65 J. A. NANNFELDT 



not to be denied. Is there any better explanation for Dryopteris fragrans in 

 Utsjoki (northernmost Finland) or for Oxytropis deflexa in Kautokeino 

 (northernmost Norway) or for the short visit of Arctostaphylos alpina to a 

 spot in west Jutland? 



The "West Arctic Element", however, cannot be explained in that way. 

 The number of species is too large, and they do not occur haphazardly but 

 are restricted to special areas which are inhabited also by other species 

 possessing restricted and disjunct part-areas. A theory explaining the occur- 

 rence of the "West Arctic" species is not acceptable, if it does not at the same 

 time explain also the occurrence of these other species. In the mind of 

 numerous phytogeographers, including myself, the only possible explanation 

 is the existence during the Last Glaciation of ice-free refugia west, and perhaps 

 even north, of the ice-sheet, not too far from the areas where these plants 

 grow nowadays. How they survived the previous glaciations is impossible to 

 know, but I feel sure that they, or at least most of them, did survive in north- 

 west Europe. I feel sure that the Scandinavian mountain plants have a long 

 history with us and that few, if any, species have reached us from the south as 

 late as during or after the Wurm Glaciation. Possible exceptions are such 

 species as Campanula harhafa, Gentiana purpurea and Ranunculus platani- 

 folius. These are far from High-Alpine and are restricted to the south part of 

 the Scandes, except for a most isolated locality in Finnmark for the Ranunculus 

 (chance dispersal ? ?). Several species have certainly perished from the 

 severe vicissitudes during the glaciations, and a new glaciation would certainly 

 impoverish our flora still more. This is the only explanation of the absence 

 from the Scandes of such species as Alopecurus alpinus, Cluunaenerium 

 latifolium and Lomatogonium rotatum. 



Taxonomists and phytogeographers often plead for a broad species 

 concept, lest the general survey and the understanding of the natural con- 

 nections should be lost. It is true that a species concept such as used by 

 numerous Russian authors who give specific rank to almost every geographic- 

 ally isolated population, leads to a loss of the understanding of the phylo- 

 genetical connections. The treatment of all agamospecies as normal "full" 

 species may have the same effect, especially if the agamospecies of, for 

 example, the Ranunculus auricomus-complex were treated on the same level 

 as the species within the larger, sexual part of the genus. But the scope of the 

 species is in itself of little importance, if only the lower recognizable units are 

 not neglected, for these smaller units give often important, in many cases 

 perhaps the most important, clues to the migratory history of a species. I shall 

 here give some few examples from the North Atlantic area, especially relating 

 to the flora of the Scandes. My first examples aim at showing the closer 

 floristic connections of the Scandes with Iceland and Scotland than with 

 the Alps and other central and south European mountains. 



