90 PHILOSOPHICAL SOCIETY OF WASHINGTON. 
while the other law, though largely predominant at the present 
period, shows a few conspicuous failures at the present time, but a 
very large number of them in times past. 
Another hypothesis to account for volcanic energy supposes the 
interior of the earth to consist of unoxidized elements, which grad- 
ually become oxidized by the penetration of oxygen from the at- 
mosphere. 
The objections to this hypothesis are as follows: On the assump- 
tion that the earth acquires no oxygen from space, the primitive 
atmosphere would have been many thousand times greater than at 
present ; but the geological record argues strongly in favor of an 
atmosphere which may indeed have varied in quantity and compo- 
sition, but nowhere near so greatly as the hypothesis implies. Any 
such extravagant difference would have recorded itself legibly in 
the strata. Furthermore, on this view, the end of all volcanic ac- 
tivity is close at hand. Only three pounds of oxygen to the square 
inch of terrestrial surface are left. A few hundred or thousand 
centuries and the last volcanic beacon is extinguished, and with it 
all organic life. 
But suppose the earth gathers up oxygen in its march through 
space. This may be true, but we can make any supposition on this 
point which pleases our fancy and feel sure that no prudent scien- 
tific man will dispute it. 
A third hypothesis is that of the late Robert Mallet, which as- 
sumes the earth to be contracting interiorly by asecular loss of prim- 
itive heat. As the interior cools and shrinks, the external shell is 
crushed and crumpled together, and this mechanical crushing is a 
sufficient source of heat. 
To this hypothesis there are many answers. The most direct one 
is that the very facts which are relied upon to prove that there is - 
any interior cooling at all now going on also prove that the amount 
hitherto has been excedingly small, and has been limited as yet to 
a thin external shell, not exceeding 150 miles in thickness, while 
the great interior is about as hot as ever; but, by the terms of the 
hypothesis, if the interior has not cooled there has been no interior 
contraction. The hypothesis is refuted by taking its own premises 
and pushing them to their inevitable conclusions. 
There is a fourth hypothesis, which cuts the Gordian knot in- 
stead of untying it. It assumes, as the result of causes unexplained, 
heat is generated locally within the earth, and such local movements 
