312 THORODDSEN 



lude of 600 800 metres, there are usually gravelly tracts with scat- 

 tered individuals of rocky-flat plants, such as Silene acaulis, S. ma- 

 ritima, Cerastinm alpinum, Luzula arcuata, Polygonum viviparum, 

 Armeria maritima, Ranunculus glacialis, Saxifraga nivalis, S. oppo- 

 sitifolia and others. In small damp hollows where the snow 7 persists 

 for a long time there is often a characteristic dense growth of Salix 

 herbacea, which almost entirely conceals the moss-covering of the 

 ground, together with Sibbaldia procumbens, Gnaphalium supinum, 

 Oxyria digyna and Polygonum viviparum; in some places these are 

 associated w T ith several other plants. In other places there are small 

 patches of Grimmia-heaths with scattered specimens of Pedicularis 

 flammea or Cassiope hypnoides. Here and there streams and bogs 

 occur w r ith Carex rigida, C. lagopina, C. rostrata, C. incurva and others, 

 but most often with Eriophorum angustifolium and E. Scheuchzeri. 



As has been seen from the preceding notes on the vegetation of 

 the plateau it is not easy to determine the altitudes or upper limits 

 of the different plant-regions. Of the Scandinavian upper zones, the 

 region of conifers is entirely absent, but, on the other hand, w r e 

 may be justified in speaking of a birch-region, of an osier-willow 

 region, and perhaps a lichen- or moss-region, but these regions pass 

 into one another in many ways, and overlap. During the period 

 after the Ice Age (the Purpura-lapillus Period) when it was warmer 

 than it is now, the birchigrew everywhere in the lowlands even 

 on the northernmost headlands, but it had already retired from the 

 latter at the time the first settlers came to the island, and since 

 then, as we have already seen, its distribution has been considerably 

 limited owing to the interference of man and sheep. The present 

 Polar limit of Betula odorata in Iceland has not been fully investi- 



*/ 



gated, but judging from what I saw r on my journeys it appears, on 

 the east coast, to extend across Vopnafjordur to the w r est coast of 

 Melrakkasljetta, across Axarfjordur and Skjalfandi to Eyjafjordur 

 near the mouth of the valley of Fnjoskadalur. The stretch of land 

 between Eyjafjordur and Hunafloi is now devoid of birch coppices, 

 although these occurred there in olden times; how far out they ex- 

 tended at that time upon the peninsulas between the fjords is not 

 known. From Hunafloi the northern limit of the birch extends from 

 Steingrimsfjordur to Isafjardardjup. This is, however, only quite a 

 provisional limit; the subject requires to be investigated more closely. 

 As regards the upper limit of the birch, it differs considerably 

 in different parts of the island; it extends highest in Thingeyjarsysla 



