ISOMORPHOUS RELATIONS OF MANGANESE. 541 



more precise, the view of voltaic action maintained in this work, 

 I shall place in a note below a concise statement of the principles 

 of that view, in its amended form.* 



* This modification of the chemical theory of the voltaic circle, which dis- 

 penses with any electrical hypothesis, is founded upon the three following 

 postulates : 



I. The binary constitution of salts, which has already been fully discussed 

 (page 160.) This applies to the fluid portions of the circle, and its assump- 

 tion is equally necessary on the usually received electro-chemical theory of the 

 circle. 



II. The Sali-molecular structure of metals. By this is meant that the metals 

 are composed of molecules or groups of three atoms, having a binary or saline 

 character, as explained above. The metallic and fluid portions of the circle 

 are thus assimilated in constitution. A decomposition can be propagated in 

 any direction through the fluid portion of the circle, owing to the mobility of 

 each particle, which permits it to take the new position required with a change 

 of the direction in which the decomposing force is made to act, (page 204.) 

 But decomposition is propagated, in both directions, through a chain of me- 

 tallic molecules also, although solid, and therefore without the same power of 

 adjustment. To explain this, it must be supposed that an internal decom- 

 position can readily take place in the metallic molecule itself ; that in respect 

 of its three atoms, A, B and C, A forming the zincous element, and B + C the 

 chlorous element, a change can easily occur, in which C becomes the zincous 

 element of the sali-molecule, and A + B the chlorous element ; that, supposing 

 the three atoms of the molecules disposed in a line, A, B, C, any of its saline 

 elements may be either to the right or left, as A + BC, or AB + C. The three 

 atoms of the molecule being of one metal, and of the same nature, may admit 

 of this change of internal arrangement, by a substitution of one atom for an- 

 other. 



Several circumstances favour the idea of the existence of the assumed con- 

 dition of metals : 1. In iron the susceptibility of magnetism is confined to the 

 metal itself and one degree of oxidation, the black oxide, with its correspond- 

 ing sulphuret. This is the degree of oxidation into which iron most readily 

 passes ; it consists of single equivalents of the protoxide and peroxide, or of 

 three atoms of iron and four of oxygen. There is oxidation, in its formation, 

 without disturbance of the metallic sali-molecule, Fe + Fe 2 ; the zincous ele- 

 ment, Fe, combining with 1 eq. of oxygen, to form Fe O, and the chlorous 

 element, Fe 2 , with 3 eq. of oxygen, to form Fe 2 O 3 ; and these two oxides 

 themselves remaining in a state of union. Metallic iron having, therefore, a 

 common magnetic character with the black oxide of iron, or the loadstone, 

 which has three atoms of metal in its molecule, may well be supposed to 

 have three also. It is worthy of passing remark, that this double oxide is 

 peculiar to the magnetic metals. It may not be an idle hope to look for the 

 elucidation of the cause of magnetism in the peculiarities of the molecular 

 structure of iron. 



2. It is supported by the disturbance of chemical affinities, or the electrical 



