Mechanism of Stromboli. 505 



the walls of the crater above the funnel of loose material consist of beds 

 of solid lavas and agglomerated fragments, and appear to dip more or less 

 towards the sea, or away from the centre of the island, and were no 

 doubt formed by one of its great early and more central craters at a period 

 excessively remote. Suffioni of steam issue in some spots from between 

 these beds, and the percolation of water was seen in places not far below 

 the brim of the crater. There is a perennial spring of percolated water 

 much higher up upon the island, and under the steeps that mark the 

 position of an ancient crater, so that it is highly probable that rain-water, 

 in greater or less quantity, finds its way into the funnel, and even the 

 tube, of the volcano, although the percolation of sea-water is no doubt 

 the chief source of the supply, which is blown out as steam, and perhaps 

 in part as pulverized water. Each outburst, while we continued to 

 observe them, was preceded by several distinct low detonations, with in- 

 tervals between each of from 4 or 5 seconds to as much as 80 seconds : 

 these, though of a far deeper tone, greatly resembled the cracking 

 noises that are heard when steam is blown into the water of a locomotive 

 tender for the purpose of heating it. These detonations sensibly shook 

 the rock beneath our feet. 



The outburst, when it comes, does not rush forth quite instantaneously 

 or like that of exploded gunpowder. It begins with a hollow growl and 

 clattering sound of breaking or knocking together of fragments of hard 

 material, which very rapidly increases to a roar at its maximum, continues at 

 about the same tension for a period varying from a few seconds to a minute 

 or two, and then rapidly declines, but less rapidly than the augmentation 

 took place. At the first instant of the outburst, the rock on which we 

 stood was very sensibly shaken, the vibrations being both vertical and 

 more or less horizontal ; at the end, and after the fragments have ceased 

 to fall and the dust has cleared away, all tension of vapour in the tube 

 seems for the moment at an end, and the funnel is seen filled merely with 

 rolling clouds of vapour. The noise produced by the outburst is not very 

 loud, and more resembles that of the rush of a heavy railway-train over 

 an iron-girder bridge, when heard at some distance off, than any other 

 sound to which I can compare it, but more fluffy and flat. On examin- 

 ing the existing surface of the island, it is easily discerned, by an eye 

 educated to the observation of extinct volcanic regions, that successive 

 craters have been formed, shifting their positions posterior to the pro- 

 duction of that great and nearly central one from which the main mass 

 of the island was thrown up. The existence of three such craters may 

 be traced ; and the existing little crater is situated at the landward or 

 south-eastern side of a vastly larger crater, all the north-western or sea 

 side of which has been destroyed and buried in deep water, and of which 

 the heavy beds of lava seen under water at both sides of the Schiarrazza 

 are the only remains to seawards of the existing slope of debris. This is 

 represented by the Diagram No. 2, copied from my note-book, in which 



