Behaviour of Chemical Compounds. 163 



conform to the electrolytic deportment of the compounds. 

 While in the case of non-electrolyzable substances considera- 

 tions upon electrical contrarieties of their constituents, without 

 experimental foundation, can only be the subject of purely 

 theoretical speculations, the separation, effected by the same 

 current, of various electrolytes into their ions proves that the 

 latter contain equal quantities of positive or negative electri- 

 city (are electrolytically equivalent). This equiA'alence is not 

 everywhere expressed by the structural formula : thus in the 

 compounds FeCl2 and Fe2 Clg equal quantities of chlorine 

 (2 CI), on the one hand, and the different quantities of iron 

 (Fe and f Fe) combined with them, on the other, are equiva- 

 lent. It is the same with chloride of potassium and the yel- 

 low prussiate of potash, to which are assigned the formulae 

 KCl and 4 K + FeCyg, while yet the equivalence of equal quan- 

 tities of potassium in them is ascertained, and so on. These 

 electrolytic relations of the compounds, however, would surely 

 have as much right to be taken into account, in the conside- 

 ration of their structure, as their chemical properties ; and it 

 is obvious that a complete knowledge of the constitution must 

 give a full account of them as well as of the latter. 



Similar relations present themselves on the consideration 

 of the magnetic properties of compounds. While in the che- 

 mical and electrolytic investigation of bodies it is always the 

 result of a displacement or separation of their atoms that is 

 discovered, and consequently only at the beginning and end 

 of certain states of motion is the way in which they are com- 

 bined considered, the magnetic investigation of compounds 

 offers the advantage that we can study a physical property 

 altogether characteristic of the individual atoms of the com- 

 pounds without any decomposition, on the unaltered compounds, 

 and can therefore draw a conclusion respecting the particular 

 behaviour of those atoms in them. 



Permit me, in the following, to epitomize the communica- 

 tions which I have previously published on the magnetic be- 

 haviour of chemical compounds, together with an account of 

 more recent observations on the same subject. Perchance 

 this combination may furnish a fresh instance of even a purely 

 physical method of investigation sometimes giving a revela- 

 tion, in certain directions, respecting the constitution of che- 

 mical compounds, such as could never be attained through 

 purely chemical reactions exclusively. 



The magnetic moment of the bodies experimented on was 

 measured by means of a torsion balance. The head of it could 

 be rotated about its axis, in a vertical hollow cylinder, by 



M 2 



