163 



ably smaller on the stony flats at Vassijaure than in the 

 lowlands, often only half as large, bnt comparatively broader. 

 A somewhat similar effect may be observed in other Arctic 

 regions, bnt it should be remembered that the development of 

 the leaves is, to a very great extent, dependent upon the nature 

 of the habitat, and that small, poorly developed leaves can as 

 well occur on Danish heaths as on those in Arctic regions; 

 or vice-versa, the leaves may be as well developed on Disko, 

 for example, as in Denmark. 



Towards the apex of the shoots the leaves are very close-set. 

 There is no other protection for the young leaf-rudiments than 

 this dense crowding of the leaves.^ 



As is well-known, the leaves expand long after the flowering 

 of the plant; in Denmark, in April or early in May, but in 

 the Arctic regions naturally much later still; thus as regards 

 Greenland the leaves do not expand until some time in July 

 (this statement is based on the material to hand ; 1 have in 

 vain searched in the literature upon the subject for information 

 regarding the leaf-expansion). As a result, the shoots have 

 only a very short time in which to grow, for during September, 

 without doubt, all growth is stopped. It will easily be under- 

 stood that, as mentioned above, the growth-increment can on 

 the whole be but small. 



In connection with the fact that the leaves are evergreen, 

 is their xerophylic nature ; their peculiar structure is described 

 by several authors, as Gibelli, Grcber, Warming, Borgesen (3), 

 Wagner (18), etc. The notes by Warming (20) and Borgesen (3) 

 on this subject are particularly interesting , because their 

 investigations were carried out upon Arctic material; otherwise, 



^ Engler (Die natürl. Planzenfam. Ill, 5, p. 127) says: — "Knospen- 

 schuppen lang bewimpert." But we cannot be justified in talking of 

 budscales propei-, although it is true that the leaves, when quite young, 

 are furnished \\ith long hairs, which are matted together very densely 

 in the space above the shoot-apex; this may easily be ascertained by 

 a median section through the apex of the shoot. 



1Г 



