1818.] the Atoms of Chemical Bodies. 345 



Weight of an atom. 



38 Peroxide of nickel 9-750 « 



39 Cobalt 3-625 



40 Protoxide of cobalt 4-625 w 



41 Peroxide of cobalt 10-250 



42 Manganese 3-500 



43 Protoxide of manganese 4*500 * 



44 Peroxide of manganese 5-500 



45 Uranium 15-625 



46 Protoxide of uranium 16-625 * 



founded on the best data to be had, and cannot be very far from 

 the truth. Tupputi found sulphate of nickel composed of acid 

 100 + 87-26 protoxide of nickel. Now 100 : 87-26 :: 5 : 4-362 ; 

 so that the weight of an atom of protoxide of nickel is 4-362 : 

 4*375 scarcely differs from this number. I prefer it, because it 

 is a multiple of 0*125. The average of the analyses of the prot- 

 oxide of nickel (omitting Proust's) makes it a compound of 100 

 metal -f 28-74 oxygen. This shows us that it must be a com- 

 pound of 1 atom metal + 1 atom oxygen. If so, the true pro- 

 portions must be 100 metal + 29-63 oxygen. 



11 This supposes it a compound of 2 atoms metal + 3 atoms 

 oxygen. It contains ll times as much oxygen as the protoxide. 



w This number is derived from the analysis of protoxide of 

 cobalt by RothorF, who found it a compound of 100 metal + 

 27*3 oxygen. If it be a compound of 1 atom metal + 1 atom 

 oxygen, an atom of the metal must weigh 3-663 ; but the atoms 

 of carbon, oxygen, azote, phosphorus, and sulphur, which have 

 been ascertained with the greatest precision, are multiples of 

 0*125, or an atom of hydrogen. Hence I suspect that all the 

 atoms, if accurately ascertained, would be multiples of hydrogen: 

 3*663 not being a multiple of * 125, I take 3*625, which is the 

 nearest multiple as the weight of an atom of cobalt : of course 

 the protoxide must be 4*625. Rothoff has shown that the oxygen 

 in the peroxide is 1^ times. as much as that in the protoxide. 

 Supposing the peroxide composed of 2 atoms metal -f 3 atoms 

 oxygen, its weight will be as in the table. 



x From John's analysis of sulphate of manganese, the weight 

 of protoxide comes out 4*6 ; from his analysis of the carbonate, 

 it comes out 4*495. I conceive, therefore, that 4*5 is probably 

 the true weight. The peroxide of manganese has never been 

 accurately analyzed ; but there can be little doubt that it contains 

 just twice as much oxygen as the protoxide. I believe the red 

 oxide of manganese, which is intermediate between the other (wo, 

 not to be a particular oxide, but a compound of protoxide and 

 peroxide. It does not seem capable of uniting with other bodies. 

 r The experiments of Schoiibert lead to the conclusion that 



