ON THE ACTION OF AIR AND WATER UPON IRON. 257 
hence it is the more moistened, and therefore the more cor- 
roded. These modifying conditions apply, whether iron be in 
contact with zinc or not. 
238. Water is always decomposed by iron or zinc in metallic 
contact as soon as oxidation of either metal has commenced, 
and hydrogen is at first absorbed by the water, and when this 
is saturated, evolved; but oxidation will not commence at all 
either on metal in an hermetically sealed vessel of water free 
from air, or on a body acting inits capacity as a peroxide. Hence, 
while air is constantly requisite to maintain the power of de- 
composing the water, it is not by the decomposition of the air 
alone that the iron or the zinc is oxidized, as was maintained 
by Dr. Marshall Hall*. 
239. In my former Report+, I alluded to the effect of 
covering surfaces of neutral solids, as glass, &c. (beneath which 
the solvent fluid penetrated), in arresting corrosion. In the 
progress of these experiments, however, I have observed some 
curious modifications of this condition. 
240. If a clean surface of iron immersed in water be covered 
with a parallel surface of plate glass, leaving a film of water 
between, oxidation will not take place between the glass and 
the iron, at least for a great length of time; it very gradually 
creeps inwards from the edges, forming patches of green inter- 
mediate oxide ; but if in place of the plane of glass a glass lens 
of large curvature, and thus making very small angles with the 
surface of the iron, be placed upon it, oxidation will commence 
at the point of contact, and will spread from thence, although 
the iron may be in such a condition, that if no glass or other 
neutral solid were in contact with it at all, oxidation would 
_ just not take place. 
241. So that, in general, whether a neutral solid prevent or 
_ promote oxidation, depends upon its position in relation to the 
surface of the metal. This fact seems to belong to the as yet 
not understood power, in promoting chemical action, which 
extremely small orifices or fissures seem to possess, as in the 
action of spongy platina, of pyrophorus, of porous bedies in 
_ the condensation and the diffusion of gases, endosmose and 
exosmose through capillary tubes, and so forth. It has long 
_ been observed, that in a crystallizing solution, crystals first 
form at acute angles and on salient points. 
242. If a plane of polished iron, or other oxidable metal, be 
fixed, forming a very acute angle with a plate of glass, ivory, 
&c., and both plunged into water, oxidation commences at the 
angle first and spreads from it, whatever be the position of the 
* First Report, § 11. + § 121. 
1840, S 
