PROFESSOR FORBES ON THE TEMPERATURE OF THE EARTH. 219 



(which is equivalent to the expression in the Theorie de la Chaleur, p. 499, Eq. (26) ), 

 gives the following numerical result : — 



Trap. Sand. Sandstone. 



14-124 16137 24-750 



but if referred to the French Metre instead of foot as unity (the centigrade de- 

 gree has been already employed), they become 



4-588 5-242 8-040 



W which are comparable with Poisson's result, 5-11655 for the Observatory of Paris. 

 M Now, the specific heat c having been found in the last section, we may eliminate 

 ■ it, and obtain the following numerical values of k, the conducting power of the 

 1^ strata, which it may be presumed has rarely been so accurately determined for 

 any kind of matter. 



Trap. Sand. Sandstone. 



k 11-120 8-260 29-884 



There is another constant b employed by Poisson, which involves the character 

 of the recipient surface of the ground as well as the interior conductivity, and 

 which is determinable from the retardation of epochs by equation (27) of page 499 

 of the Theorie de la Chaleur. 



where 6 and 6^ are the epochs of maximum and minimum temperature at any 

 given depth, reckoned from the 21st March in fractions of a year (= 1), the metre 

 being also the unit. Instead of taking observations at a single depth, we may 

 take the epoch for 24 French feet from the interpolating lines in Plate X., which 

 represent not merely the observations at that depth, but the result of their com- 

 bination with all the others. 



Trap. Sand. Sandstone. 



Voar. Year. Year. 



Maxima at 24 F. ft., July 8.= -515 July 1. =-496 May 6. = -234 



Minima, . . . Jan. 4.= l-008 Dec. 26. = -984 Nov.4. = -841 



Mean reckoned from 1st Jan., -761 -740 -591 



Beckoned from 21st Mar. = -J (6 + 0j) -545 -524 -375 



Substituting the values of a?=: 7-7961 metres (24 F. ft.), and of a before found, we 

 obtain 



Trap. Sand. Sandstone. 



b 0-4972 , 0-1007 0-0772 



M. PoissoN finds for b at the Paris Observatory, the value 1-057. If we examine 

 the circumstances which influence the value of 6, we shall admit that its determi- 

 nation in this manner is liable to so great en-ors as to render it almost worthless. 



