Tetravalency of Oxygen. 225 



property of the atom is most striking. We might usually 

 confine this notion of the thing or force to " atomicity," and 

 reserve the idea involved in the number or ratio to " valency." 

 I propose that we should always look upon valency as a 

 number and atomicity as a property, but both still to be 

 investigated. That the valency of all known atoms is a small 

 integer, and most probably 1, 2, 3, 4, 5, or 6, is surely one of 

 the best assured facts in the present state of chemical know- 

 ledge. It is a separate question whether in particular com- 

 pounds the valency of a selected atom may vary within the 

 limits of these numbers. It is also a separate question what 

 this number really means. But I would particularly point out 

 that this valency is apparently independent of the discussions 

 on and decimal revisions of our atomic weights. For valency 

 as a number must be a ratio and referred to monovalent H'; 

 and practically we have referred the valency of all atoms to 

 either monovalent H', or to that unit weight of chlorine which 

 combines with the unit weight of hydrogen in the H/Cl'i Un- 

 fortunately the word equivalent has had and has several 

 meanings attached to it. But it is most convenient to take it 

 as a weight, viz. that weight of the non-metal which combines 

 with unit weight of the gaseous metal hydrogen, or that 

 weight of a metal which combines with the chemical unit 

 weight of the gaseous non-metal chlorine. 



Our first most obvious and useful definition of valency is, 

 first, that it is the numerical ratio, 



atomic weight ' 

 equivalent weight ' 

 e. g. (1) valency of antimony (in antimony trichloride) is 



^°=3,i.,.Sb'", 



(2) valency of oxygen (in steam) is 



16 15-96 _ _.. 



T or W =2 > or0 ' 



As some elements have more than one equivalent weight, 

 but can only have one atomic weight*, it is not surprising that 

 the valency of an element may be a varying number. But 

 although this is really a practical rule in the case of almost all 

 the metals, we meet with difficulties in the non-metals. Thus, 

 on the hydrogen standard, this kind of valency of oxygen is 

 II. and I. (cf. mercury and chlorine, giving II. and I. alsof), 



* Dr. "W. Crookes has recently speculated that the atomic weights of 

 individual atoms may slightly vary, assuming that they contain the 

 hypothetical protyle. 



t Usually formulated Hg" and (Hg 2 )". 



