Tetravalency of Oxygen. 227 



/O 



(2) that it is obtained from Ba | , and (3), in the words 



x O 

 of Wurtz (I. c.) : — " If 0=0 represents a saturated couple, 

 the symbol — — — will represent a couple which is 

 unsaturated and capable of attaching, for example, two atoms 

 of hydrogen. This conception explains the constitution of 

 hydrogen peroxide, H — — — H." 



Now this is one of the most numerous class of bodies whose 

 true molecular weight we have at present no means of ascer- 

 taining, or at least of comparing with gaseous molecules. It 

 is, like Ba0 2 , Pb0 2 , Mn0 2 , Sn0 2 , Si0 3 , versus C0 2 , simply a 

 chemical unit formula. As such formulae are not strictly 

 comparable like our two-volume formulas, the chief, if not 

 the only, use of constitutional formulas in such cases should 

 surely be to explain or indicate reactions, and thus to con- 

 tribute something from the purely chemical standpoint to the 

 solution of the problems of valency. It can scarcely be said 



OH /O 



that | or Ba. | are, on these grounds, useful formulae. 



There is no chemical equality, so to speak, in the oxygen atonis, 

 and no suggested explanation of why some peroxides have what 

 is technically termed " available oxygen " and others have not. 

 One chief stumbling-block in our theories of quantivalence 

 seems to be due to an apparently unanimous faith in the 

 constant divalency of oxygen, as well as the constant mono- 

 valency of the halogens. This has been curiously illustrated 

 in the adoption by many chemists of extreme views in two 

 directions. Either the valency has been very largely varied 

 between 1 and 7 in the case of the halogens, or the valency- 

 power or atomicity of the multivalent metals has been largely 

 increased. The late Professor Wurtz advocated the first 

 view, and Professor Williamson's address at York furnishes 

 an instance of the second standpoint. Certainly, rather than 

 recognize so many valency powers* for the non-metals, it 



* Thus Wurtz gave to chlorine the valencies 1, 3, 4, 5, and 7, and to 

 nitrogen 1, 2, 3, 5. Heptavalent chlorine is also suggested by several 

 philosophers, especially in connexion with the periodic arrangements of 

 the elements. In such cases, at least equal weight must be given to 

 aluminium and gallium as trivalent in the periodic series, whereas the 

 tetravalency is constantly asserted, based on the two-volume molecular 



formula, M^Clg. I have suggested the alternative, A1-C1=C1-A1, in 



x ci=or 



these and similar cases as more probable than the usually assumed 



metallic nucleus, | or ||| . 

 M= M= 



