206 REPORT— 1890. 
anode solution and the dilute solution. But I found that with the electro- 
motive force at my disposal (50 volts) I was unable to determine any 
such layer of hydrate. The experiments were, however, not conclusive, 
for the hydrate is to a certain extent soluble in copper sulphate; over- 
looking this defect, the dilution seemed so to diminish the current as to 
make it weak enough for the sulphate molecules to carry it. 
The rough agreement of Weber’s results with the values of ionic 
velocities deduced by other methods is inconclusive, for the impoverish- 
ment of the solution would be itself dependent upon the ionic velocities, 
and hence the results deduced from the limit of current density would 
depend on the velocities—directly, upon the one hypothesis, and indirectly, 
upon the other. 
Summing up the results of this section, so far as regards a fused 
electrolyte or the solution of a single salt, we may say that the ions have 
hitherto been determined from the results of the chemical analysis of the 
liquid in the anode and cathode vessel, with the tacit understanding that 
the electrolysis shall be regarded as that due to the resolution of single 
molecules into ions which are atoms, or their chemical representative 
radicles, unless the observations are such as to make such a view entirely 
untenable, even when the dilution is referred to unequal motion of the 
ions and not to complex molecular decomposition. Thus every chemical 
determination of the ions should imply a determination of the constant of 
migration; and when the dilution at one electrode is so rapid that to 
apply the hypothesis of unequal ionic motion successfully would require 
us to assume the velocity of one ion to be negative, that is, that the ion 
would have to be moved against the electrical forces acting upon it,' then 
the decomposed molecule may be regarded as compound; one ion is 
assumed to have associated with it one or more molecules, as may be 
necessary, of undecomposed salt. Thus a critical consideration of the 
ions in electrolysis leads us to the question of the migration of ions. 
There are, however, cases in which the ions corresponding to the simple 
molecular decomposition can be comparatively easily inferred. The 
results of a number of determinations of ions are given in the table on 
the next page (Table IV.). 
A confirmation of the results obtained may be derived from the 
electrolysis of solutions in series, in which case the anion of one solution 
combines with the cation of the adjacent one. The results exhibited in 
the table show the amounts of the respective ions corresponding to the 
deposition of one equivalent of hydrogen in a voltameter, so that they 
may be also regarded as showing the application of the second part of 
Faraday’s law. 
One warning must be given about such determinations. In order to 
determine both the ions, both the anode and cathode vessel must be 
separately analysed. The analysis of one alone is not sufficient. For a 
salt such as Na,;PO, may be decomposed by solution into NaH,PO, and 
Na,HPO,, and the electrolysis be different from what it would be if the 
deposition of Na were established, and the second ion inferred from the 
composition of the original salt. I do not think that all the results 
quoted in the table have been subjected to minute criticism from the 
1 The force upon an atom or group of atoms carrying a charge +e would be 
av 
“da 
the direction towards the cathode. 
dV 
» when Te. is the slope of potential per unit of length from anode to cathode, in 
