Thermochemical Constants. 57 



the corresponding solid ammonium salt ? This same case is 

 also an instance of an entirely fictitious constant ; from the 

 fact that, in calculating the value of N, H 4 , CI, the quantity 

 H, CI, aq. is involved, and also that the heat of neutralization 

 of ammonia by the three haloid acids is identical, it follows, 

 as Thomson himself clearly shows, that each of the differences 

 N, H 4 , CI — H, CI, aq. is made up of x— A, where x is the same 

 quantity in each case, and A is the heat of dissolution of the 

 ammonium salt concerned. So that the constancy depends 

 entirely on the three values for A being nearly equal (which 

 we might well expect to be the case) ; these values are 3,880, 

 4,380, and 3,550 respectively, showing a variation of as much 

 as 800 cal., which amounts to 20 per cent, of the whole, but, 

 when imported into the large quantity of 36,000 cal., appears 

 as little more than 2 per cent. 



There are other points which must raise suspicion as to the 

 reality of these constants. Why, for instance, amongst the 

 many acids containing sulphur, should dithionic and tetra- 

 thionic acids alone show numerical relations? Why should 

 P 2 5 , As 2 5 , and As 2 3 form a group to the exclusion of 

 P 2 3 ? Then, again, there may be found other quantities 

 where there is as evident a numerical relationship as in many 

 of those quoted by Thomsen, but in which this relationship is 

 evidently accidental; thus, immediately opposite the page on 

 which No. 4 is given, we find 



iSOf s ,aq. = 7,700, 

 SO^aq. = 1,500, 

 BOJ*,aq. =39,170; 

 which numbers may be expressed as 



5 x 1,540, 



1 x 1,500, 



26 x 1,507; 



although the difference between the first and second quan- 

 tities is due to purely physical causes, and the other differences 

 to chemical, or combined physical and chemical causes. Again, 

 it may be observed that the numerical relations are in many 

 cases scarcely of a very simple character, such fractions as ^, 

 ■fa, and -j 8 ^ occurring ; and, moreover, that the error is often 

 of very considerable magnitude. Thus, in No. 12, P 2 , O s 

 should be 386,650 cal., if the constant supposed to be shown 

 in As 2 , 3 is correct ; whereas, if the constant given by P 2 , 5 

 be correct, then As 2 , 3 should be 147,960 cal., an error of 

 either 16,750 or 6,710 cal. occurring; certainly out of all 



