4 ART. 17.— EDWARD DIVERS. 



printer's error for 0:N:(S0 3 K) 2 is uncertain, but if it is not it indicates 

 some recognition by these chemists of the presence of univalent ' 

 oxygen. However this may be, the possibility of the nitrogen 

 being quadrivalent being inadmissible, the only solution of the matter 

 seems to be to consider that both mononitric peroxide and porphy- 

 rexide are compounds of univalent oxygen, although still peroxides. 

 The fact that a molecular quantity, such as that formulated by HO, 

 N0 2 , (S0 3 K),NO, or (C 5 H 9 N 3 )NO, is never met with singly in 

 chemical interactions militates against the acceptance of this expla- 

 nation. Piloty and Schwerin, in discussing the quadrivalency of 

 nitrogen, conceal this fact by stating that porphyrexide is produced 

 from porphyrexine by the action of half an atom of oxygen. It is 

 much more correct to hold with Haga that the molecule of 

 peroxylaminesulphonate, and therefore also of mononitric peroxide and 

 of porphyrexide, is not less than that represented by [N"0(S0 3 K)o]..„ 

 (NO,)'ö, or (CfHgN^O);,, as the case may be, if by molecule is meant 

 the smallest chemically active weight of a substance. But still the 

 fact remains that, when measured by comparison of their physical 

 properties, the morecular weights of porphyrexide and red nitric 

 peroxide are expressed by half the above formula', and probably the 

 bluish- violet from of a peroxylaminesulphonate has also a molecular 

 weight, which when physically considered, should be expressed by 

 the formula, NO(S0 3 K) 2 . These weights, only physically determined, 

 have no chemical significance, and should be distinguished from their 

 doubles, the truly chemical molecular weights. But, as pseudo- 

 chemical molecules, they must be represented to contain a univalent 

 atom of oxygen. 



It is, after all, not so difficult to admit that oxygen may be 

 univalent in a peroxide, [n fact, true peroxide may be defined and 

 differentiated from other oxides, as being a compound in which some or 



