THE ICE-FREE REFUGIA OF NORTHWESTERN SCANDINAVIA 329 



strating that the lower hmit of such detritus — a limit not necessarily repre- 

 senting the ice surface of Wiirm time but rather that of the maximum Pleisto- 

 cene glaciation (loc. cit., p. 92) — is only hundreds of meters above present 

 sea level in the coastal areas, whereas it reaches values of 1700 to 1900 m in 

 the southwestern part of the investigated region. This hypothesis of Dahl's 

 does not agree with the distribution pattern of mountain top detritus in 

 Sweden. For instance, in the southern part of the Scandes at about the same 

 latitude and at altitudes between 900 to 1200 m, many of the peaks have well 

 developed boulder fields of autochtonous character. From the province of 

 Dalarna, Lundqvist (1951, p. 81, etc.) mentions Slugufjallet, Nipfjallet, 

 Stadjan, Drevfjallet, Fulufjallet and Getsjohoa. An early reference to these 

 boulder fields on mountain tops was made by Samuelsson (1914). Glacial and 

 glaciofluvial features of diiferent kinds demonstrate also that the peaks must 

 have been covered by the last inland ice sheet at a very late stage. 



From more northern parts of the Swedish mountains, Rapp and Rudberg 

 (1960) have assembled some data on the lower Hmit of the boulder field zone: 

 southern Jamtland, 1 500-1600 m, Yasterbotten, 1 100-1200 m, the amphibolite 

 mountains at Abisko, 1000-1200 m. These figures refer to areas where 

 covering by the ice sheet is indisputable. In addition, in Finnish Lapland 

 low-lying mountain top detritus has been observed, for instance, on Levitun- 

 turi (about 20 km. north of Kittila) 531 m above sea level (V. Okko, pers. 

 comm. 1961). 



The formation of mountain top detritus certainly is a complicated process 

 with many variables, such as the properties of the bedrock, its condition before 

 glacierization, the velocity and eroding capacity of the ice sheet in different 

 situations, the conditions of deglaciation, postglacial chmate, etc. I think 

 that something similar can be said about the more fine-grained weathering 

 material, including the clay minerals from mountain tops which have also 

 been used by Dahl as a proof of nunatak situations. Here as elsewhere, 

 however, Dahl stimulated the debate in an effective way. 



Initially it was stated that nunataks very probably existed during Wiirm 

 time. Finally it should be added that plants and animals, if any survived, 

 must have lived under extremely harsh conditions. 



2. Foreland refugia in western Norway. The presence of glacial striae and 

 erratics in many places in areas postulated as foreland refugia has been taken 

 as evidence of glacierization by the opponents of this concept. The proponents 

 of the refugium hypothesis reply, however, that such evidence may remain 

 from a previous glacial stage, but this is energetically denied by the opponents. 



The successive destruction of striated surfaces can be followed everywhere 

 in Sweden and plotted against a quite detailed time scale. In many areas it is 

 absolutely impossible to find any striae on bedrock surfaces which have been 

 exposed only a few thousand years (e.g. since the emergence of the land 



