Vol. 65.] THE CAULDRON-SUBSIDENCE OF GLEN COE. 667 



heart of the cauldron, but the chance of proving such a connexion 

 has been destroyed by the continuation of the movement of 

 subsidence. On the other hand, any subaerial products that may 

 have been supplied by the main fault-intrusion have long since been 

 removed by erosion. 



It must also be pointed out, that sources for the great out- 

 pourings of augite-andesite and rhyolite, so characteristic of Glen 

 Coe, can nowhere be recognized ; if these outpourings were fed from 

 marginal intrusions, their conduits have apparently been obliterated 

 during the final stages of the volcanic history, when the fault- 

 intrusion and the Cruachan Granite came to flood the country. 



We had reached this position when Dr. Flett drew our attention 

 to an important paper by Dr. Hans Spethmann x on the Askja 

 caldera, ' Iceland's greatest volcano '. Spethmann's descriptions 

 show so clearly that Askja is in many features nothing more than a 

 modern example of a Glen Coe cauldron-subsidence, that it seems 

 advisable to set forth briefly the main conclusions which this 

 author has derived from its study. Askja and Glen Coe, in fact, 

 furnish two most important stages in what Prof. Judd has happily 

 described as a ' denudation series.' 



The Askja caldera is a roughly circular hollow, about 4± miles 

 broad, situated in the midst of a basaltic mountain-mass of subdued 

 swelling form, the Dyngjufjoll (fig. 13, p. 668). The walls of the 

 caldera rise steeply some 1300 feet, measured from their base, and 

 are interrupted only by the valley of the Askja Op, which itself is 

 bounded on both sides by cliffs. Dr. Spethmann regards Askja 

 as a subsidence-caldera, rather than an explosion-caldera, because 

 of the great size of the hollow, and because of the absence of 

 any corresponding pyroclastic accumulation (op. cit. p. 403). 

 Round the margin of the subsidence one notices at once the small 

 rim- volcanoes which have poured out basalt in a great flood into 

 the caldera, whence a united stream has issued by way of the 

 Askja Op. These rim-volcanoes we may regard as the local 

 uneroded representatives of the deeply eroded fault-intrusion zone 

 of Glen Coe. As Spethmann puts it, the sinking mass has 

 squeezed up the magma along the marginal fracture 

 {op. cit. p. 410). 



The subsidence of the Askja cauldron dates from late Glacial or 

 post-Glacial times, for the walls of the hollow consist largely of 

 glacio-volcanic products, while the surfaces of the lavas, poured out 

 from the rim-volcanoes, are entirely untouched by glaciation. The 

 previous history of the area is doubtful ; but Spethmann believes 

 that the Dyngjufjoll was originally built up as a great shield- volcano 

 of Hawaian type, with its centre corresponding more or less with 

 the centre of the later formed Askja. 



Subsequent developments have an added interest, since they have 

 occurred within the last half-century. On March 29th, 1875, a 



1 ' Vulkanologische Forschungeia im ostlichen Zentralisland ' JSfeues Jahrb, 

 xxvi. Beilage-Band (1908) p. 381. 



