212 PROFESSOR FORBES ON THE TEMPERATURE OF THE EARTH. 
imperfectly reduced to theory, that we cannot satisfactorily compare it with ex- 
periment ;* but one law of great simplicity has long been known from theory 
to be approximately true, namely, that the retardation of epochs is uniformly 
greater as the depth increases. This is also easily verified graphically. By tak- 
ing the depths in a vertical direction, and setting off the day of greatest or least 
temperature horizontally, a series of points is obtained through which a straight 
line should pass. I have not engraved these projections for each year, but that 
for the mean of the whole will be seen in the lower part of Plate X., where the 
interpolating lines in general pass so nearly through the dots that they cannot be 
distinguished. From these projections the mean rate of propagation downwards 
is easily determined, and affords a palpable illustration of the conducting powers 
of the soil. 
TABLE XIV. SHEWING THE NUMBER OF DAYS REQUIRED BY THE IMPRESSION 
oF HEAT TO PASS THROUGH ONE FooT oF SOIL. 
MAXIMA. MINIMA, 

Observa- |Experimen- 
Observa- |Experimen- 
tal Garden. 
Craigleith. tory. tal Garden. 
Craigleith. 


Days. Days. Days. 
1837 : 3 4:9 nate : 34: 
1838 : : 3°6 6°5 ; 3°6 
1839 : : 4:6 6:0 - 3-6 
1840 35 | 61 3-05 
1841 . ; 3°0 6-4 : 3°6 




Means : 2 3°92 6:25 i 3:46 





It must be added, that in the several years the law of wniform progression is 
by no means accurate, although, in the mean of five, the accidents are nearly 
compensated. And here, again, we find the good conductor, the sandstone, gives 
by far the most regular and consistent results. 
E.—On the Form of the Annual Curves. 
With a view to approximate more nearly to the form of the annual curves of 
temperature at different depths, I have had the mean temperature for each week 
of the year taken by the mean of five years, which has the effect of disposing of 
the more irregular fluctuations, as may be seen in Plate VIII., the curves in which 
are taken from the following Table : 
* See the Appendix. 
