150 



Prof. J. Prestwich. 



probability connected with this eruption, opened on the coast. This 

 fissure was traced inland from the shore for nearly 3 miles, varying 

 in width from a few inches to 3 feet. " In some places the water was 

 seen pouring down the opening into the abyss below." This is not 

 strictly an analogous case, but I mention it to show how the activity 

 of the eruption may be promoted by the influx of surface waters. 



Similarly, should the water-level in and under the mountain fall 

 below both the sea-level and also below the general level of the water, I, 

 in the permeable strata p, then not only will the springs at the surface 

 of p fail and the wells run short or dry, but the outward and seaward 

 current will also be reversed, and the water will flow in from the sea 

 to the seat of the volcanic disturbance, through the same channels as 

 those by which the inland waters before escaped. This later con- 

 dition of the eruption is represented in section 3, where the level of I 

 in the strata p falls below the sea-level. With the fall of the water- 

 level, I, the available supply of water becomes gradually exhausted or 

 the channels of communication impeded, and this continues until, 

 with the cessation of the extravasation of the lava, the eruption comes 

 to an end. 



To return to the first stage of the ' eruption. The lava, as it rends 

 and crashes through the plug, comes into contact with the water 

 lodged in the cavities and porous strata about and above the plug ; 

 explosions and detonations follow, violent in proportion to the supply 

 of water. More or less of the lava is hurled into the air and scattered 

 as scoriae and ashes, and the explosions continue so long as water finds 

 its way to the escaping lava, but as the supply becomes gradually 

 exhausted the detonations diminish in power and number, until they 

 finally cease from want of supply. But the extravasation of the lava 

 may, and often does continue, long after this exhaustion of water 

 supply, showing, as I believe, the independence of the two causes ; for 

 otherwise the flow of lava would be accompanied to the last with 

 explosions and detonations due to the escape of the extruding 

 agent. 



This flow of water from the surrounding beds into the volcanic 

 duct, its sudden flashing into steam and the violence of the explosions 

 during the first period of the eruption, are easy to conceive; but 

 greater difficulties attend the following stages when the column of 

 lava has ascended higher and fills the duct, and the level of the 

 underground water has become lowered. The only way water can 

 then gain access is through the walls of the duct into the fluid lava 

 as it ascends. 



This involves some little-understood problems. Of the actual 

 underground conditions we must ever be ignorant, and experiment at 

 present guides us but a short way. Any inquiry must therefore for 



