a Traveller on the Glacier of Buet. 59 



taula, therefore, applied to the mean temperature at the 

 level of the fea, in the parallel of 46' (the latitude of Buet), 

 as enablifhed in Dr. Kirwan's Work on the Temperature 

 of the Globe, viz. 56-4° nf Fahrenheit, or io c ,8 of Reau- 

 mur, will give for 1453 to '^ cs I 4'53 degrees to be deducted ; 

 which makes the mean annual temperature of the lower 

 boundary of the mow, in thai parallel, to be 3*63 degrees 

 below zero*. 



It is not furprifing, therefore, that thi? mountain fhoulcl 

 be crowned with a glacier, fined the fnow which falls there 

 during the cold feafon, never entirely melts in the fummer. 

 The water produced by the partial melting of the furface of 

 the fnow filters through the porous fnow beneath it, and 

 freezing in its interfaces gradually converts it into ice. In 

 this manner is formed an accumulation, the thicknefs of 

 which M. De Luc endeavoured to eftimate from the follow- 

 ing obfervation : 



" We judged," faid he, " bv the pofition of thefe fmall 

 rocks, about 200 feet lower than the higheft part of the 

 ice, that they formed a part of the real fummit of the 

 mountain. The whole mafs above them was nothing but 

 ice, in the form of a cone, cut through its axis, 200 feet in 

 height, with a very extenfive bale, and retting on the im- 



* I had occafion not long ago to difcufs this formula with a phi- 

 lo r o,>her, who obferred to me, that, according to its nature, it was 

 impolfible it could be corredt ; becaufe the detifuy of the air, an ele- 

 ment on which the prefcrvation of heat in the different ftrata of the 

 atmofphcre effentially depends, decreafes in geometrical progreffion ; 

 while the heights in toifes, which reprefent the temperatures, proceed in 

 arithmetical progreffion. I admitted the juftnefs of the oMervation, 

 fpeaking mathematically ; but, in a phyfical point of view, as the for- 

 mula is compofed of co- efficients, fome of them unknown or inappre- 

 ciable, which gives to the tempcratnrean arithmetical progreffion, decreafing 

 from the bottom upwards, it is no lefs true, th it rhis formula, however 

 deceptive it may be, represents the mean relults of obfervations fuffi- 

 ciently well to be employed with convenience, when an approximative 

 q lantity only is necelfary : and this was exactly the cafe. The queftion 

 was the mean temperature of mount Saint Bernard, the philofopher 

 was Bonaparte, and the difcuffion took place at table, and even in the 

 apartment of tiie celebrated man whofe theory and calculations I \va> 

 endeavouring to defend. 



I 2 roenfe 



