638 E. BE K. LEFFINGWELL 



The waves had undercut this bank the summer before, making 

 a long, low cave, perhaps ten feet deep. During the following 

 winter this cave had been flooded at high tide and partially filled 

 with ice. Early in the summer the face of the bank had been 

 masked by slumping, leaving only a few glimpses of true ground- 

 ice under the turf. Shortly before the time of observation the 

 lower part of the bank had been washed clear of debris, exposing 

 the continuous layer of new sea-ice. This appeared to be ground- 

 ice, when coupled with the scattered exposures of undoubted 

 ground-ice above. Luckily the cave was exposed to view at one 

 point, so that the mistake in interpretation could be corrected. 



The writer went into the field in the summer of 1906 with the 

 idea that the coastal ground-ice occurred in horizontal sheets, and 

 in consequence of faulty exposures did not learn its real distribution 

 until 1914. During the first eight summers, although the ground- 

 ice was examined at every opportunity, little insight was gained 

 into the method of its formation. There has been no opportunity 

 to go anew through the literature before writing this paper, but such 

 illustrations as are at hand do not bear out the inferences drawn 

 from them as to large horizontal beds of ice. 



The usual theory advanced in the literature is that bodies of 

 snow or ice were buried by peat or wash material and thus pre- 

 served. The writer sought to interpret the Alaskan coastal ground- 

 ice in the Hght of this theory, but could neither postulate a 

 satisfactory source for the ice, nor find any workable hypothesis to 

 account for its preservation. It was not until the summer of 1914 

 that the fact was forced upon him that most of the ice was formed 

 in place in the ground. A vertical wedge of ice within a peat bed 

 first drew his attention to the fact; for such a dike of ice could not 

 have stood up in the air for the hundreds of years that were neces- 

 sary for the formation of the peat (Fig. 5). 



PROST CRACKS 



During the Arctic winter, frequent reports are heard, coming 

 apparently from the ground. Often the sound is accompanied by 

 a distinct shock, which is in fact an earthquake of sufl&cient inten- 

 sity to rattle dishes, etc. One is justified in ascribing this phe- 



