on Subterranean Temperatttre. 279 



vities ; and, as the results have been similar to those which I am 

 about to examine, the inferences at which I shall arrive are 

 equally applicable to them. 



Such are the experiments whose merits we have to appreciate ; 

 and in doing this criiicism has nothing to neglect. As it is pro- 

 posed definitively to apply to the great the inferences deduced 

 from the small, it is obvious that the shghtest errors will have a 

 prodigious influence upon what is to be inferred regarding the 

 entire mass of the globe. Thus, for example, proceeding from 

 the approximate law which is deduced from the experiments hi- 

 therto published, one degree of Fahrenheit of error more for a 

 depth of 180 feet, in a given country, will raise to 1,600 feet (nearly 

 half a quarter of a league), the point at which it is to be presumed 

 that the temperature of boiling water exists under the place of 

 observation. These considerations will be a sufficient apology 

 for the details into which I shall sometimes be obliged to enter. 



By means of the precautions to which I have had recourse, I 

 trust that my own experiments may be regarded as sufficiently 

 accurate. Most of them were made in three coal mines in France, 

 very distant from one another, which I selected as presenting the 

 most favourable circumstances, and which are : 1st, the mine of 

 Littry, situated eight miles W. SW. of Bayeux, in the Depart- 

 ment of the Calvados, and of which the openings have an eleva- 

 tion of about 200 feet above the sea ; 2dly, the mine of Decise, 

 situated seven and a half miles to the north of the city of that 

 name, and of the banks of the Loire, in the Department of the 

 Nievre, and of which the elevation above the sea is about 490 

 feet ; 3dly, the mine of Carmeaux, situated in the Department of 

 the Tarn, eight miles to the north of Alby, and nearly 820 feet 

 above the sea. I shall revert to the local circumstances of these 

 mines as I proceed. At present it is sufficient to add, that my 

 experiments took place, in the first in August 1823, in the second 

 in September 1825, and in November 1822 and September 1825 

 in the third. In all, I made use of mercurial thermometers, which 

 I carefully provedand compared with one another, and which, in all 

 cases, where I shall not mention the contrary, were applied with 

 the ball naked. With the kind assistance of MM. Arrago and 

 Mathieu, I have been enabled to reduce all my results to the gra- 

 duation of the normal thermometer of the Observatory of Paris, 



