ON THE SOURCES OF HEAT WHICH INFLUENCE CLIMATE. 
463 
gravity would be obtained greater than that which, from astronomical data, we 
know our globe to possess; and as their density would go on increasing as we 
continue to recede from the surface, he was led to conjecture that the terrestrial 
sphere had a hollow centre. lie further came to the striking result, that air will, 
at considerable depths, from its higher elasticity, become heavier than water; if 
we then grant that water at the greatest depths of the ocean contains air in 
solution, as we find it to do on the surface, Leslie’s theory will tend to the 
conclusion that the ocean rolls on a bed of air. He adopts this view, and 
endeavours to connect it with some of the phenomena of volcanoes. 
To suppose the seats of volcanic fire at the depths Leslie has done, vvould 
bring us to internal pressures on parts of the crust of our globe exceeding, on 
the most moderate scale of computation, 20,000 atmospheres. This appears to 
me to render it extremely improbable that the seats of volcanoes are deep in the 
earth’s surface. In practice we know of no pressures equal to 100 atmospheres; 
we also find how difficult it is to construct metallic vessels able to resist the force 
of a few atmospheres ; the materials of the crust of the globe are, moreover, of 
a composition not well calculated for the resistance of pressure. Yet there is no 
doubt, from the data furnished us by volcanic eruptions, that during such events 
a small portion at least of the earth’s crust must be capable of overcoming a con¬ 
siderable force. 
We have only to suppose the centre of volcanic fire to be 2,100 feet below the 
orifice of a crater in activity, to have there an internal pressure of 100 atmos¬ 
pheres^ There are on record many eruptions where lava was discharged in large 
quantities that must .have proceeded from depths greater than this; but I am 
much inclined to assign to the centres of volcanic eruptions a position as near the 
earth’s surface as is consistent with recorded phenomena. I am very hostile to 
connecting them with the incandescent matter occupying the earth’s centre, 
chiefly from the consideration that they would then communicate when in an 
active state an enormous force over every point of the interior of the inclosing 
crust ; there are other reasons that render such an hypothesis extremely im¬ 
probable. 
Fluids, wherever situated, must be influenced by the all-pervading law of 
gravitation. The attraction of the sun and moon will develope tidal waves in 
every considerable mass of fluid matter; hence, in the event of a crater com¬ 
municating -with such a mass, we should look for its discharge being varied by 
tides ; from this circumstance I think it can be proved that they are not in 
connexion with any great quantity of fluid. 
The existence of internal tides bears on another question. If the solid surface 
of the earth be comparatively only a thin crust, we should have a powerful 
diurnal tide, which would occasion violent dislocations in the strata. This view 
3 p 2 
