686.  That these mineral acids should confer facility of conduction and decomposition on water, is no proof that they are competent to favour and suffer these actions in themselves.  Boracic acid does the same thing, though not decomposable.  M. de la Rive has pointed out that chlorine has this power also; but being to us an elementary substance, it cannot be due to its capability of suffering decomposition.

687. Chloride of sulphur does not conduct, nor is it decomposed.  It consists of single proportionals of its elements, but is not on that account an exception to the rule (679.), which does not affirm that all compounds of single proportionals of elements are decomposable, but that such as are decomposable are so constituted.

688. Protochloride of phosphorus does not conduct nor become decomposed.

689. Protochloride of carbon does not conduct nor suffer decomposition.  In association with this substance, I submitted the hydro-chloride of carbon from olefiant gas and chlorine to the action of the electric current; but it also refused to conduct or yield up its elements.

600.  With regard to the exceptions (679.), upon closer examination some of them disappear.  Chloride of antimony (a compound of one proportional of antimony and one and a half of chlorine) of recent preparation was put into a tube (fig. 68.) (789.), and submitted when fused to the action of the current, the positive electrode being of plumbago.  No electricity passed, and no appearance of decomposition was visible at first; but when the positive and negative electrodes were brought very near each other in the chloride, then a feeble action occurred and a feeble current passed.  The effect altogether was so small (although quite amenable to the law before given (394.)), and so unlike the decomposition and conduction occurring in all the other cases, that I attribute it to the presence of a minute quantity of water, (for which this and many other chlorides have strong attractions, producing hydrated chlorides,) or perhaps of a true protochloride consisting of single proportionals (695, 796.).

691. Periodide of mercury being examined in the same manner, was found most distinctly to insulate whilst solid, but conduct when fluid, according to the law of liquido-conduction (402.); but there was no appearance of decomposition.  No iodine appeared at the anode, nor mercury or other substance at the cathode.  The case is, therefore, no exception to the rule, that only compounds of single proportionals are decomposable; but it is an exception, and I think the only one, to the statement, that all bodies subject to the law of liquido-conduction are decomposable.  I incline, however, to believe, that a portion of protiodide of mercury is retained dissolved in the periodide, and that to its slow decomposition the feeble conducting power is due.  Periodide would be formed, as a secondary result, at the anode; and the mercury at the cathode would also form, as a secondary result, protiodide.  Both these bodies would mingle with the fluid mass, and thus no final separation appear, notwithstanding the continued decomposition.