February 24, 1898J 



NATURE 



393 



below. Should that lake by any great shock be suddenly 

 let out — not an improbable catastrophe — what a terrible 

 deluge would sweep the plain of the Araxes ; and, if the 

 lake were wholly or partly drained, what a history of the 

 gradual heaping up of waters on the flanks of an active 

 volcanic region might be studied in its depths. 



Here we saw not a mountain valley dammed, but the 

 whole foot of a mountain region enclosed by volcanic 



f\T»yoJ — 



Fig. I. — Diagram to explain the mode of occurrence and origin of Lake Gokcha 



ejectamenta having their origin along a belt of country 

 lying a little oflf and in front of the high ground. From 

 this the inference is obvious that the outlet was not 

 always where we crossed it, and we were therefore pre- 

 pared to find here and there curious conditions of the 

 rock pointing to a different line of exit for the percolating 

 and overflowing water. 



What struck us most in examining the outflowing 

 stream at Elenofka was its very small volume con- 

 sidering the extent of the lake that feeds it, and we 

 could see no evidence of any considerable difference of 

 level between the late summer's flow and the winter's 

 flood waters. The stream filled its channel up to its 

 reedy green bank, and no bare beds of sand and gravel 

 or heaps of torrent debris suggested that it ever rose any 

 higher. Perhaps when the waters do rise, they find their 

 way through broken or scoriaceous rock to other outlets. 



Driving on through ridge and tumbled crag of brown 

 and black cindery rock — the great barrier of lava that 

 holds back the lake — we suddenly came to the southern 

 flank. We looked out through an opening in the lava 

 over the rich valley of the Araxes, and first caught sight of 

 Ararat rising grandly alone into the soft southern sky. 

 Distance did not dwarf its 17,000 feet ; but whether it 

 was active at the same time that the Elinofka volcanoes 

 were locking up Lake Gokcha, I cannot say. Perpetual 

 snow now covers this hill of ancient fires. There it has 

 stood for ages, looking 

 down on the region where 

 tradition carries us back 

 to the earliest homes of 

 civilised man. 



As we descended from 

 the great masses of lava 

 around Lake Gokcha, we 

 suddenly came upon a 

 most interesting and sug- 

 gestive section, where the 

 road winds under a cliff, 

 partly natural and partly 

 cut back for the road 

 itself, so as to exhibit _ 



now a face of newly- 'ii*^ 



bared rock. In this we 

 saw alternations of dark 

 green or brown -black 



obsidian, and a white product of decomposition of 

 volcanic rock standing in vertical planes (see Fig. 2). 

 The white rock is, especially in the upper part of the 

 cliff, apt to be broken up into a kind of breccia made up 

 of pieces of all sizes and shapes. The first impression 

 produced by the section would be that it consisted of a 

 number of bands of obsidian alternating with a coarse 

 agglomerate of fragments of one and the same kind 



NO. 1478, VOL. 57] 



of decomposed igneous rock, and all turned on end 

 by subsequent movement. But an examination of the 

 ground soon showed that view to be untenable, because 

 the section was cut through the terminal margin of a lava- 

 flow which had come down from the great volcanic group 

 forming the southern rim of Lake Gokcha (see Fig. 3, 

 in which the arrow indicates the direction of flow). The 

 obsidian bands are therefore at right angles to the 



bedding. 



Another view was there- 

 fore suggested, namely, 

 that the obsidian was in- 

 jected or found its way 

 along vertical fissures in 

 the crushed and brecciated 

 rock of an earlier flow. 

 But had this been so, 

 there seemed no reason 

 to limit the injection to 

 the vertical planes. The molten liquid mass would 

 have penetrated in all directions through the frac- 

 tured rock. Besides, an examination of the rock in 

 detail showed that the obsidian was not injected, for 

 it occurred in isolated masses in the midst of the 

 white porcelanous or powdery rock. The plates as- 

 sumed a fan-shaped arrangement in places, as if their 

 position had been determined by joints converging down- 



FlG. 2. 



wards. The mode of occurrence of the rock suggested the 

 true explanation of the phenomenon. The white rock was 

 decomposed obsidian. The whole mass was a quickly 

 cooled lava traversed by shrinkage joints at right angles 

 to the cooling surfaces, and roughly parallel to the line 

 of flow. Water followed the cracks, acted on the 

 silicates, and changed the glassy lava to the white chalk- 

 like mass in which bands of obsidian, preserved between 



the joints, still remained samples of what the whole rock 

 once was. 



A similar change may be observed at Obsidian Cliff 

 in the Yellowstone Park in America, where the same 

 alternation of bands of decomposed and sound obsidian 

 may be seen as in the Armenian section, though not 

 on such an extensive scale. The same process picks 

 out by decomposition the spheroidal shells, which are 



