206 PROFESSOR FORBES ON THE TEMPERATURE OF THE EARTH. 
B. General Observations on the Thermometric Curves. 
Some of the most important results depend upon the annual ranges of tem- 
perature at different depths. But the determination of the extremes is no easy 
matter. I will first direct attention to the curves in Plate VII, which convey a 
great deal of valuable information, which can here be only slightly touched upon. 
They are reduced to one-sixth of the size of the original projections, in which one 
degree occupied two-thirds of an inch vertically, and one day occupied one-tenth 
of an inch horizontally. The corrected temperatures are those which have been 
projected. 
The curves extend over five years; and are placed in the order of depths 
(vertically) to which they belong: the uppermost undulating curves shewing the 
variations at the three stations 3 French feet below the surface, the lowest set 
shewing the variations 24 French feet below the surface. 
The most obvious results are the following :— 
1. In the upper set of curves, though the irregularities are greatest, yet the 
three curves follow one another with singular fidelity throughout these irregulari- 
ties. The curves separate a little in summer, and regularly in the same direction 
every summer, shewing the influence of exposure, the Experimental Garden being 
most heated, then Craigleith, and lastly the Observatory, which is also the order 
of the elevations of the stations above the sea. It may also be added, that the 
diurnal change may possibly have some slight influence upon the Experimental 
Garden, where the observations were made fully two hours later than at the other 
stations. (See Section V.) 
2. As the focal irregularities diminish at increasing depths, the range dimi- 
nishes, and the times of maxima and minima are continually retarded. 
3. At increasing depths, the curves, which followed one another so closely 
and exactly amidst the irregularities of temperature near the surface, systemati- 
cally separate from one another, both owing to a variation in the range or degree of 
undulation of the curve, and owing to a greater or less degree of retardation in 
the maxima or minima of the different curves. 
4, The effect last described is least sensible in comparing the observations 
at the Observatory and Experimental Garden, but most sensible if either of these 
be compared with the Craigleith observations, for which last the range diminishes 
more slowly, so that, at 24 French feet, it is about double that of either of the 
others, and the retardation of the maxima and minima is much less. 
5. In the trap and loose sand, the range is diminished to one-tenth part in 
descending from 3 feet to 24 feet; but in the sandstone it is not quite diminished 
to a fifth part. The epoch of maximum temperature is retarded in the two former 
cases nearly five months, in the latter only three. 
From these statements it is easy to see that the influence of the conDUCTING 

