18 Experiments for ascertaining [Jan., 
in the results. From the following analysis, it will be observed 
that at the lowest depths the rate of imcrease was nearly as great as 
at any portion of the descent :— 
The first observations gave 51° as the invariable temperature at a depth of 164 
feet, as already stated. 
Between 693 feet and 710 feet the temperature was nearly uniform at 58°, 
Between 710 and 927 feet the rate of increase was 1° for 62°4 feet. 
Between 927 and 1,257 feet the rate was 1° for 60 feet. 
Between 1,257 and 1,839 feet the rate was 1° for 86°91 feet. 
Between 1,839 and 2,055 feet the rate was 1° for 65°6 feet. 
And the mean of the whole series of observations gives 1° for every 
83-2 feet, which is under the average of the observations. 
The increase of temperature appears to be independent of alti- 
tude, of the place, or of the density of the air, as proved by the 
observations of Humboldt.* ‘Thus, in a silver mine in the Andes of 
Peru, at an elevation of 11,875 English feet above the sea, the 
temperature was found to be 25:4° Fahr. higher than the external 
air. Inanother mine at the same elevation, the difference between 
the temperature of the internal and external air was found to be 
15°8° Fahr., and the water streaming down the rock showed 52°3° 
Fahr. These observations were repeated in other localities with 
similar results. 
The concurrent testimony, therefore, of all the observations 
which have been made, is in favour of an increase of temperature, 
though at rates varying considerably from one another, and this 
has been established as holding good down to a depth of about 
2,400 feet from the surface. What seems to us now to be required, 
and for several reasons, is a series of experiments made at even 
greater depths. 
First.—With so many recorded observations, all approximately 
concurring towards one conclusion, any sinkings or borings to a 
depth of less than 2,000 feet would, in all probability, only afford 
another series of similar results, and would be so much money use- 
lessly expended ; and secondly, it seems probable that at greater 
depths several sources of error would be avoided. Of all these 
sources of error—the percolation of water through the strata—is 
probably the most formidable ; now it has generally been found that 
the quantity of water in coal-mines derived by percolation from the 
surface, decreases in proportion to the depth, and that at depths 
greater than 500 or 600 yards the mines are dry. It is somewhat 
difficult to account for this where there are thick beds of porous 
sandstone, but it is to be recollected that a rock which is full of 
joints and fissures at the surface, is often found extremely solid at 
a considerable depth, and the friction, or resistance to percolation, 
* ¢ Cosmos,’ Sabine’s Trans., vol. iv. 
