170 FORMATIONS AND GUILDS [Part II 



supplied by Kihlman, very hazy ideas prevailed in geographical botany 

 as to the causes of the absence of trees within the arctic zone. At one 

 time it was ascribed to the cold, then to the shortness of the vegetative 

 season, then to a combination of both these factors, although no character 

 founded on the physiology of trees could in any way support such an 

 assumption. That severe and persistent winter cold is not incompatible 

 with the growth of trees follows from the fact that the lowest degree of 

 cold known anywhere occurs in the Siberian forest district l . 



The significance of the wind in relation to tree-growth was already recognized 

 by Middendorff, though not on physiological grounds, as appears from the 

 following extract from his work on Siberia 2 : 'I am ready to maintain that in 

 the extreme north a favourably formed shelter against the wind is of much 

 greater importance than the geographical latitude or altitude above sea-level. 

 A shelter a few fathoms high favours tree-growth there much more than fifty 

 thousand or a hundred thousand fathoms less of northern latitude.' 



It is well known that the north polar tree-limit does not form a 



sharp line of demarcation 

 between forest and treeless 

 tundra. Tree-growth be- 

 comes gradually reduced 

 before it entirely disap- 

 pears, as was clearly pointed 

 out by Middendorff and 

 especially by Kihlman. 



Fig. 91. Jnniperus communis. Tabular growth. At the Middendorff gives a rough 



limits of tree-growth. After Kihlman. . ° 



description of the pheno- 

 menon without entering into its causes : — ■ 



' If we follow the tree-limit over large tracts of country and observe all the 

 different species of trees appearing on them round the North Pole, we see that 

 they all in like manner become stunted and degenerate into gnarled growths 

 (Fig. 90) : both broad- leaved species and conifers eventually become dwarfed 

 into veterans, two feet or even one foot in height 3 .' 



The deformities that tree-growth experiences near its polar limits 

 originate, as Kihlman shows, from desiccation in winter, the increase of 

 which in the northerly direction finally checks all tree-growth : — 



'On observing the development of the juniper, as it appears in the higher 

 forest region or in the interior of the tundra (Fig. 91), it will be found that the 

 tip of the main axis regularly dies as soon as it has attained a certain, somewhat 

 variable, height above the ground. The side branches however continue to grow 

 obliquely upwards, or almost horizontally, until their tips also die at the fixed 



1 See p. 40. 2 Middendorff, op. cit., p. 683. 3 Id. p. 675. 



