302 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1917. 



posited in a shallow f resli-water basin in a steep-sided valley at right 

 angles to the direction of ice motion; the climatic conditions "were 

 arcto-alpine part of the time, but must have been somewhat better 

 when Betula odorata, Pinus silvestris, and Picea excelsa grew. Its 

 interglacial age is based not only on the stratigraphical evidence, 

 which is very strong, but also on the similarity of its flora to that of 

 undoubted interglacial deposits in Denmark and north Germany. 

 Picea excelsa is a very characteristic interglacial fossil in Denmark. 

 This Hernosand deposit lies exactl}^ at present sea level. 



J. Geikie (53) quotes a letter from M. Tornebohm describing a 

 section in Wermland showing two bowlder clays superposed, the 

 lower darker in color and containing fewer big stones than the 

 upper. There is usually a sharp line of demarcation between them, 

 and in some places the lower till has been partly broken up and 

 denuded before the upper till was deposited, indicating an interval 

 when the ground was free from ice. 



A. G. Nathorst and II. Lundbohm have shown that earlier than 

 the ice sheet which moved from north-northeast to south-southwest 

 and formed the " lower diluvium" (middle glaciation) of north Ger- 

 many, there must have existed another ice sheet which glaciated 

 southern Sweden from east to west and accumulated a ground mo- 

 raine with blocks from the east and southeast. 



In southwest Norway, at Jaederen, south of Stavang'er, K. Bjor- 

 lykke (54) found a Cyprina clay passing under and disturbed by 

 bowlder clay and seen in borings to be underlain by other glacial 

 deposits ; he considers it to represent the cold part of an interglacial. 

 Farther east, between Stavanger and the Christlania Fjord, A. M. 

 Hansen (55) finds evidence of two main glacial periods, in addition 

 to the later readvances ; these he attributes to Geikie's Saxonian and 

 Mecklenburgian. 



Thus we see that there is considerable, though scattered, evi- 

 dence that at least once during the course of the glacial period Scan- 

 dinavia became practically ice free. At present it is not possible 

 as a rule to allocate the interglacial deposits to definite horizons, 

 but the Cypriiia clay of Jaederen is considered by Bjorlykke to be- 

 long to the older interglacial, on both stratigraphical and paleonto- 

 logical grounds, and this would bring it into good agreement with 

 the Cyprina clays of Denmark and north Germany. On the other 

 hand, the flora of the Hernosand Gyttja is more in accordance with 

 the upper interglacial of north Germany, and this is supported by 

 the relative freshness of the deposits. We may accordingly suppose 

 that there were at least three entirely distinct glaciations of Scan- 

 dinavia, separated by intervals when the ice melted well back into 

 the mountains, if it did not disappear altogether. During the earlier 

 of the intervals the coast of Norway lay lower than at present; 

 during the later both Norway and Sweden lay higher. From the 



