804 



ANNUAL EEPOET SMITHSONIAN INSTITUTION, 1917. 



The correlation of the bracketed stages is uncertain, but the se- 

 quence of events was probably something like this: Denmark was 

 ice free before Scania, so that the older Dryas period of Denmark 

 may be unrepresented in Sweden, or it may be only slightly earlier 

 than the arctic plant beds of Schonen (57), whose position is doubt- 

 ful. The Baltic ice lake evidently corresponds to elevation in Den- 

 mark, which shut out the sea. As it was a period of rapid ice melt- 

 ing, and therefore relatively warm, it may correspond to the Allerod 

 oscillation, in which case the fall of temperature and readvance or 

 still stand of the ice marked by the Scanian end moraines must be 

 correlated with the younger Dryas period, and the subsidence marked 

 by the Told/a Sea in the Baltic is marked in Denmark not by the 

 younger Yoldia clay, but by the Zirphaca sands. This correlation 

 is rendered more probable when we remember that the late glacial 

 subsidence reached its maximum in Denmark, the more peripheral 

 area, earlier than in Scania, the more central area. To represent the 



Denmark 



Primary Curve 



Secondary Curve 



Fig. 1. 



changes diagrammatically, we have to superpose two curves of 

 changes of level — a primary, which in Denmark was alwa5^s a phase 

 earlier than Scania, and a secondary, which was the same in both 

 districts (fig. 1). 



In Scandinavian late glacial, the depression of the Yoldia Sea 

 and its equivalents and the succeeding elevation appear to have 

 progressed inward to a central zone over the east of Sweden ; here, 

 where the ice was thickest and most permanent, the consequent de- 

 pression was the greatest, reaching more than 250 meters, and the 

 land took the longest time to recover from the effects of the load. 



