of the perpetnat Snows in the North. 293 



respects a measure for the strength of vegetation, which must de- 

 pend only on temperatures above zero. 



The height at which the perpetual snows are found imder dif- 

 ferent latitudes, and the distance from their limit to that of the 

 trees, have been recently the subject of a profound discussion in 

 the Prolegomena, de Distrlbulioiw Geograpluca Plantarum of M. 

 <!e Humboldt. We shall give in a future number of the Annales 

 de Chimie et de Phvsique an extract of that part of the work 

 v.hicli contains some new and ingenious considerations on the di- 

 rectiou oi isothermal l/res. At present we shall conhna ourselves 

 to such of M. de Humboldt's results as refer to the interesting 

 question which is the subject of the article. 



1. Between the tropics from 0^ to 10^ latitude in the Cor- 

 dilleras of the new world, the limit of the snow = 4795 metres 

 (2460 fathoms). The mean temperature of the air at this height 

 is not zero, as Bouguer and after him all have admitted, but ra- 

 ther + 1*5° centigrade. 



2. Between the parallels of 19' and 21" of north latitude, 

 Mexico, at the entrance of the torrid zone, we find perpetual 

 snow at 4580 metres (2350 fathoms). 



3. Under the temperate zone, on Caucasus lat. 42° — 43°, the 

 height according to Messrs. Euglehardt and Parrot is 3216 me- 

 tres (1650 fathoms). 



4. In the Pyrenees (lac. 42=1—43"), M. Ramond finds the 

 permanent snows 2729 metres (1400 fathoms*). At this height 

 the mean temperature of the year = — • 3"5". 



5. The 



* We might have heen astonished to find ii difference of 487 metres (250 

 fathoms) between the heij^ht of tlie snow on Caucasus iind the Pyrenees, 

 almost under the same parallel, if the discussion to xvhicii M. dc Buch de- 

 votes his memoir had not shown how great is the influence of local circuni- 

 staiices. The following passage, taken from a letter, now very old, of ^I. 

 Jlamond, will have the ilouhle advantage of affording a considcrahle degree 

 of certainty as to the determination of altitude which the text contains for 

 the Pyrenees, and of making known the peculiar causes which in this chain 

 render the observation dii!icult. 



" I have remarked that the inferior snow limit descends lower the hiy:!ier 

 the mountaitis rise, and tliat ihemountaiiis which form the skirts of a chain 

 are deprived of snow at a lieiglit atwliich rhe centre mountains are covered. 



" Thus the Peak du Midi, alrliough 2935 metres (1506 fathoms) high, has 

 a permanent snow, if it is not m a rav ine exposed to the north-west, and 

 •till it disappears in warm summers. iS'eouville and Peak Long have perma- 

 nent snows at 2820 metres (]'150 fatlioms) only, and these snows contain 

 filacieres which never ap|>ear above them. But in order to have the charac- 

 teristic limit of the Pyrenees we must go to the centre of the chain, i. e. to 

 the environs of Mont Fcrdu, Marboii, and VigueMuk. I remark as follows 

 ;:i places favourable to those kinds of observations : — there are ca'rpfts of 

 |iermanent snow above the I^ort dc Fmei/e, Col de Nivelle, the lake of Mont 

 I'crdii, and above the peaks oiVort dcGarvurnia, Ike. The mean estimate 



T 3 ^ of 



