122 T. 8. Hunt — Chemical Integration. 



very large numbers (which, as was subsequently shown, are 

 multiples by many thousands of that of hydrogen) and if, as was 

 then further maintained, there are found among such species 

 bodies belonging to progressive series having as a common differ- 

 ence not only CH 2 , OH a and OM, but M 2 , H 2 and 0, it becomes 

 apparent that the chemical formulas of bodies so related may 

 present ratios much less simple than those hitherto generally 

 admitted. If, in fact, the relations perceived in surbasic salts, 

 such as the so-called subnitrates and subsulphates, which differ 

 among themselves by ??(M0) be extended to other compounds, 

 and if oxyds with elevated formulas differ among themselves 

 by ??H 2 , wM or nO we may have not only n(MO), n(M s 4 ), 

 n(M 2 0* and n(MO s ) which, if ?i=24, will be 



M 24 24 ---M 24 3 -s----Mo 4 3e ____M, 4 46 

 but bodies intermediate between these, the general formula 

 being M a4 24 -f nO. The coefficient of M for oxyds like peri- 

 clase. magnetite, hematite, polianite, quartz and cassiterite, will 

 however not be 24, but approximately 2400. 



§ 12. The existence of oxyds of such intermediate compo- 

 sition has lately been insisted upon by Schiitzenberger, who 

 shows that oxyds of tin, mercury, copper, lead, zinc, manganese 

 and iron, when prepared in different ways, vary considerably 

 in their proportions of oxygen. Thus cupric oxyd from the 

 calcined nitrate evolves oxygen when dissolved in nitric acid, 

 while ferric oxyd got by a similar process, if we assume the 

 formula to be Fe 2 O s , gives the value of Fe = 54, while for a sim- 

 ilar oxyd from the calcined oxalate, if we admit the same 

 formula, we get Fe=56. In other words, 24 parts of oxygen 

 are in the first case united with 54, and in the second with 56 

 parts of iron. 



Boutlerow, who confirms the results of Schiitzenberger, 

 concludes from similar studies that the combining weight of 

 carbon may vary from 12"0 to 11*8, and supposes a change in 

 the chemical value of that element ; that is to say, the amount 

 of carbon united with 32 parts of oxygen may vary from 12 - 

 to 11*8 parts, the resulting compounds, though not identical, 

 being similar in chemical properties. Schiitzenberger, in like 

 manner, admits the variability in value on either side of a 

 point of maximum stability, which is in most cases attained. 

 He concludes that " all of these results lead directly to the con- 

 clusion that the law of definite proportions is not rigorously 

 exact, unless we are willing to admit in each particular case 

 the existence of compounds more or less oxygenized than those 

 hitherto known to us, which may occur mixed with the pro- 

 ducts of the reaction."* This suggested explanation of the 



* See for the papers of Schiitzenberger and Boutlerow, Bull, de la Societe 

 Chimique de Paris, 1883, xxxix., 257-263, also an extended analysis of them in 

 this Journal, in the same year, xxvi, 63. The results obtained, according to 



