2Iixter — Thermochemistry and the Periodic Law. 523 



atomic weight from one element to the next is small, that 

 any variation from linear in the diagram signifies a large 

 difference from what the heat effect would be if it were 

 a linear function. For example, if silicon were on the line 

 connecting aluminum and phosphorus its heat of oxidation 

 would be nearly as high as that of aluminum. 



While it is not yet possible to account for the irregularities 

 shown by the diagram, this may be stated. The rise or 

 fall in the quantities of heat evolved by 16 grams of oxy- 

 gen and 71 grams of chlorine is fairly regular in a series. 



Thermochemistry of the Groups. 



The heat equivalents of the members of a sub-group vary 

 more regularly than they do in a series in the Periodic System. 

 Some marked peculiarities in the thermochemistry of small 

 groups have been observed. For example, Thomsen* states 

 that in the groups, sulphur, selenium, tellurium, and phos- 

 phorus, arsenic, antimony, the heat of oxidation of the middle 

 member exceeds that of the last. Also — " The affinity of 

 water is greatest for baryta and least for lime " ;f and in each 

 of the groups of non-metals the affinity of hydrogen is greatest 

 for the first member and falls with an increase of atomic 

 weight.":}: Thomsen§ discusses at length the " Dependence of 

 the Thermal Effect upon the Molecular Weights of the React- 

 ing Substances," and states, " There is at present no satisfac- 

 tory answer to the question as to the relation between the 

 thermal effect and atomic weight, although some such connec- 

 tion may be observed in the material under consideration." 

 He apparently did not notice that in a group of three closely 

 related elements the one with an atomic weight nearly the 

 mean of the other two has a heat of combination nearly the 

 mean of the others. For example, hydrogen bromide has a 

 heat of formation the mean of that of hydrogen chloride and 

 hydrogen iodide. 



The data used in the discussion are, unless otherwise stated, 

 from the Physikalisch-Chemische Tabellen, edition of 1912. 

 The investigators quoted are indicated by initials or otherwise. 

 They are Thomsen, Berthelot, Beketoff, Rengade, deForcrand 

 and some others. The groups are taken up in order. 



The most closely related elements in the first group are 

 potassium, rubidium and caesium. The fundamental values 

 in the thermochemistry of these metals are derived from the 

 quantities of heat evolved when they are dissolved in water. 

 As the results of different investigators vary, the values in 

 Table II are somewhat uncertain. 



* Thermochemistry, 217. flbid., 261. 



% Ibid., 337. § Ibid., 360. 



