Atomic Laws of Thermochemistry . 45 



The heat of formation of pure N 2 4 at 18° C. is given by 

 Thomsen as —3*8, so that the heat of formation of N0 2 is 

 — 8*4. It will be noticed that Thomsen has made quite an 

 arbitrary assumption as to the constitution of N0 2 , which 





 could be = N=0 rather more reasonably than N\ | ; and 





 in his further treatment of the oxides of nitrogen, assump- 

 tions become more abundant than facts. Thus he assumes 

 that, because in the case of carbon v 1 —-v 2 , or/(C*C)=/(C:C), 

 therefore it is likely that the single binding of the two 

 atoms in his assumed constitution of N0 2 is the same in 

 thermal value as the binding of two atoms to form the 

 molecule ; and accordingly the heat of formation of 2N0 2 , 

 namely— 16*8, is four times the thermal value of a single 

 binding between N and 0, minus the heat required to break 

 the N molecule into two N atoms, namely 13. Thus the 

 thermal value of a single binding between N and is 

 obtained as — *95 ; or, as this is small, it is taken as zero by 

 Thomsen. Turning now to nitrogen monoxide NO, the heat 

 of formation of which is — 21*6, Thomsen assumes that the 

 double binding between and N which its structure involves 

 has double the thermal value of the single binding and is 

 therefore zero, and he then deduces for the thermal value of 

 a single binding between two oxygen atoms the value 26*3. 



It is evident that a satisfactory investigation of the oxides 

 of nitrogen is not easy, and the accumulated assumptions in 

 Thomsen' s attempt being untested in any cases beyond those 

 which they were first devised to meet remain pure assump- 

 tions. He does not apply them to the interesting cases of 

 N 2 and N 2 5 , where they fail unless supported by still 

 further assumptions. For on Thoinsen's principles the struc- 





 ture of N 2 ought to be taken either as N — — N or iy^V 5 



according to the first of which the heat of formation would 

 be 2x0 — 13 — ^ x 26'3 or — 26 • and according to the second, 

 2x0 — ^x26*3, or —13, while the experimental number is 

 — 18. Again, according to Thomsen, the structure of N 2 5 



/) 



would probably be I N — — N | , the heat of formation of 



/ \) 



which, according to his values, ought to be 6 x — \ x 26*3 — 13, 

 or —26, whereas the experimental number is zero. The 



structure q~N— 0— NZq would give the value 

 10x0-13-f x26-3, 



