8 REPORT—1843, 
described; the action of moist air, however, produces both upon cast and 
wrought iron totally different forms of corrosion. Plumbago is not formed 
at all; rusting takes place with almost complete uniformity over the whole 
surface, and coat after coat of adherent rust is removed; at first each coat 
leaves a surface parallel to that of the original metal, but the whole surface 
gradually becomes fretted with minute concavities or indentations, at first 
scarcely visible, but gradually enlarging by several falling into one, until, after 
a lengthened period, the surface, originally flat or plane, is found covered with 
nearly hemispherical indentations often a quarter of an inch in diameter. 
This takes place whether atmospheric moisture lodges on the surface or not, 
and on both cast and wrought iron; its cause seems difficult of explanation. 
304. The eudiometrical properties of water, by which, whether in the state 
of snow-water, as observed by Boussingault, or of rain, it contains a portion 
of free oxygen, render it, as before observed, a powerful agent in promoting 
oxidation. 
Fresh-fallen rain after a time of drought, especially in cities, comes down 
so loaded with free oxygen, carbonic acid and ammoniacal salts, that it produces 
instantly a coat of red rust upon any iron placed in contact with it. Pure 
water, however large a portion of common air alone it contains, does not seem 
capable of producing any immediate oxide higher than the magnetic Fe O 
+ Fe, O,. The contrary is the case if carbonic acid or a minute quantity of 
any saline substance be in solution. 
305. But the deposition of dew under certain circumstances originates the 
most immediate and powerful oxidation, as the following observation testifies. 
On the 14th March, 1842, the temperature at Dublin, at 12 o'clock at 
noon, was high, and the day fine, but the air was nearly saturated with moist- 
ure, and dew rapidly collected on the polished parts of a large steam-engine 
which stood unfinished in a shady open building, whose temperature was 
considerably below that of the open air. In two hours time after, being 
wiped clean with cotton waste, all its bright work had a moist coating of red 
rust upon it. The rusty moisture could be swept off with the finger. 
Whether electrical disturbance of the atmosphere is concerned in this, or 
whether occasionally the atmospheric moisture is loaded with saline matter, 
especially near the sea, Iam unable to say ; but the fact of such rapid action 
of deposited dew is remarkable, and is not confined to a single instance, having 
been noticed also to me by engineers as occurring frequently at Liverpool. 
306. No second immersion of the cast iron series y was practicable, the 
arrangements of the Dublin and Kingstown Railway Company, which enabled 
the first to be made, having been unfortunately discontinued. 
307. Table XIV. shows the average results of the corrosion of all the cast 
irons in water on the second immersion, and those of the corrosion of cast 
iron in air (£). The numbers indicate that in general, in moist air, chilled 
cast iron corrodes much more slowly than that cast in green sand; the con- 
trary being, as before stated, the result in water. ' 
That the average loss on all varieties of cast iron in moist air is not much 
below that which takes place in clear sea water in an equal time, and is much 
above that produced by clear fresh water in the same period. 
In the case of cast iron with the skin removed by planing, the loss in moist 
air is almost precisely the same as in clear sea water. 
These facts show that the preservation of structures in iron exposed merely 
to the weather is much more important than has been hitherto presumed, and 
that without paint or some other more efficient covering, they perish almost 
as fast as if in the open sea. 
308. We now proceed to some remarks upon the series of experiments upon 
