22 Temperature of the Terrestrial Globe. 
percolating through crevices of the rocks, become new sources 
of partial heat, which being found incomparably nearer the sur- 
face than the centers of volcanoes, may sensibly increase, in cer- 
tain places, the interior temperature at depths which have been reach- 
ed with the thermometer. ‘These are the waters which furnish us 
with mineral springs, warm if the water has only a short distance to 
flow before it comes to light, and if the action of the water upon the 
pyrites is energetic and the quantity of water small,—cold if the in- 
tensity of the action is weak, the course long and the quantity of 
flowing water considerable. I do not rank in this class the Geyser, 
the Ricum and other spouting fountains of Iceland, which I regard as 
immediate products of the volcanoes of that island. 
We must also reckon among the number of causes of continental 
heat, fossil coal, mineral coal of all kinds, the antique remains of an 
immense vegetation which had existed in the latter periods of the 
formation of the crust, inflammable substances which are not (even 
at a moderate temperature) indifferent to the oxygen, either of the 
atmosphere, or of the water which finds access to them, and whose 
action may increase even to a kind of volcanic activity, such as are 
offered us by the sacred fires of Bakou, and other analogous appear- 
ances along the shore of the Caspian and Dead Seas; to which 
we may add those parts of the earth where petroleum and bitumen 
are formed, as in France, Italy, Germany, Bohemia, China, and 
North America. 
I do not instance the heat produced by the great work of general 
precipitation, for the first idea of which we are indebted to M. 
de Humboldt, because this temperature, occurring over the whole 
surface of the globe, cannot serve to explain the local anomalies, 
which observation proves to exist. 
These considerations explain the great anomalies which are ob- 
served in mines relative to the progression of temperature in the in- 
terior of the earth. We may add to the principal causes now cited, 
the slow oxidation of metals in the metallic state, and the chemical 
changes which some oxides, and even rocks, may undergo, by air, 
water, and carbonic acid. From these it will necessarily result that 
the temperature of the exterior strata of continents should be a little 
higher than the corresponding beds of the ocean, even independently 
of the difference of action of the solar rays upon liquids and solids, 
and of the greater evaporation from the surface of the sea than from 
the surface of the earth, and that the difference of temperature should 
