386 SIGURDUR THORARINSSON 



especially those of Betula, were much corroded. The count of pollen in three 

 samples is shown in Table 1 . It may be added that one grass pollen is probably 

 Elymus arenarhis and the Filicinae spores are mainly those of Dryopteris 

 Linneana. 



On the whole the pollen spectrum gives the same picture as the macro- 

 scopic plant remnants. The most characteristic feature is the dominance of 

 A Inns. Possibly the Betiila pollen is somewhat underrepresented, compared 

 with Alnus, as it seems to be more sensitive to corrosion. 



Further palynological studies of the sediments are highly desirable, both 

 in order to identify more species and to find out possible changes in the 

 flora during the period represented by the Svinafell layers. 



THE AGE OF THE SVINAFELL LAYERS 



Paleomagnetic dating, based on the repeated reversals of the remanent mag- 

 netism of the basaltic rocks, was introduced in Iceland by Hospers (1953) and 

 has since been applied in this country mainly by Einarsson (1957 a. b) and 

 Sigurgeirsson (1957). It has proved very useful for the establishment of a 

 chronology for the Tertiary and Pleistocene rocks in Iceland. This method 

 has also been of much help in the determination of the age of the Svinafell 

 layers. The author measured the magnetization of the rock of Svinafellsfjall 

 at many places in the field, and for further checking, a lot of rock samples 

 were measured in the laboratory by Professor Th. Sigurgeirsson. 



That the Svinafell layers are interglacial is, in the author's opinion, already 

 evident from the fact that they are definitely younger than the basalt-globe 

 breccia of Snidabrekka, which is a very typical subglacial breccia. The 

 question is then, to which Interglacial do the layers belong? Hospers (1953, 

 pp. 472-473) and Einarsson (1957b, pp. 223-224) agree with Pjeturss (1903), 

 Bardarsson (1929) and Askelsson (1938, 1960) in regarding the Biilandshofdi- 

 Brimlarhofdi (Stodin) fossiliferous sediments (cf. Fig. 1) as interglacial, and 

 state that they are covered by series of basalt layers of reverse polarity. 

 According to Einarsson, moraine-like sediments are found between some of 

 these basalt beds. The Svinafell layers, on the other hand, rest on normally 

 magnetized basalts that have been greatly eroded, presumably mainly by ice, 

 before the deposition of the plant-bearing sediments. Thus, they cannot 

 belong to the same Interglacial as the Biilandshofdi-Brimlarhofdi layers, 

 which in all probability belong to the Giinz-Mindel Interglacial (cf. also 

 Askelsson, 1938, 1960). The Svinafell layers therfore cannot be placed lower 

 than the Mindel-Riss Interglacial. As stated above, the rocks of Svinafellsfjall 

 resting on the Svinafell layers represent two thick series of subglacially formed 

 rocks, separated by subaerial lavas. All these rocks are normally magnetized. 

 In the author's opinion there is hardly any doubt that these rocks represent two 

 Glacials, and the Svinafell layers therefore cannot be placed as high as the 

 Riss-Wurm Interglacial. Consequently they must be placed in the Mindel-Riss 



