324 



Mr. W. Herapath's Contributions 



Table of' Protoxides. 



Here we see that the numbers in the 3rd column have an 

 evident tendency to increase downwards : — that platinum, a 

 metal 21^ times as heavy as water, combines with about 4 per 

 cent, of oxygen ; while arsenic, about 5\ times as heavy as 

 water, combines with 31 percent. Tungsten, nickel, cobalt, 

 manganese, and iron, are greater numbers than they should be; 

 and it is strange that all of them, except manganese, are such 

 as we should expect not to have found the protoxides of, be- 

 cause the oxides in the table contain oxygen, compared with 

 the peroxides, as 2 to 3, leaving it probable that there is an 

 undiscovered oxide of each, containing oxygen as 1. In the 

 oxides of manganese there is great confusion. Berzelius finds 

 two with less oxygen than the one in the table : but the theory 

 of equivalents and the experiments of good chemists plead 

 against the first of them; while the second of them I have not 

 introduced, because it is not well defined, although it would 

 better accord with the table. 



My experiments on the oxide of tin from muriatic acid, 

 make it composed of 100 metal +24-8 oxygen (instead of 

 13-56), in which case that metal would accord: there only re- 

 mains antimony, the oxides of which are still more confused 

 than those of manganese. 



I am the more induced to think that this law will be proved, 

 from the circumstance that the unmetallic bodies I find, as far 

 as they have been investigated, seem to be governed by the 

 same laws. 



I make this 100+24-92. 



Table 



