Tetravalency of Oxygen. 229 



two, which valence-powers probably all conform to the re- 

 markable law of odd and even numbers, or, as Prof. Odling 

 ■termed atoms with such valencies, perissads and artiads. If, 

 therefore, I ask that oxygen, chlorine, iodine, and fluorine 

 should be placed in this second class, it is not to complicate 

 but to simplify the great valency problem. Monovalent 

 atoms in such gaseous molecules as WC1 6 * and PF 5 seem 

 highly probable, but they may be questioned in the oxides of 

 chlorine and in the allied oxy-acids. Thus, in the simplest 



ci N ci x Cl^ 



oxide we really have the choice of .0, || /O, and | 0. 



J cf ci 7 Cl/ 



Cl x K H x 



Many chemists have accepted Jd and ,0 and ,0 



oK cr or 



somewhat hastily, because of their historic place in the 

 water-type theory, rather than from positive evidencef. 



If the KO CI formula be adopted there is the alternative 

 between K — — CI and K — 0=C1. But the chemical evi- 

 dence seems rather in favour of (K . CI . 0) x , in which case 

 chlorine is trivalent or tervalid as K — C1=0 ; or in K 2 C1 2 2 



K— C1=0 

 as {} . The remarkable increasing chemical stability 



K— C1=0 

 of the series KC10, KC10 2 , KC10 3 , KC10 4 remains to be 

 explained ; but I believe that the hypothesis of the valency 

 variation of one or both the negative elements must be 

 adopted, so long as valency language is useful in focusing 

 mysteries enshrined within a chemical molecule. Possibly the 

 oxygen atoms are in some sort of " closed ring/' Certainly 

 the old tandem arrangement K — — — — — CI must be 

 abandoned. In KC10 3 and KC10 4 , the K and CI may each 

 be separately associated with an oxygen atom ; or, seeing that 

 KG is such a constant product whether by wet or dry reac- 

 tions, there is something to be said for Cl /// in some such 



* Just above the boiling-point. 



t Undoubtedly the formula Cl(OH), that is, CI'— 0"— H, and notably 

 the CN(OH) formula for cyanic acid, established themselves under 

 shelter of a somewhat dangerous, or at least premature, extension of the 

 hydrate theory, or (OH), so well established in the case of organic acids. 

 The tendency to introduce unwarrantable OH-groupings is especially 

 shown in such cases as Ag(OH) and even NH 4 (OH). The distinctions 

 between hydrates and hydroxides should also be more carefully observed ; 

 the former are really quite in a minority. The reaction against the 

 indiscriminate use of OH-groupings in inorganic chemistry in many 

 " not proven " cases is now beginning to show itself clearly. 



Phil. Mag. S. 5. Vol. 25. No. 154. March 1888. R 



