264 
THE HUSTINGS OF IKON AND STEEL. 
in the case of iron. As Professor McLeod has remarked, it is true that the 
hydrogen peroxide can scarcely exist for a second in the presence of iron; directly 
the iron is immersed in a weak solution it decomposes the hydrogen peroxide 
so fast that it is hardly to be expected that you would be able to detect it. 
It is true, that I spoke cautiously about the probable electrolytic nature of 
this process and it is not quite so clear as he seemed to think. We appear to 
have practically no second electrode. It is rather unusual to have a metal acting 
as one electrode and a gas acting as the other and oxygen negative to the iron. I 
think there is a great plausibility about it, but I spoke with some caution because 
it is a view that runs contrary to our common notions about electrolysis and it 
would need many more facts to prove that it was really electrolysis in the 
ordinary sense. 
Then Professor McLeod referred to the deposits of rust on some pieces of iron 
which he has kept in solution for a very long time,—I think he told us they 
were put in water in a solution of sodium chloride so long ago as 1863, 
and there he noticed in certain cases the greenish black stain. I may 
say that it is very difficult indeed under ordinary circumstances to detect 
the early formation of any ferrous oxide in the process. I have never succeeded 
in convincing myself that it can be detected ; but it is found if the action 
has gone on for some time, and it is always noticed on the surface of the 
iron which is below the rust; you have a layer of rust formed on the surface of 
the metal and you generally find that after a time some ferrous oxide has been 
formed at that point. It may be due, of course, to the reduction of ferric oxide 
by the metal itself, I do not think it is at all clear that the greenish stain he 
noticed is the first formed ferrous oxide of my equation ; I think it is much 
more likely that it is the result of reduction of the rust which has been lying in 
contact with the metal. 
Basic zinc hydroxide is formed when hydrogen peroxide acts on zinc. 
There is one other point that Professor McLeod referred to, to which he said I did 
not allude, namely, the support that my theory gains from the fact that sodium 
chloride and sulphate and other salts accelerate rusting. That may indicate, as he 
supposes, that the process is electrolytic. On the other hand some of these salts 
that prevent rusting also assist electrolysis and it is not yet proved that all 
sorts which in solution are good conductors accelerate rusting. 
Professor Tilden has asked me an important question, whether the old theory 
of rusting is to be thrown over altogether. 1 should say that so far as ordinary 
rusting is concerned undoubtedly it must be because carbonic acid gas is not 
necessary for rusting to occur. Ordinary rust, on analysis, contains a very 
small amount of carbonate and under ordinary conditions the part played by 
carbonic acid gas must be, I think, comparatively small. I think the 
amount of the effect which is produced by carbonic acid gas contained in the air 
is very small indeed as compared with the amount of action which is going on 
under the conditions summarised in these equations. The correct statement is 
that carbonic acid gas is not essentially concerned in the process of rusting. 
Water that has been exposed to air contains carbonic acid gas and that 
assists the process of rusting, in the way summarised by the equations of 
Professor Crum Brown; but rusting goes on quite well in the absence of it and 
I think there is hardly any doubt that the main effect must be attributed to the 
kind of change I have indicated. 
I accept Professor Crum Brown’s view that when iron is put into contact with 
carbonic acid gas ferrous carbonate is formed and hydrogen given off; and I 
believe that there are a number of statements by different observers to the 
