) = 15, /(OC) = 14*2, we get 

 /(C -N) -fa=l*3 in acetonitrile, and # 9 in propionitrile, or 

 1*1 in the mean. Now if we proceed with the value for HON 

 in the same way we get /(CH) +/(Cj N) + a = 10'9, whence 

 /(C ;N)+a= — 4*1. The reason for the discrepancy in these 

 two values of /(C \ N) + cc is not hard to find, for in HCN we 

 have taken the thermal value of the junction CH to be the 

 same as it is in CH 4 , although the association of the N with 

 the C has conferred on the H the character of the H in acids ; 

 thus, then, we have to do with another case of a mutual 

 influence of atoms that are not directly connected. In most 

 cases we have been able to neglect these mutual influences 

 in our first approximations, but the amount of it here, 

 namely 5*2 kcal., shows the importance of proceeding to 

 second approximations as soon as the data permit of making 

 the attempt. It can be seen that there is no need to accept 



