PHYSICAL GEOGRAPHY 259 



Avhere for some reason or other no ice-layer is formed in the ground, 

 neither a rudemarks" nor knolls seem to occur. 



The depth at which this ice is present in spring differs greatly 

 in the different parts of the island, and - according to the weather 

 - in the different years. In the northernmost districts this ice may 

 remain throughout the summer for years during cold and damp 

 periods. As a rule, in the first half of June, frozen ground is met 

 with at a depth of 1 I 1 /? metres over the greater part of the island; 

 on the plateau a thicker or thinner layer of this sub-surface ice is 

 no doubt always present throughout the summer, and there, in 

 several places, it gives rise to the formation of swamps, hogs and 

 lakes, as the melting snow and ice cannot drain off. In some districts 

 with very warm ground, where hot springs or other secondary effects 

 of volcanoes occur, the ground is never frozen. 



As is well-known clay and clay-soil develop, by contraction, 

 numerous intersecting clefts; such cracks are also formed during 

 Avinter by the action of frost, and in severe winters loud cracks are 

 constantly being heard, announcing the rending of the ground. The 

 surface layer of soil is therefore traversed by a network of numerous 

 rents and cracks which divide the clay -soil into irregular frag- 

 ments or a number of prisms. On closer investigation these cracks 

 are not only found to occur on gravelly and clayey flats, poor in 

 plants, but also in the clay-humus of the home-fields ; some of them 

 are as fine as hairs, others have a breadth of 2 3 cm. Both these 

 factors, the sub-surface ice and the cracks in the ground, are neces- 

 sary to the formation of u rudemarks" and knolls, and to these should 

 be added two other very important factors, viz. frost and unequal 

 surface-evaporation. 



When, in spring, the snow melts on the cracked and netted 

 surface of a flat consisting either of clayey gravel or of plant-covered 

 clay-soil, and the flat itself thaws at the surface, the water percolates 

 through the ground and the cracks, but cannot escape on account 

 of the sub-surface ice so that the entire laver of soil becomes 



*. 



saturated with water. Where the flat consists exclusively of sand 

 and gravel, without any mixture of clay, the entire soil-layer is 

 evenly saturated and the surface of the water can sometimes rise 



tj 



to a level with that of the ground. Evaporation then takes place 

 evenly over the entire surface, and when the sub-surface ice melts, 

 the surface-water drains away or evaporates, and nothing further 

 happens. But on a clayey gravel-flat intersected by a network of 



