286 M. Gay Lussac’s Reflections on Volcanoes . 
gen of volcanoes, should occur in lavas, since hydrogen has the 
property of reducing its oxides at a high temperature ? It is 
certain, at least, that the iron does not occur in the state of a 
peroxide ; for they act powerfully on the magnetic needle, and 
the iron appears to be precisely at the degree of oxidation which 
the water alone determines ; that is to say, in the state of a deut- 
oxide. Further, to prevent the hydrogen from reducing the 
oxides of iron, it would be sufficient, as I have shewn, that it 
were mixed with several times its volume of aqueous vapour. 
The necessity, as appears to me, of water penetrating into the 
foci of volcanoes, the presence, in lavas, of some hundred parts 
of soda, that of muriate of soda, and of several other chlorides, 
render it very probable that it is the water of the sea which most 
commonly penetrates. But an objection occurs which I must 
not pass over : it is this, that the lava ought to make its egress 
by the canals through which the water has entered, since it 
would find in them a less resistance than in those by which it is 
raised above the surface of the earth. It should, in this case, 
also frequently happen, that the elastic fluids formed in the foci 
of volcanoes, previous to the ascent of the lava to the surface of 
the earth, should boil forth by the same canals in some places of 
the surface of the sea, — -a circumstance which I am not aware of 
having been yet observed, although it is probable that it is these 
elastic fluids which produce the mephitic gases so common in 
volcanic countries. 
On the other hand, it might be remarked, that the long inter- 
missions of volcanoes, and their state of repose during a great 
number of years, seem to shew that their fires are extinguished, 
or at least smothered ; then the water would gradually penetrate, 
by its pressure, into imperceptible fissures, to great depths in the 
interior of the earth, and would accumulate in the vast cavities 
which it contains. The volcanic fires would gradually kindle 
again ; and the lava, after having obstructed the canals by which 
the water penetrated, would be raised by its usual orifice, the 
diameter of which would increase continually by the fusion of 
its walls. These are merely conjectures ; but it is pretty certain 
that water actually exists in the foci of volcanoes. 
We see how the theory of volcanoes is still uncertain. While 
we have strong reasons to think that the earth contains sub- 
