226 Mr. J. F. Heyes on the 



but the oxygen chlorides give 2, §, and \, which ratios are 

 those of 12, 4, and 3. It would be illogical to infer the multi- 

 valency of oxygen from these figures, but they seem to 

 support clearly the integral character of valency. On this 

 account I would venture to deprecate the use, except with 

 extreme caution, of such phrases as " residual affinities " or 

 "residual valencies/' at any rate so long as we fully and 

 strictly accept the fundamental laws of definite, constant, and 

 multiple proportions. Whatever may emerge as the precise 

 meaning of that well-known little . in those formulas where 

 the joint influence of reactions and of notions will neither 

 consent to the distant chemical acquaintance of + , nor the 

 intimate familiarity of the valence-link — (so prematurely 

 thrust upon us in inorganic text-books), it would nevertheless 

 seem clear that we are observing t^e action of integral units 

 of some sort, and not residual or small decimal quantities of 

 chemism, affinity, or attraction — whichever our mysterious 

 force be called. 



It is in such apparently simple cases as the oxides of 

 chlorine and nitrogen that the combined doctrines of valency 

 and linking seem to break down, or are at least not obvious. 

 Here is the great paradox in our constitutional formulae, even 

 from the low standpoint of reactions' shorthand. The doc- 

 trines have been applied with marvellous success to analysis, 

 synthesis, and prevision of thousands of organic compounds, 

 including dozens of liquids which, although they give the 

 same analytical and vapour-density results, are absolutely 

 distinct in their properties and reactions. And, basing then 

 work on the idea of the geometrical stability of a molecular 

 system containing strictly tetravalent atoms, Van 't Hoff 

 and others have gone still further. Yet in the apparently 

 simple molecules of inorganic chemistry, we are still in 

 a state of comparative chaos. At least we may say that, 

 whereas in organic chemistry the young student finds the 

 constitutional formulas the most helpful shorthand-memo- 

 randa possible of reactions and relationships, the beginner in 

 inorganic chemistry is usually disastrously affected, if not 

 hopelessly muddled, by the presumed necessity of writing 

 these so-called graphic (?) formulas, in the case, for instance, 

 of the oxides of nitrogen, chlorine, and manganese, or even 

 for what are commonly called ammonia, hydrochloric acid 

 and nitric acid. 



In the case of hydrogen dioxide, H 2 2 , the only reasons 

 assigned, so far as I know, for the usual formula (OH) 2 or 

 0— H 

 | or H— 0— 0— H, are (1) that it is f < free hydroxyl," 



