recession of the pack boundary northvfard. The rapid melting of the ice 

 is attested by several observers in past years and represents a probable 

 change in currents according to the following reasoning. The rapid melting 

 can only occur if the ice in question is winter ice, since polar ice melts 

 slowly due to its physical characteristics. If the quickly-melted ice is 

 winter ice, the edge of the polar pack must be far north of the Alaskan 

 coast. But by hypothesis, during the previous summer the pack boundary 

 was close to shore. Therefore, the movement of the polar ice away from 

 the coast occurred during the previous winter and must have been the re- 

 sult of a steady slow movement of the ice. A current is suggested, and 

 it is possible that during the winter the current from Bering Sea and the 

 anticyclonic Beaufort Sea whirl may have been stronger than usual, so 

 that all the polar ice was removed from the Alaskan coastal area, leaving 

 only winter ice to be melted after the wind breakup. Very little is known 

 of winter changes in currents around Alaska, 



The packing of the ice due to the wind is greatest at the edges, 

 since the differential frictional coefficient between water and ice is 

 greatest at the edge. The wind forms windrows near the boundary but not 

 deep within the pack. Homogeneity of the surface and the resulting 

 decreased differential frictional coefficient are responsible for the 

 slovf changes in concentration at the center of the pack. 



Much of the pressure on the coasts of the North American side of 

 the pack, from Greenland to Banks Island, probably is caused by currents 

 which drive the ice against the land, rather than by wind stresses. The 

 most impenetrable ice is found along these coasts and in the outer por- 

 tions of the Canadian Archipelago, Within the archipelago, the warming 

 effect of the land induces melting during the summer, so that most of the 

 ice is winter ice. The shore lead which is so prominent along the Alaskan 

 coast and Banks Island usually is absent betvreen Prince Patrick Island and 

 Greenland; of late years no observations from the Lincoln Sea have ever 

 shovm any wide lead comparable to Peary's "Big Lead", and it is doubtful 

 if there ever are large leads along this coast, due to the pressure of 

 the ice. 



A trend in warming of the Arctic has been discussed at length in the 

 literature (e.g., Zubov, 1948). As regards the polar pack, the average 

 thickness of the ice has been estimated to have decreased about one meter 

 in the past fifty years (Zubov, 1940). The implication of this statement 

 for ice forecasting is that the summer coverage of the Polar Basin must 

 be decreasing slowly, as the thinner ice is more easily reduced in area 

 ("melted") than thicker ice. The thermal structure of the surface layers 

 of the PoLar Sea has also changed within the past few decades, the upper 

 layer of cold water being much shallower now than in 1900. In the fore- 

 seeable future, if the trend continues, the thickness of the polar ice 

 will continue to decrease until such time as the ice potential during a 

 year equals the summer melting, when the regime will suddenly pass to a 

 winter ice type, and immediate great changes in climate will take place. 

 This change in thermal structure may presage the end of the current ice 

 age. 



