. AFFECTING THE PILING-UP OF VOLCANIC CONES. 747 



have been formed by falling4n of superincumbent masses deprived 

 of support beneath, although in the one case the removal of support 

 has been produced by volcanic erosion, and in the other by that of 

 running water in subterraneous channels. For a more particular 

 account of these curious cavities see vol. i. of my Eeport, to the Royal 

 Society, on the great Neapolitan earthquake of 1857. Besides large 

 areas, such as those of the Cisterna and Yal del Bove, which by the 

 mechanism assigned had fallen in, though gradually or piece- 

 meal, yet with sufficient rapidity to be denominated per saltum move- 

 ments (which can only occur where the surface-rocks of the cone 

 are massive and coherent), it cannot be doubted that the continual 

 evisceration always proceeding from active vents by the blowing-out 

 of dust and lapilli must be attended with gradual creeping-in of 

 subterranean ducts and equally imperceptible descent in the whole 

 body, and more especially in the lowermost parts of the mountain 

 mass. 



We have thus, in conclusion, established the existence of two 

 separate but cooperative causes, operative in all active volcanic cones, 

 unless perhaps those few which are based upon solid and practically 

 unalterable rock (such as the Puys of Auvergne), tending to alter 

 the position and level of the strata of deposition of volcanic cones. 

 The modifying forces have therefore seemed to me of sufficient im- 

 portance to deserve the somewhat fuller, though still imperfect, ex- 

 planation of them which I have here attempted. With the ex- 

 ception of the very brief reference, in application of the facts, to be 

 found in my paper on the Dykes of Somma (Quart. Journ. Geol. Soc. 

 vol. xxxii. page 493, commencing at line 32 from the top) I am 

 not, as has been already stated, aware that the matter treated of in 

 this paper has been noticed by any previous writer on volcanos *. 



* Assuming, as may safely be done, that volcanic cones are on the whole dis- 

 continuous masses, the pressure upon the unit of surface beneath the base of such 

 a mountain will be everywhere proportionate to the vertical depth of the material 

 superimposed thereon. If, therefore, we assume a cone 20,000 feet in height, 

 and take the density of its material to be as low as 112 lb. per cubic foot (the 

 figures which I have adopted in my paper on " the Nature and Origin of Vol- 

 canic Heat and Energy," Phil. Trans, pt. i. 1873), the pressure on the base 

 would be 1000 tons per superficial foot ; and if the whole mass were of a 

 density as great as that of pyroxene, the pressure would be about one half more. 

 This is but a rough approximation, and probably much below the truth, as the 

 discrete material must of itself acquire increased density with time by the 

 pressure of its own superincumbent mass. It is a difficult physical question to 

 determine conjecturally whether if, given discontinuous material, as actually 

 found in nature, the pressure on the base, the depth being given, will decrease 

 inversely as the surface thereof; for there is little to guide us to an accurate 

 opinion as to whether the base of an entire mountain can compress the 

 material beneath it to a greater or less extent than the same amount of pressure 

 would compress a single square foot of the same base. The base in such large 

 instances, however rigid, unless it be of continuous and coherent rock, must 

 transmit the vertical pressure outwards laterally in all directions, somewhat 

 after the fashion of the quaquaversal pressure of liquids ; and many complex con- 

 siderations enter into the question of how far in any given case those lateral 

 pressures may extend. 



To contrast this roughly with the actual pressure visited upon the unit of 



