210 



PROFESSOR FORBES ON THE TEMPERATURE OF THE EARTH. 



On these results it may be remarked, that A, which is the logarithm of the 

 superficial range, is necessarily variable according to the season, and that it ap- 

 pears, singularly enough, to have been constantly on the decrease throughout the 

 period of these experiments. This gives a great probability that the mean of 

 these will be very nearly an average result for this climate. The depth at which 

 the annual variation disappears is also evidently dependent, in part, on the qua- 

 lity of the season. B is the only proper constant, depending solely upon the spe- 

 cific heat and conductivity of the soil ; and the mean results of Table X. are evi- 

 dently near approximations to the truth. 



These computations have been made on the supposition that the logarithmic 

 law of the diminution of the range is correct, and that the deviations from it are 

 due to accidental errors. These deviations appear, however, to be too systematic 

 to admit exactly of this conclusion. The observations at Craigleith coincide most 

 nearly with theory ; those at the Observatory much less so, although there is every 

 reason to believe that the observations there were in every respect the most un- 

 exceptionable of the three. At the Observatory, the observations at great depths 

 indicate a less rapid contraction of the range than do those at the surface, as an 

 inspection of the curves in Plates IX. and X., and the points through which they 

 have been drawn, sufficiently proves. 



To illustrate this difference, I had the constants A and B separately computed 

 from all the possible combinations by pairs of the observations of 1837-38, with 

 the following results. 



Table XII. 



The curves of Plate VII. plainly shew that the periods of maximum and 

 minimum temperature occur later and later as we descend. The epochs of maxi- 

 ma and minima were obtained graphically at the same time with the greatest and 

 least temperature, in the manner akeady described. The results are contained in 



the following Table : 



i 



I 



