GROUND-ICE WEDGES 



651 



indications pointed to new loci. About 20 per cent of the tundra 

 is found to be probably underlain by ice of greater or less thick- 

 ness. 



On a 250-mile boat trip from Flaxman Island to Point Barrow 

 in the summer of 19 14, good exposures of muck banks always 

 revealed ice. Many miles were examined closely on foot or from 

 the boat, and very little ice was observed which was not definitely 

 in the form of vertical wedges associated with frost cracks on the 

 surface of the tundra. 



Fig. 20. — Tundra block broken off. Ice wedge at left 



Ground-ice wedges with their accompanying surface features are 

 typically associated with muck formations, and none were seen else- 

 where. River silts, elevated sand and gravel deposits, and soft 

 shales have been carefully examined and the only ice found in such 

 places was evidently of another form and of a different origin. 

 Straight lines leading across the gentle surface undulations of sand 

 spits have frequently been observed, and they could be explained 

 only by frost cracks. No polygonal forms have been seen in such 

 places. The writer is unable to say whether ice wedges develop 

 in such sands, for the exposures made by fresh wave-cutting are 

 seldom more than 2 or 3 feet deep, which is less than the depth 

 reached by the summer's thawing. 



