464 GENERAL OBSERVATIONS. 



nation. I have sometimes thought that the ano- 

 maly might be obviated by admitting, that oxy- 

 Atom of gen in reality has an atomic weight amounting 



oxygen may . , 



be = 0-5. to 0*5 instead or 1 ; on that supposition, it will 

 have the property of usually entering into com- 

 binations by atoms at a time. All those com- 

 pounds which I have considered as containing 

 only 1 atom of oxygen, will in reality contain 2 ; 

 and those which I suppose to contain li atom 

 will contain 3 atoms. This supposition, if ad- 

 mitted, will make no alteration in the atomic 

 weights given in this work ; it will only alter 

 our way of viewing them, and 1, instead of de- 

 noting the weight of 1 atom oxygen will repre- 

 sent 2 atoms. It will be more convenient to 

 retain these atomic weights even if this new view 

 of the subject should be ultimately adopted ; 

 because, it is but very seldom that a less quanti- 

 ty of oxygen enters into combination with other 

 bodies than what we have denoted by 1. One, 

 then, may still be considered as the representa- 

 tive of a compound atom of oxygen, because it 

 is the common multiple of it which enters into 

 combinations ; though in a few rare instances we 

 find that 0-5, or a simple atom of it, also unites 

 with other bodies. Such combinations consti- 

 tute the peroxides of iron, nickel, cobalt, ceri- 

 um, and sodium, and likewise the deutoxide of 

 manganese. 



If we embrace this view of the subject, which 



