673.  On the other hand, bodies dependent on weak affinities very rarely give way.  Take, for instance, glasses:  many of those formed of silica, lime, alkali, and oxide of lead, may be considered as little more than solutions of substances one in another[A].  If bottle-glass be fused, and subjected to the voltaic pile, it does not appear to be at all decomposed (408.).  If flint glass, which contains substances more directly opposed, be operated upon, it suffers some decomposition; and if borate of lead glass, which is a definite chemical compound, be experimented with, it readily yields up its elements (408.).

  [A] Philosophical Transactions, 1830, p. 49.

674.  But the result which is found to be so striking in the instances quoted is not at all borne out by reference to other cases where a similar consequence might have been expected.  It may be said, that my own theory of electro-chemical decomposition would lead to the expectation that all compound bodies should give way under the influence of the electric current with a facility proportionate to the strength of the affinity by which their elements, either proximate or ultimate, are combined.  I am not sure that that follows as a consequence of the theory; but if the objection is supposed to be one presented by the facts, I have no doubt it will be removed when we obtain a more intimate acquaintance with, and precise idea of, the nature of chemical affinity and the mode of action of an electric current over it (518. 524.):  besides which, it is just as directly opposed to any other theory of electro-chemical decomposition as the one I have propounded; for if it be admitted, as is generally the case, that the more directly bodies are opposed to each other in their attractive forces, the more powerfully do they combine, then the objection applies with equal force to any of the theories of electrolyzation which have been considered, and is an addition to those which I have taken against them.

675.  Amongst powerful compounds which are not decomposed, boracic acids stand prominent (408.).  Then again, the iodide of sulphur, and the chlorides of sulphur, phosphorus, and carbon, are not decomposable under common circumstances, though their elements are of a nature which would lead to a contrary expectation.  Chloride of antimony (402. 690.), the hydro-carbons, acetic acid, ammonia, and many other bodies undecomposable by the voltaic pile, would seem to be formed by an affinity sufficiently strong to indicate that the elements were so far contrasted in their nature as to sanction the expectation that, the pile would separate them, especially as in some cases of mere solution (530. 544.), where the affinity must by comparison be very weak, separation takes place[A].

  [A] With regard to solution, I have met with some reasons for
  supposing that it will probably disappear as a cause of transference,
  and intend resuming the consideration at a convenient opportunity.