April 2$, 1885] 



NATURE 



593 



siderable depths, but with the increasing heat capillarity loses 

 its power. 



mg these various conditions into consideration, the author 

 whether the surface-waters can penetrate to depths of 

 more than seven to eight miles, and feels it im; ossible to accept 

 any hypothesis based upon an a-sumed percolation to unlimited 

 depths. That there should be open fissures through which 

 water could penetrate to the volcanic foci, he also consi'i 

 bility. 



But the objection to which the . weight 



• the extravasation of the lava being due to the presence of 

 vapour in the volcanic foci, is, that if such were the case, there 

 should be a distinct relation between the discharge of the lava 

 and of the vapour, whereas the result of an examination of a 

 number of well-recorded eruptions shows that the two operations 

 are in no relation and are perfectly independent. Sometimes 

 there has been a large discharge of lava and little or no escape 

 of steam, and at other times there have been paroxysmal explo- 

 sive eruptions with little discharge of lava. 



There are instances in which the lava of Vesuvius has welled 

 out almost with the tranquility of a water-spring. A great 

 eruption of Etna commenced with violent explosions and ejec- 

 tion of scoria:, which, after sixteen days, ceased, but the flow of 

 lava continued for four months without further explosions. In 

 the eruption of Santorin, 1S66, the rock-emission proceeded for 

 days in silence, the protruded mass of lava forming a hill nearly 

 500 feet long by 200 feet high, which a witness compared with 

 the steady and uninterrupted growth of a soap-bubble. The 

 eruptions of Mauna Loa are remarkable for their magnitude, 

 and at the same time for their quiet. Speaking of the eruptions 

 of 1S55, Dana says there was no earthquake, no internal thun- 

 derings, and no premonitions. A vent or fissure was formed, 

 from which a vast body of liquid lava flowed rapidly but 

 quietly, and without steam explosions, for the space of many 

 months. 



On the other hand, paroxysmal eruptions are generally ac- 

 companied by earthquakes, and begin with one powerful burst, 

 followed rapidly by a succession of explosions, and commonly 

 with little extrusion of lava, although it is to be observed that a 

 large quantity must be blown into scorise and lost in the ejec- 

 tions. Such was the eruption of Coseguina in 1835, and of 

 Krakatoa in 1SS3. Sometimes in these paroxysmal eruptions 

 there is absolutely no escape of lava, scoria; alone being pro- 

 jected. A common feature in eruptions, and which indicates 

 the termination of the crisis, is the stopping of the lava, though 

 the gaseous explosions continue for some time with scarcely 

 diminished energy. 



There is thus no definite relation between the quantity of ex- 

 plosive gases and vapours and the quantity of lava. If the 

 eruption of lava depended on the occluded vapour, it is not easy 

 to see how there could be great flows without a large escape of 

 vapour, or large volumes of vapour without lava. The extrusion 

 of lava has been compared to the boiling over of a viscid sub- 

 stance in a vessel, but the cases are not analogous. 



The only logical way in which it would seem possible for 

 water to be present is on the hypothesis of Sterry Hunt who 

 supposes the molten magma to be a re-melted mass of the earlier 

 sedimentary strata, which had been originally subject to surface 

 and meteoric action. But in the end the preceding objections 

 apply equally to this view. 



There is the further general objection to the presence of water 

 in the molten magma, in that were the extrusion of lava due to 

 this cause, the extrusion of granite and other molten rocks 

 (which do not, as a rule, lie so deep as the lava magma) should 

 have been the first to feel its influence and to show its presence. 

 Vet although water is present, it is in such small quantities that 

 these rocks never exhibit the scoriaceous character which lava so 

 commonly possesses. 



Nor is lava always scoriaceous, as it should be if the hypo- 

 thesis were correct. Many lavas are perfectly compact and free 

 from vapour-cavities, and so also are especially most of the great 

 sheets of lava (basalt) which welled out through fissures in late 

 geological times. These vast fissure eruptions, which in India 

 and America cover thousands of square miles, and are several 

 thousand feet thick, seem conclusive against water agency, for 

 they have welled out evidently in a state of great fluidity, with 

 extremely little explosive accompaniments, and often without a 

 trace of scoria: mounds. The general presence of non-hydrated 

 rocks and minerals is also incompatible with the permeation of 

 wa'er which the assumption involve 1 . 



