382 THE NORTH POLAR BASIN. 



but this may be a temporary or local variation caused by exceptionally 

 severe winters. The Hunt of forest growth does not coincide with the 

 isotheruKS of mean annual temperature, nor with the mean temperature 

 for January nearly so closely as it does with the mean temperature for 

 July. It may be said to approximate very nearly to the Jnly isotherm 

 of 53° F. We may therefore assume that a 6-foot blanket of snow pre- 

 vents the winter frosts from killing the trees so long- as they can be 

 revivified by a couple of mouths of summer heat above 50° F. 



The limit of forest growth is thus directly determined by geograph- 

 ical causes. In Alaska and in the Mackenzie Basin it extends about 

 300 miles above the Arctic Circle, but in eastern Canada the depres- 

 sion of Hudson Bay acts as a vast ice-house, and the forest line falls 

 500 miles below the Arctic Circle, whilst on the east coast of Labrador 

 the Arctic current from Baffins Bay sends it down nearly as far again. 

 On the other side of the Atlantic the limit of forest growth begins on 

 the iNorwegian coast on the Arctic Circle, gradually rises until it 

 reaches 200 miles farther north in Lapland, is depressed again by the 

 ice-house of the White Sea, but has recovered its position in the valley 

 of the Pechora, which is rather more than maintained until a second 

 vast ice-house, the Sea of Okotsk, combined with Arctic currents, 

 repeats the depression of Labrador in Chuski Land and Kamchatka. 



There are no trees on Novaya-Zemlia. Two or three species of wil- 

 low grow there, but they are dwarfs, seldom attaining a height of 3 

 inches. Novaya-Zemlia enjoys a comi)aratively mild winter, the mean 

 temperature of January, thanks to the influence of the Gulf Stream, 

 beiug 15° r. above zero in the south and only 5° F. below zero in the 

 north. The absence of trees is due to the cold summers, the mean tem- 

 perature of July not reaching higher than 45° F. in the south, whilst 

 in the north it only reaches 38° F. 



The Lidians of Canada have discovered that when they want lo tiiid 

 water in wanter it is easiest reached under thick snow, the thinnest ice 

 on the river or lake being found under the thickest blanket of snow. 

 On the same xn-inciple the tree roots defy the severe winters juotected 

 by their snow shields; but they must have a certain temperature (above 

 50° F.) to hold their own m summer. 



The influence of the snow blanket is very marked in determining the 

 depths to which the frost penetrates beneath it. Thus we And that a 

 Norwegian writer, alluding to latitude ()L'°, remarks '-that the ground 

 is frozen Irom 1 to 2i feet m winter, but this depends upon how soon 

 the snow falls. Higher up the mountains the ground is scarcely frozen 

 at all, owing to the snow falling sooner, and in fact if the snow falls 

 very early lower down it is scarcely frozen to any depth." Similar 

 facts have been recorded from Canada in latitude 53°. " On this 

 prairie land, when there is a good fall of snow when the winter sets in, 

 the frost does not penetrate so deep as when there is no snow till late." 

 Another writer a little farther south, in latitude 51°, says: " 1 am safe 



