Br. Johnston- Lavis — Vesuvius and Monte Somma. 303 



area of section of a lava stream flowing from it ; then before that 

 part of the mass that enters the lower extremity can reach the surface, 

 a lava stream will have been thrown out and have flowed 25 kilo- 

 meters long. 



We therefore see that the igneous magma must remain for a very 

 considerable time during its passage upwards, in contact with the 

 walls of the chimney, even when we have to deal with a rapid out- 

 pour of lava. If, on the other hand, the upper outlet be blocked 

 either by the cooling of the contained mass, or the crumbling in of 

 *its sides, a very extensive column of igneous magma may remain for 

 an indefinite time without much disturbance. 



Now the walls of this tubular channel will be composed of varied 

 kinds of rock, many of which may be rich water-bearing strata. 

 We have therefore a superheated fluid igneous magma under 

 enormous and varying pressure in contact with solid rocks of different 

 chemical composition ; conditions more favourable could hardly be 

 conceived for chemical interchange. 



There would take place the gradual assimilation of water into the 

 fluid rock in quantities proportional to the temperature, pressure, 

 supply, and chemical composition of the fused mass, but probably 

 most dependent upon the length of exposure. When a great erup- 

 tion of lava was taking place, a given mass might traverse the whole 

 of the chimney in a few days, or if the volcano were semi-extinct, as 

 Somma was before a.d. 79, the contact of the two materials might 

 be prolonged over centuries. In the first instance, where little time 

 was given for water absorption, we should expect a gradual outpour 

 of fluid rock, such as is usually the case in that type of eruption, 

 whereas in the second the enormous expansion of volatile matter due 

 to the extensive imbibition of many years, would be so great and 

 take place with such rapidity that in escaping from the tube it would 

 carry the fluid magma with it. 



The mass would naturally have lost much of its heat by raising 

 the absorbed water to its own temperature, and during the eruption 

 this loss would be great by the conversion of the liquid into gaseous 

 water. The rock torn into fragments as the result of this rapid 

 expansion would still experience diminution of pressure, and there- 

 fore temperature, in their ascent, so that, long before crystals or even 

 microliths could form, the rock must have solidified as a vitreous or 

 at least semi- vitreous mass. 



If the conditions for the absorbing of water be less favourable, or, 

 when two paroxysmal eruptions rapidly follow each other, then 

 the explosions will take place with much less violence. At the same 

 time less heat will be lost in expansion, and so the cooling of the 

 mass will take place more gradually, allowing the formation and 

 growth of microliths and crystals. As the rock is less rich in 

 volatile matter, it remains in a plastic state longer, so as to permit of 

 the escape of what few bubbles there are, the rock that results will 

 assume a more compact and crystalline character. By such gradations 

 we pass through what has been called vitreous, microlitic, compact 

 pumice, to pumiceous scoria, and thence to true lava and scoria. 