It has been suggested by some writers that large subterranean 

 cavities may exist at depths in which the vapour of water is 

 stored under high pressure, but the author shows that such 

 natural cavities are highly improbable in any rocks, and im- 

 possible in calcareous strata. 



The author proceeds to account for the presence of the 

 enormous quantity of the vapour of water, so constantly present 

 in eruptions, and which, in one eruption of Etna, was estimated 

 by Fouque to be equal to about 5,000,000 gallons in the twenty- 

 four hours. He refers it to the surface-waters gaining access 

 eruptions to the volcanic ducts either in the volcanic 

 mountain itself, or at comparatively moderate depths beneath. 

 He describes how the springs and wells are influenced by 

 volcanic outbursts. By some observers, these effects have been 

 referred to the influence of dry and wet seasons, but there are so 

 many recorded instances by competent witnesses, as to leave 

 little doubt of the fact. This was also the decision of the 

 inquiry by the late Prof. Phillips, who asks, Why is the drying 

 up of the wells and springs an indication of coming disaster? 



The author then considers the hydro-geological condition of 

 the underground waters. He points to the well-known fact, 

 that on the surface of volcanoes the whole of the rainfall 

 disappears at once, and shows that when the mountain is at 

 rest, the underground water must behave as in ordinary sedi- 

 mentary strata. Therefore, the water will remain stored in the 

 body of the mountain, in the interstices of the rocks and scorise, 

 and in the many empty lava-tunnels and cavities. The level of 

 this water will rise with the height of the mountain, and he 

 estimates that it has at times reached in Etna a height of 5000 to 

 6000 feet, while the permanent level of the springs at the base 

 of the mountain seems to be at about 2000 feet. The water does 

 not, however, form one common reservoir, but is divided into a 

 number of independent levels by the irregular distribution of the 

 scorire, lava, &c. These beds are traversed by verticil lykes 

 running radially from the crater, so that, as they generally 

 admit of the passage of water, the dykes serve as conduits to 

 carry the water to the central duct. 



Little is known of the sedimentary strata on which volcanoes 

 stand. In Naples, however, an artesian well found them under 

 the volcanic materials in usual succession, and with several 

 water-bearing beds, from one of which, at a depth of 1524 feet, 

 .1 spring of water rose to the surface with a discharge of 440 

 gallons per minute. When in a state of rest the surplus under- 

 ground waters escape in the ordinary way by springs on the sur- 

 face, or when the strata crop out in the sea, they then form 

 submarine springs. 



During an eruption these conditions are completely changed. 

 The ascending lava, as it crashes through the solid plug formed 

 during a lengthened period of repose, comes in contact with the 

 water lodged arouud or, may be, in the duct, which is at once 

 flashed into steam, and gives rise to explosions more or less 

 violent. These explosions rend the mountain, and fresh fissures 

 are formed which further serve to carry the water to the duct 

 from which they proceed ; or they may serve as channels for the 

 sea-water to flood the crater, when, as in the case of Coseguina 

 and Krakatoa, the volcano is near the sea-level. As the erup- 

 tion continues, the water-stores immediately around the duct 

 become exhausted, and then the water lodged in the more distant 

 parts of the mountain rushes in to supply the void, and the ex- 

 plosions are violent and prolonged according to the available 

 volume of water in the volcanic beds. When this store is 

 exhausted, the same process will go on with the underlying 

 water-bearing sedimentary strata traversed by the volcanic duct. 



The author gives diagrams showing the position of the water- 

 levels be/ore, during, and after eruption ; and deser s the 

 manner in which, if the strata surrounding the duct and below 

 the sea-level become exhausted, the efflux of the fresh water 

 which passed out to sea through the permeable beds, when the 

 inland waters stood at their normal height above the sea-level, 

 these same beds will in their turn serve as channels for the se,.- 

 water to restore the lowered water-level inland. Thus, the ex- 

 current channels which carried the land waters into the sea-bed, 

 and there formed, as they often do off the coasts of the Mediter- 

 ranean, powerful fresh-water springs, now serve as channels for 

 an in-current stream of sea-water, which, like the fresh water it 

 replaces, passes into the volcanic duct. This agrees with the 

 fact that fresh-water remains are common in many eruptions, 

 and marine diatomaceous remains in others ; also that the pro- 

 ducts of decomposition of sea-water are so abundant during and 

 at the close of eruptions. With the fall of the water-levels, the 



