and Electrolysis of Chemical Compounds* 449 



a copious amount of gas was immediately evolved, but no 

 vibratory motion was perceptible in the liquid : this (as 

 already remarked) appears only to come into play with a 

 certain degree of insulation and a determinate difference of 

 potential between the electrodes. It is very possible that we 

 have then to do with the discharges named by Faraday 

 " carrying discharges," in which the greater portion of the 

 electricity is mechanically conveyed by the molecules without 

 their splitting up as in the case of electrolysis. 



10. Conclusions. 



The chief object of this investigation was to establish the 

 possibility of electrolysis of a great variety of compounds, in 

 order to determine the part which might be taken by hydrogen 

 in electrolytic conductivity in consequence of its chemical 

 character. Now it appears from the foregoing experiments 

 that upon this definitive conclusions cannot be established ; 

 and even elective affinity has no preponderating influence. 



It is true that the hydracids, with the exception of hydro- 

 cyanic acid, are very bad conductors, and their electrolysis is 

 far from being proved ; their chemical properties in the 

 liquefied state are very feeble ; but this may result from ex- 

 traneous causes — for example, the very slight solvent power 

 of the liquefied gas. The frequent coincidence therewith of 

 the insulating property may therefore be regarded as to some 

 extent accidental. 



On the other hand, a certain number of organic compounds, 

 as the organo-metallic radicals and some haloid compounds of 

 the alcohol-radicals, readily exchange their metal or hydrogen 

 for other metals (or groups of atoms, as, for instance, in the 

 benzol series and most of the amines), and yet are bad con- 

 ductors. It is, however, remarkable that no compound with- 

 out hydrogen or a metal is known as a conductor, although 

 the mere presence of one of these elements seems not sufficient 

 to determine conductivity. The behaviour of the anhydrides 

 particularly supports this opinion. Also there do not appear 

 to be any substances that are good conductors without being 

 readily decomposed by chemical reactions. 



The view which is now more and more generally admitted 

 in physics, that electric conductivity is not a simple function 

 of the chemical composition, therefore receives fresh confirma- 

 tion from my experiments. An analogous conclusion in 

 regard to another class of compounds — namely, fused salts, 

 consequently bodies that are also liquid without a solvent — 

 has been established by Braun *. He concludes from his in- 

 vestigation that no connexion is recognizable between galvanic 

 * Pogg. Ann. vol. cliy. p. 195. 



Phil. Mag. S. 5. Yol. 5. No. 33. June 1878. 2 G 



