﻿Yol. 66.~] VOLCANO OF MATAVANU IN SAVAII. 629 



Places are common where the surface is covered with fresh sheets 

 of lava measuring an inch or even less in thickness ; and the surface 

 is honeycombed with channels along which the lava has flowed, 

 channels which are occasionally deep and open, but often have a 

 crust, sometimes thick, sometimes so thin as to give way under a 

 man's weight, and form dangerous pitfalls for the unwary. There 

 are, too, caves and bubbles of all sizes up to many feet in diameter, 

 •often also covered with only a thin roof equally liable to collapse 

 (PI. XLIX, fig. 1), and many areas, sometimes extensive, where the 

 liquid lava beneath has found a vent lower down after a crust has 

 formed on the surface, and the whole crust has subsided after the 

 manner of the plain of Thingvalla. Tunnels can be seen by which 

 the lava has escaped. 1 



This class of subsidence by lava flowing out from under a crust 

 seemed to me common on the lower parts of the lava as well as on 

 the upper, and there are two notable examples at the eastern foot 

 of the cone (PI. XLV, fig. 2). The lava-tunnels and evidences of 

 rapid flow, superficial and deep, are numerous along the line of great 

 fumaroles. On my ascent of the cone I had occasion to cross the 

 line three times, in order to find a practicable route. From the causes 

 named above, the surface, especially in the upper part, was so rotten 

 that progress was extremely difficult ; while the certainty that the 

 river of molten lava was running somewhere below at an uncertain 

 depth and under a crust of unknown thickness, combined with the 

 approach of night, rendered delay undesirable. I was, therefore, 

 unable to examine the fumaroles carefully 2 ; but, so far as I could 

 see, one of them was a large hollow with precipitous sides, 

 apparently of the nature of a subsidence from the collapse of a 

 crust of lava by combined remelting and withdrawal of fluid 

 support below ; and subsequent examination of similar places near 

 Kilauea has confirmed this opinion. At Matavanu few of them 

 appear to have been examined, and in still fewer was the liquid 

 lava visible, but I understood that in one it had been seen flowing 

 at a depth of 30 to 40 feet. 3 



The slope of these lava-fields is very gradual. The distance of 

 the crater from the sea is about 7 miles as the crow flies ; but, 

 following the sinuosities of the underground current, it may be 

 estimated as somewhat over 10 miles. If the foot of the cone be 

 taken at a height of 1800 feet, this gives an average fall of one in 

 30, or about 6°, a slope which is of the same order of magnitude as 

 that of many of the Icelandic streams. The slopes of some of the 

 Hawaiian volcanoes formed of the same class of lava are stated by 



1 Compare T. Anderson ' Volcanic Studies' 1902, pis. lviii-lviii a & lxxii. 



2 The next day the wind blew the fumes against us, and prevented further 

 examination. 



3 I have since received from Mr. Barts of Fagamalo a photograph of one 

 such ' pit crater,' showing precipitous walls composed of very numerous thin 

 layers of lava, presumably both surface-flows and intrusive sheets, comparable 

 to those in the wall of the crater; also another photograph showing a pit with 

 its walls coated with lava frozen on to them, somewhat like the wall of the 

 crater shown in PI. XLVII. This pit appears to have been an orifice by 

 which lava rose and flooded the surrounding lava-field. 



