958 REPORT—1885. 
will be felt in granting this of cuprous and stannous chlorides, and even of 
cadmium, lead, silver and zine chlorides; but opinions will differ as regards the 
metals of the alkalies and alkaline earths. Assuming the constitution of metallic 
electrolytes to be such as I have suggested, it is not improbable that on electrolysis 
a part only of the metal is determined to the one pole, the remainder being trans- 
ferred alone with the negative radical to the opposite pole. Hittorf, indeed, has 
already put forward this view in explanation of the remarkable results he obtained 
on determining the extent of transfer of the ions in aqueous and alcoholic solutions 
of the chloride and iodide of cadmium and zinc. 
Again, an argument in favour of a connexion between chemical constitution 
and electrical conductivity is the fact that carbon, sulphur, selenium and phos- 
phorus each exist in conducting and non-conducting modifications, as it can 
scarcely be doubted that the so-called allotropic modifications of these elements 
are differently constituted. 
It appears, as I have already said, to be the current belief that when aqueous 
solutions are submitted to electrolysis, as a rule, the dissolved substance, aud not the 
water, is the actual electrolyte. Without reference to the question I have raised as 
to the constitution of an electrolyte, it appears at least doubtful whether this view 
can be justified by appeal to known facts; at all events, I have failed to find satis- 
factory evidence that such is the case. Moreover, as sulphuric anhydride dissolves 
in water with considerable development of heat, it would appear that more work 
has to be done to separate hydrogen from sulphuric acid than to separate it from 
water; on this account we might expect that the water rather than the acid would 
be decomposed. Are not perhaps both affected according to the proportions in 
which they are present? The marked variation in the extent to which the negative 
ion is transferred to the positive pole, as observed by Hittorf, when solutions of 
different degrees of concentration are electrolysed, would appear to support this 
view. The difference in the products, according as dilute or very concentrated 
solutions of sulphuric acid are used, may also be cited as an argument that the 
chemical changes effected vary with the concentration; but, on the other hand, it 
is quite possible that the observed differences may result from the occurrence of 
purely secondary changes. Ostwald has recently put forward the view that one or 
more of the hydrogen atoms of certain acids are split off according to the concen- 
tration of the solution. 
I call attention to this because I conceive that it has a most important bearing 
on the discussion of the nature of the chemical changes which occur during the 
dissolution of metals. Formerly it was said that, when zinc acts upon dilute sul- 
phuric acid, the zine displaces the hydrogen of the water and the resulting zinc oxide 
dissolves in the acid, forming zinc sulphate ; the modern explanation advocated by 
most chemists has been that the metal directly displaces the hydrogen of the’acid: in 
fact, that this is the nature of the change whenever an acid is acted upon by « metal. 
Tf in a solution of sulphuric acid, of whatever strength, the acid be the actual elec- 
trolyte, I imagine that we are right in accepting this modern view ; but if the water 
be the electrolyte, we must, to be consistent, return to the view that the oxide—more 
probably in most cases the hydroxide—is the primary product. And if it can be 
1 We may regard.as evidence in support of this explanation the fact that neither 
beryllium chloride, which fuses at 600°, nor mercuric chloride, is an electrolyte, as 
both of these, at temperatures not far removed from their boiling-points, exhibit the 
simplest possible molecular composition. It should be pointed out, however, that 
Nilson and Patterson found it possible to determine the density of beryllium chloride 
gas at a temperature 100°-150° below the melting-point found by Carnelly; but they 
were not able to say that fusion took place. Clarke’s recent interesting observations 
on mercuric chloride and iodide do not, I think, suffice to prove that these com- 
pounds are electrolytes; it is more than probable that electrolysis is preceded by 
the formation of mercurous compounds. Even an aqueous solution of mercuric 
chloride does not conduct appreciably better than water (Buff). I should perhaps 
add that the mere presence of more than a single atom of metal in the molecule 
does not, I believe, alone constitute the compound an electrolyte; much depends 
probably both on the nature of the metal and on the structure of the molecule. 
