Habdoastle. — On the Cause of Volcanic Action. 339 



proportion to their specific heat. The pressures needed to hft a 

 mountain 20 miles wide and 1 mile in average height above its 

 base, from a crust 20 miles thick, would be about 270 tons per 

 inch, givmg a temperature, if converted into heat, within silica, 

 of 3,348^. In neither case is the initial temperature of the rock 

 taken into account. The fusion of rock and extrusion of lava 

 are the more important geologically, but it is not necessary that 

 rock should be fused to give rise to volcanic phenomena. Tem- 

 peratures of 550° and 1,000^, which would not affect a rock, 

 give steam pressures of l,0001b3. and 4 tons per inch, respectively, 

 either of which, but especially the latter, would have great disrup- 

 tive or explosive power, provided a vent was opened for them. 

 The writer contends that volcanic steam, or fused rock, cannot 

 open their own way to the surface ; this must be provided for 

 them by the movements which produce the heat fissuring the 

 rocks above. He contends, also, that volcanic steam results 

 from the heating of a wet rock ; that violent eruptive phenomena 

 cannot be caused by the access of water to heated rocks. It is 

 suggested that in steam eruptions, (such as that at Tarawera,) 

 the steam in escaping tears and crumbles up the free surface of 

 the heated rock as frost acts on a clay bank : hence the fineness 

 of the bulk of the ejecta. A rule is found to hold good in so 

 many cases as to be worth further study — that volcanoes only 

 appear where upheaving forces have acted about more than one 

 axis, the volcanoes being found, not where the lines intersect, 

 but in one or more of the angles formed by them. 



The paper then proceeds to offer a history of the recent out- 

 break at Tarawera, on the lines thus laid down : — 



Crust pressures, acting (as shown by the great fissure-lines) 

 upon an axis lying north-east and south-west, accumulated in 

 the elastic compression of certain beds until they were able to 

 bring about movements of some kind in the rocks within which 

 they acted, and which were at no great depth beneath the sur- 

 face, but whose extent and thickness I make no attempt to 

 estimate. During a fortnight or more before the outbreak these 

 movements were going on, as was shown by the earthquakes 

 experienced in the locality. (That the focus of action was 

 situated at no great depth is indicated by the fact that the shocks 

 were merely local.) The movements affected a considerable 

 mass of wet rock, and were only effected by the exertion of con- 

 siderable force. Judging from the resulting great amount of the 

 ejections, it is probable that the action involved such a deforma- 

 tion of some part of the area of rock compressed as would have 

 amounted to crushing at the surface, and the heat developed in 

 such a case would be proportional to the force employed in the 

 crushing. While this was going on below, the upper rocks were 

 being cracked and fissured by the movements. The line of 

 crushing appears to have passed under the Tarawera, or very 



