86 ICE RECESSION IN NEW ENGLAND 



•lays can be very high, causing the glaciers to melt considerably 

 and the rivers to swell strongly (see also Hann, 191 1). 



The nourishment of the ice sheet, which, as Hobbs (191 1, p. 

 287; 1911a, 191 5) points out, largely occurs in such a way that 

 the snow, fallen during a relative calm, is by violent anticyclonic 

 surface air currents swept from all central portions of the ice 

 sheet and deposited near and about the margins of the ice 

 shield, may perhaps have become greater in western Sweden, 

 but that does not explain the rapid recession in the Baltic region. 

 Nor can the discharge of bergs from the ice front explain the 

 different rate of retreat. In the present Baltic area, as well as 

 in the surrounding lowlands, which were submerged below the 

 Baltic of that time, calving doubtless played an important role, 

 particularly where the waters were deep and wide and the lifting 

 power and wave action was great. But since the supply of ice 

 in these regions was greater than in the supramarine parts of 

 eastern Sweden, the recession was almost as fast in these 

 latter. 



The important part played by sunshine as compared with that 

 of rain is evident, for the Swedish west coast, in late glacial time, 

 had doubtless much more rain than had the Baltic district. 

 Rains in Arctic regions are also usually cold and sparse. 



Indication of the great importance of temperature seems also 

 to be found in the thicknesses of the varves, which varied greatly 

 from year to year. If the recession had been mainly determined 

 by conditions of precipitation, it is to be expected that the 

 recession and the clay sedimentation should have been prac- 

 tically the same from year to year and should have undergone 

 only slow variation. 



Variations of temperature seem also usually to be the causes 

 of the long and marked fluctuations in the rate of recession, which 

 themselves, to be sure, can be readily explained by changes in 

 precipitation, since in Sweden and Finland the thicknesses of 

 the varves are found as a rule to stand in good relation to the 

 rate of retreat and the thin varves deposited during halts indi- 

 cate inconsiderable melting (cf. p. 84). 



