294 On the Limits of the perpetual Snows in the North. 



5. The mean of the observations recently published by M. 

 WaJilenbeig i^ives for the snow region in the Alps (lat. 45°-J 

 — 40)°^) 2o70 metres (1370 futhunis). At this height the mean 

 annual tt!n])erature= — 4'', the mean of winter is — 10", that of 

 summer + G°. 



6. The mean temperature of the year, at the height at which 

 M. de Buch has found the limit of the snows under the 6Sth de- 

 gree of latitude, is — 6^ : that of winter = — 20'5°, that of summer 

 + 9-5°. 



From the parallel of Popocatepec in Mexico, to that of Etna, 

 the inferior limit of the snow has not yet been determined by 

 direct measurements. According to the researches of M. de 

 Humboldt, this limit, by the latitude of 28- 17'^, which is that of 

 the Peak ofTeneriffe, must be 3S00 metres (1950 fathoms), but 

 the heiijht of this ridge is only 37 1 1 (1904 fathoms) ; so that if it 

 is clear of snow during sunnner, it is less on accoimt of being 

 heated (as was thought) by the volcanic fire, than on account of 

 its elevation, which is not considerable enough. 



In order to show how much this department of general physics 

 has gained latterly by the researches of travellers, we shall sub- 

 join the heights which Kirwan and all subsequent authors assign 

 to the limits of snow, and by their side the results which have 

 been furnished by direct observations : 



J. .•, , Heig/it of the Snow Limit Height of the Snow Limit 



from actzuil observation. according to Kii-ivun. 



metres. jiieMes. 



C^ .. 4795 .. 4783 



20"' .. 4580 .. 4186 



45° .. 2729 .. 2343 



61° .. 1690 



62^ .. 1582 .. 943 



eS'' .. .. 766 



67** .. 1169 



67i= .. 1072 .. 619 



70'' .. 1060 



of these elevations is about 2632 metres, ("or 1350 fatlioms), and all tliincs 

 considererl, I do not think it oii£;ht to l-e fixed lower : jjcrliaps it luay evci) 

 be carried tiie length of 2729 metres (1400 fatlioins), for below tliis tbe 

 carpets of snow are not continued, and their being mnintained depends on 

 local circunnstances, of ravines, or exposure, &c. Besides, notiiing is more 

 difficult than to establish this fart in the Pyrenees, because the whole cen- 

 tre of the chain is composed of mountains vertirally broken : there are 

 ahnoit no moderate slopes ; tlie snows roll down and form heaps which re- 

 sist the warmth of summer by their thickutss, whilsc the heights, which 

 ought to be covered with ihem, are bare." 



LIX. On 



