276 O. GJ/EREVOLL 



Nannfeldt {he. cit.) also includes the endemic, viviparous Poa stricta in 

 the same complex. This race has an abundant and very concentrated occur- 

 rence in the Dovre Mts. (Fig. 18). 



I have tried to show that the greater part of the problematic plant species 

 of the south Scandinavian alpine flora is confined to the mountains some 

 distance from the coast. Many of the plants are fairly evenly distributed 

 within this southern area, but there are also striking irregularities. Rhododen- 

 dron lapponicum and Braya linearis are known only from the northern part of 

 Jotunheimen, Stellaria crassipes and Carex bicolor only from Dovre. Poa 

 stricta has a small area in Trollheimen, but its main distribution is in Dovre, 

 and it is not found at all in Jotunheimen. Ranunculus nivalis, Carex arctogena, 

 and Draba nivalis do not occur in Trollheimen. 



Saxifraga hieraciifolia with a single coastal locality in the Romsdal Mts. 

 has its main distribution in Sunndalen, Jotunheimen, and the southern part of 

 Dovre. It is not known from the rich areas of central Dovre and Trollheimen. 



How is it then possible to explain this pecuUar distribution and the dis- 

 junctions within the southern area? 



It has been assumed commonly that the species must have survived in the 

 coastal districts of More. When the ice began to retreat, the plants followed 

 the retreating ice to the mountains of Sunndalen, Trollheimen, Dovre, and 

 Jotunheimen where they are met with today. The humid, oceanic cUmate is 

 unfavorable to alpine species, due for instance to peat formation. Therefore, 

 these species have suffered competition and disappeared. 



I myself have not been content with the assumption of only a coastal 

 survival area. It is not possible to explain the total absence in the coastal 

 mountains of all the species mentioned by referring to peat formation and 

 soil conditions. It is a fact that there are high mountains close to the coast 

 where neither peat formationn or climate may represent any danger to these 

 species. I have had the opportunity to make some investigations in the Roms- 

 dal Mts., and there are plenty of suitable localities, e.g. for Artemisia norvegica. 



There are also coastal mountains consisting of calcareous rocks and schist, 

 where the calciphilous Dryas octopetala grows abundantly. This is usually the 

 best indicator of "rich areas". In this connection Mt. Talstadhesten in outer 

 Romsdal is very interesting. It is a limestone area and harbours a flora 

 comparatively rich in species, but none of those mentioned above are present 

 even though the geological conditions ought to ht suitable. Large areas are 

 not at all subject to critical humus production. Most likely this mountain 

 has been a nunatak, and, according to Nordhagen (1952), Euphrasia lapponica 

 is represented here by a variety known only from this locality and the moun- 

 tains of Hardangervidda. Furthermore, according to Porsild (1958) the 

 Dryas growing very abundantly here is not D. octopetala s.str., but D. 

 Babingtoniana, otherwise known only from the British Isles. 



I should like also to call attention to the fact that a snow-bed species like 



