THE RUSTING OF IRON AND STEEL. 
261 
long disquisition about forming chromic oxide and chromic chromates, and when I 
told him that no action at all takes place he was very much surprised. 
Professor Tilden : It is very good of you, Sir, to allow me to join in an ex¬ 
pression of great pleasure at the interesting discourse that we have listened to 
this afternoon. I entirely agree with Professor McLeod that Professor Dunstan 
has gone a long way towards establishing his proposition. At the same time I 
feel, as I suppose Professor Dunstan will be the first to admit, that the in¬ 
vestigation is not yet quite complete. I am not quite clear indeed from what he 
said whether in bringing forward what appears to be a perfectly efficient theory 
of rusting he intends it absolutely to displace the old theory in all cases, or 
whether he would still allow that in some circumstances a solution of carbonic 
acid and oxygen alone would be competent to act upon iron. I was quite sur¬ 
prised to hear that carbonic acid actually gave rise to the evolution of hydrogen, 
and when Professor Dunstan shewed us the powder stirred up in carbonic acid and 
water I wondered whether he had taken the trouble actually to ascertain whether 
it was really hydrogen gas. The notion I have always had about the action of 
carbonic acid is that it co-operated with oxygen and I should not have expressed 
the action by the equation which is first written down on the diagram and which 
is attributed to Dr. Calvert and Professor Crum Brown. I should have preferred 
to add to the first equation “ oxygen 55 which would result in the formation not 
of hydrogen but of water. I should have thought that carbonic acid alone was 
scarcely competent to produce ferrous carbonate. However, this is a point which 
Professor Dunstan has probably already investigated. 
I have never myself studied specially the question of the rusting of iron but I 
devoted some considerable time a few years ago to the somewhat kindred in¬ 
vestigation of the action of water and air and dissolved salts upon copper and 
copper alloys and I learnt a number of facts from the investigation, which were 
shortly afterwards published in the “ Journal of the Society of Chemical Industry,’’ 
and which seemed to bear to a certain extent upon some of the points which Pro¬ 
fessor Dunstan has raised to-day. Bor instance with regard to the influence of these 
salts, among other agents that I made use of was a solution of common salt, 
Sodium chloride. I compared the action of Sodium chloride with that of Mag¬ 
nesium chloride, which I quite expected would have been much more effective as 
the corrosive agent; to my surprise the common salt acted more efficiently than 
the magnesium salt in the form of solution in water and this, after a little con¬ 
sideration, it was not difficult to explain. What happens really is that atmospheric 
oxygen in contact with the Sodium chloride gives rise to a separation of the 
alkaline metal sodium in the form of oxide, while the chlorine of the chloride 
attacks the surface of the metal, and that salt was found to be most efficient 
which contains the metal which is most positive, that is to say which tends to 
combine oxygen most energetically, and I imagine that some other salts in that 
table on the wall would perhaps act in a similar way. 
I also found unquestionable evidence of the electrical influence of the substance 
which is formed as the product of the rusting on the surface of the metal, pro¬ 
vided that it remains in a solid form attached to the metal; that is to say oxy¬ 
chloride or oxy-carbonate or whatever it was, in fact what people call verdigris, 
in the presence of and contact with the copper tends to the further development 
of the same action ; and I wonder whether Professor Dunstan has considered 
whether it is not probable that rust in contact with iron forms a voltaic couple 
which may result possibly in the slow decomposition of the water. Oxide of iron 
is certainly capable of acting negatively towards the metal which remains un¬ 
rusted and I suppose it is quite certain that when once a metal has begun to rust 
the process goes on at an accelerated rate. 
