RECENT WORK ON VOLCANOES 97 



stands. In the Askja the volcano was formed in the same way, 

 but towards the end of its activity the mass of lava collapsed, 

 leaving a rim of the pelagonite tuff standing all round. 



In the south-eastern corner lie the two crater lakes, the von 

 Knebel and the Rudloff lakes. The former is much the larger ; 

 it lies against the marginal hills which rise 1,500 ft. above the 

 level of the water in step-like or vertical cliffs. On the north 

 and west the walls are made of the Askja basalt in which the 

 lake is sunk 180 ft. Owing to the great steepness of the sides, 

 there is a continual falling of stones, some of which shoot out a 

 couple of hundred feet into the lake. Along the southern shore 

 there are a great number of solfataras. 



The Rudloff lake, so named after the artist who was with 

 von Knebel when he met his death, is of much more recent 

 origin. It was formed in the eruption of 1875, and the pumice 

 thrown out of this small orifice still covers all the eastern side 

 of the island. There is a small crater ring round it, rising some 

 35 ft. above the Askja lava, but the level of the lake is 180 ft. 

 below it. The water is milky white and still steaming, while 

 from the surrounding walls solfataras gush forth, covering the 

 rocks with sulphur and gypsum crystals. 



The Dyngjufjoll with its Askja caldera stands isolated and 

 almost in the centre of eastern Iceland. No vegetation grows 

 upon it and there is none within many miles ; all around are 

 the plains of bare, black lava, covered in places with the grey 

 pumice of the Rudloff crater. The ponies carrying supplies 

 have to be driven back to grazing-ground immediately they have 

 been off-loaded, and should an expedition be cut off by storms 

 or by other mishap from relief from outside, it would be quite 

 impossible for the members of the expedition to reach safety. 

 Caldera are now known from many examples, such as the above- 

 mentioned case of the Ries in Swabia, and there is an excellent 

 instance of one in Glen Coe in Scotland (15). These types 

 simply show a central plunger with crush zones and volcanic 

 products round the rim. In the Hegau(i4), not far from the 

 Ries, we have an example where the floor of the depression is 

 flooded with lava. All these are circular pits ; it was not till the 

 Icelandic occurrences were described that the relationship 

 between the caldera and faulting became clear. In the Askja, 

 in addition, we have two sets of faults : an outer set by which 

 the block was elevated either relatively to the surrounding 

 7 



