46 Explanatory Reports on [January, 
film holds the full dose that the metal can obtain by such ex- 
posure, and forms a compaCt continuous varnish, protecting the 
metal below from further oxidation. Copper and its alloys 
oxidise in a similar manner, and lead protects itself still more 
effectually by ultimately obtaining a varnishing film of sulphide. 
Iron behaves very differently. It is capable of forming a series 
of oxides, and some of these oxides combine with each other as 
acid with base, besides which the higher oxides may act as 
oxygen carriers for the oxidation of the metal or the further 
oxidation of its lower oxides. 
The through rusting of iron depends on these actions : — 
Exposed to air and moisture the protoxide, or as some affirm the 
protocarbonate, is first formed ; this rapidly passes on to the 
hydrated sesquioxide, or well-known red rust. This red com- 
pound, besides being a feeble acid, readily gives up some of 
its oxygen to bodies capable of taking it, and thus a coat of iron 
rust, instead of protecting the metal below, as in the case of the 
oxide films of zinc, bronze, &c., has the opposite effect of acting 
as oxygen carrier. Dr. Miller says that when rusting has begun 
“ moisture is absorbed from the air by the oxide, and thus a 
species of voltaic action is produced, the oxide performing the 
part of an electro-negative element, whilst the iron becomes 
electro-positive and the atmospheric moisture acts as the exciting 
liquid.” Certain it is that, once started, the rusting of iron pro- 
ceeds inwards, and but limited success has attended the efforts 
hitherto made to prevent it. 
In this, as in so many other instances, the study of natural 
phenomena might have solved the problem, for there is a natural 
compound that is not at all addicted to rusting, even under cir- 
cumstances of the severest exposure. This is the “ loadstone ” 
or magnetic oxide of iron of composition intermediate between 
the protoxide and sesquioxide, and by some regarded as a com- 
pound of protoxide and sesquioxide. I have now before me a 
sample of the ironsand of Taranaki (New Zealand) which, if I 
am rightly informed, is collected on the sea-shore. It is com- 
posed of grains of this black oxide that are separable by the 
magnet from the associated siliceous grains. The most careful 
examination reveals no indication of rusting on any of these, in 
spite of the excessively severe test to which they have been 
subjected for ages. 
Prof. Barff’s process consists in forming this oxide upon the 
surface of manufactured iron articles. Its mere production is 
easy enough : the essential element of the practical problem is 
to obtain it as an adherent and perfectly continuous coating, for 
if it is perforated by the smallest pin-holes, communicating with 
the metal below, further oxidation will commence at these 
openings, and the resulting rust will presently undermine the 
protective coating and spread itself around. 
A film of magnetic oxide may be formed by merely heating 
iron in air ; ordinary iron scale is a rude example of this. 
