JVofices of Memoirs — Dr. T. Anderson — Subsidences in Iceland. 517 



the giving way of tbis and the escape of the naolten lower layers 

 into the deeper parts of the lake caused the subsidence. 



Similarly the lava which escaped from Asbergi may have been 

 that which now occupies the low ground near the estuary of the 

 Jokulsa, in the direction of Lon. 



On the east and south-east of Lake Myvatu a very extensive 

 eruption, or series of eruptions, has taken place from a chain of 

 craters locally called Gardr Borgir ("the castles of Gardr," which 

 is the name of a farm). The lava flow has occupied nearly all the 

 bed of Lake Myvatu, and flowed down the valley of the Laxa to 

 its mouth at Laxamyri. All this stream of lava is very remarkable 

 for the number and size of the spiracles with which it is studded, 

 and a regular gradation of sizes exists, between spiracles the size of 

 a haycock and cones some of which cannot be less than 200 feet high. 

 These cones and craters, which constitute such a striking feature 

 of Lake Myvatu, may probably be nothing more than spiracles 

 formed by the escape of steam generated by the conflict of the hot 

 lava with the water of the lake. The barrier which holds up the 

 water of the present lake consists of this lava, and caves exist in 

 it which are obviously channels by which molten lava has escaped. 

 These and deeper-seated ones would be those by which the lava 

 escaped and left the depression occupied by the present lake. 

 Between the craters of eruption and the lake no spiracles were 

 noticed, but there is a vei'y remarkable series of rocks — the Dim- 

 muborgir — masses of lava of fantastic shape, 30 or 40 feet high, 

 which have remained standing while the intervening portions have 

 •subsided. They present slickenside marks where the subsiding 

 portions have scratched the masses that have remained standing, 

 and tide-marks where the crust has halted in its descent; also in 

 many places bulgings, where the lava has been scarcely stiff enough 

 to stand, and others where it has actually formed stalactitic masses. 



So much for actual lava subsidences. 



The special object of this paper is to draw attention to a sub- 

 sidence on the slopes of Leirnukr, a volcano sevei'al miles north of 

 this spot, where a large strip of land, perhaps 200 yards wide, and 

 one mile or more long, has been let down to a varying depth, 

 averaging perhaps 60 to 80 feet. 



The faults bounding it, like nearly all the fissures in this district, 

 run north and south ; and the east face, which is most perfect, cuts 

 right through a thick stream of old columnar lava and through a 

 large boss of tuff, round and over which the lava has bedded itself, 

 and also through the tuff rocks at each side of the lava stream. It 

 would appear worthy of consideration whether this great depression, 

 which thus affects all the crust of the volcano impartially, may not 

 have been caused by the falling in of one of the steam cavities which 

 may be presumed to exist under volcanoes after the lavas have been 

 expelled by the steam pressure. 



This would accord with the observation that sedimentary rocks 

 near volcanoes often dip towards those volcanoes. Mr. Goodchild 

 has informed the author that the sedimentary rocks round Arthur's 



