ALTITUDES OF SNOW-FIELDS AND GLACIERS. 135 



of all the causes tending to its accumulation over those 

 which tend to its waste or fusion, is indeed a very com- 

 plicated fact, and cannot be taken as the simple expression 

 of any one of the elements of the climate. The snow 

 line is far from having invariably a mean temperature of 

 32, as was at one time supposed. Under the equator it 

 is about 35 ; in the Alps and Pyrenees about 25 ; and 

 in latitude 680 in Norway it is (according to Van Buch), 

 only 21. Yet there are regions, both in the extremity of 

 Siberia and in Arctic America, of which the mean tem- 

 perature is below zero of Fahrenheit (as, e.g., Melville 

 Island). And it is quite established, on the concurrent 

 authority of those best acquainted with those regions, that 

 nowhere in the northern hemisphere does the snow-Line attain 

 the low level of the sea. The explanation is to be sought 

 principally in the intensity of the summer heat during 

 the period of perpetual day, which effectually thaws the 

 soil, though only to a trifling depth, and raises upon its 

 surface a certain amount of brief vegetation, suitable for 

 the support of Arctic animals. 



' Another cause affecting exceedingly the level of the 

 snow-line is the amount of snow which falls. The interior 

 of continents being far drier than the coasts, the snow to 

 be melted is a comparatively slight covering. The snow- 

 line on the north side of the Himalaya is at least 3000 

 feet higher than towards the burning plains of Hindostan. 

 This is chiefly due to the excessive dryness of the climate 

 of Thibet. In like manner five times less rain falls on the 

 coast of the Baltic than at Bergen. All this confirms the 

 excellent generalisation of Von Buch that it is the tempera- 

 ture of the summer months which determines the plain of per- 

 petual snow. It is thus easy to understand why the mean 

 temperature of the snow-line diminishes towards the pole, 

 because for a given mean temperature of the whole year, 

 the summer is far hotter in proportion. Also, places at 

 which the temperature of the summer is low are those 

 which have a moderated or coast climate ; but there also 

 the fall of rain and snow is most abundant, whilst in exces- 



