Atomic Laivs of Thermochemistry. 49 



approximations. In a first approximation we will neglect the 

 difference, anc^ determine a mean thermal value for the sub- 

 stitution. If in the series of differences given above for the 

 ethylamines we subtract 44*2 as due to CH 2 , then we get for 

 the formation and substitution of CH 3 for H in NH 3 from 

 the ethylamines the values 37-0, 43*2, and 44*1, which with 

 the three values from the methylamines give a mean value of 

 40. With the usually accepted constitution for the amines 

 and our notation, we have 



/(C-N) + 3/(CH)-/(NH)=40, .... (49) 



.-. /(ON) -/(NH) = - 5; 



but «+/(NH 8 ) = 11-3, 



.-.. /(NH)=3-8-g , (50) 



.-. /(0-N)=-l-2 + ? (51) 



This value for /(ON) found from the amines should be com- 

 pared with those found for /(CjN)+a, namely 1*1 in the 

 nitriles, —4*1 in HCN, and —1*5 in C 3 N 2 . It seems as though 

 /(ON),/(C; N), and a are so small as to be each practically 

 zero. Now the fact that there is little or no thermal differ- 

 ence between pentad and triad N need not surprise us, for in 

 triad N we can imagine 2 of the 5 bonds to satisfy one 

 another ; in the case of C this appeared to be accompanied 

 with a thermal effect /(C*C), so that in that of N the effect 

 would be /(N'N) which we have already inferred to be zero. 

 The case of triad N is very different from that of unsaturated 

 N in N0 2 and NO, where an odd number of bonds are left 

 unsatisfied in any way. 



In the region of first approximations this is about all that 

 we can get from the data for the amines, and there is no 

 encouragement at present to go into the minor points in 

 Thomsen's data until we know what are the limits of experi- 

 mental error in the heats of formation of the amines, for 

 there is no certainty whether a minor phenomenon is a 

 genuine one in thermochemistry or a relatively small error 

 in a heat of combustion which has become a relatively large 

 one in a heat of formation. Nor would there be any profit in 

 discussing the case of aniline in view of the uncertainty about 

 Thomsen's numbers for benzene, and the thermochemistry of 

 pyridine can hardly be discussed apart from that of the 

 benzene nucleus. 



Phil. Mag. S. 5. Vol. 40. No. 242. July 1895. E 



