498 Mr. R. Mallet on the 



reverse. It is the tension of heated steam or water disseminated through 

 the lava in and beneath the vent which occasions its eruptive action, and 

 the boiling-point of every drop or bubble must be sensibly affected by 

 every barometric variation 



" In the foul weather of winter I was assured by the inhabitants that 

 the eruptions are sometimes very violent, and that the whole flank of the 

 mountain immediately below the crater is then occasionally rent by a 

 fissure, which discharges lava into the sea, but must be very soon sealed 

 up again, as the lava shortly afterwards finds its way once more to the 

 summit, and boils up there as before. Captain Smyth found the sea in 

 front of this talus unfathomable, which accounts for the remarkable fact 

 that the constant eruptions of more than 2000 years have failed to fill up 

 this deep-sea hollow." 



Dr. Daubeny ('Volcanos,' second edition, p. 248) appears to have 

 given but a cursory examination to the crater ; and in his observations on 

 its phenomena only repeats Spallanzani, Hoffman, and Mr. Scrope, as 

 follows : 



" The unremittent character of the eruptions of Stromboli appears to 

 arise, as Mr. Scrope has suggested, from the exact proportion maintained 

 between the expansive and repressive force. The expansive arises from 

 the generation of a certain amount of aqueous vapour and of elastic 

 fluids ; the repressive from the pressure of the atmosphere, and from the 

 weight of the superincumbent volcanic products." 



The mechanism, as imagined by Mr. Scrope, fails, in the author's opi- 

 nion, to account either for the rhythmical character of the eruptions or 

 for the alleged connexion between them and the state of the weather. 

 No equilibrium between the " expansive " and the " repressive " forces can 

 possibly exist at the moment of an outburst, the circumstances of which 

 prove an excess of pressure of many atmospheres, which has been gradually 

 increasing since the last outburst became quiescent. 



To account for the actual facts, we must have such a train of natural 

 mechanism as shall cause a gradual, though rapid, increase of steam- 

 pressure within or beneath the vent or tube of the volcano, until the ac- 

 cumulated pressure suffices to overcome whatever obstacles it may 

 encounter, solid or liquid, and by blowing these away release the pressure 

 itself in a burst of steam, stones, dust, &c. The conditions producing 

 this gradual increase of steam-pressure must be such as shall give rise to 

 the rhythmical recurrence, at comparatively short intervals, of the pheno- 

 mena. These conditions are certainly not to be found, either in the 

 general nearness of balance of any expansive and repressive forces alone, 

 or in any conceivable relation between these and variations of atmospheric 

 pressure. 



Mr. Scrope has, as the author believes, greatly overrated the altitude of 

 the f undus of the crater in stating it at 2000 feet above the sea. But let us 

 suppose the height of the column of liquid lava, between the level of the sea 



