642 E. DE K. LEFFINGWELL 



may be sufficiently elastic to absorb the strain, so that no defor- 

 mation occurs. (3) The formation may be deformed and bulged 

 up, either as a whole or locally along the edge of the ice wedge. 

 (4) The ice may be deformed. 



If the summer's strain has been relieved by readjustment of the 

 material within the polygon block, the next winter will again bring 

 about the conditions which caused the first cracking of the ground. 

 Even if the first crack is full of ice, it may be still a plane of weak- 

 ness for tensive strains; and this will be especially true if the crack 

 has been only partially filled. Granted that it is a plane of weak- 

 ness, the new cracks will open at the same place and a constantly 

 growing body of ice be formed at the locus. That this is the ordi- 

 nary case in tundra formations is seen in the constant association 

 of ice wedges with definite loci of frost cracks. 



Thus the growth of the ice goes on from year to year, possibly 

 failing during mild years, when it may not be necessary for all the 

 cracks to open in order to reheve the strain. If the process were 

 not hindered, the upper edges of the wedges would eventually come 

 into contact, thus completely inclosing a cone-shaped mass of the 

 original ground-material. It is conceivable that the process might 

 still go on by bulging up the ice, as it at first bulged up the ground. 

 It is thus within the limits of possibility that a continuous horizon- 

 tal bed of ice should be formed in this manner, but nothing 

 approaching this possible stage has been observed by the writer. 



The thinnest wedge that has come under observation was about 

 a foot wide, but cracks have been seen accompanied by no surface 

 manifestations (Fig. 6) and with no visible ice below them. No 

 doubt the intervening stages exist, especially in an area such as a 

 recently drained lake bottom, where the process is being initiated. 

 The thin veins have nearly parallel sides and flat tops, as can be 

 seen in Figs. 9 and 10. As the ice increases in size, it becomes more 

 wedge-like in form, since the growth is greatest near the top where 

 the crack opens widest. There is a tendency in the large wedges 

 to spread out under the surface of the ground (Fig. 11). This is 

 exaggerated in oblique sections, as is shown in Fig. 12. 



The bottom of an ice wedge has never been observed by the 

 writer. Most of the banks on the north shore of Alaska are less 



