Ch. XXIX.] HYPOTHESIS OF UPHEAVAL. 635 



of the compass from a centre, and yet that no great fractures should 

 have been produced. We should expect to see some open fissures on 

 every side, widening as they approach the Caldera. The dikes, it is 

 true, do undoubtedly attest many dislocations of the mass, which have 

 taken place at successive and often distant periods. But none of them 

 can have belonged to the supposed period of terminal and paroxysmal 

 upheaval, for, had the caldera existed when they originated, the melt- 

 ed matter now solidified in each dike must, instead of filling a rent, 

 have flowed down into the Caldera, tending so far to obliterate the 

 great cavity. 



The second objection is the impossibility of imagining that so vast 

 a series of agglomerates, tuffs, stratified lapilli, and highly scoriaceous 

 lavas could have been poured out with a limited area without soon 

 giving rise to a hill, and eventually to a lofty mountain. Such heavy 

 angular fragments as are seen in the agglomerates, single beds of which 

 are sometimes 200 or 300 feet thick, must, when hurled into the air 

 have fallen down again near the vent, and would be arranged in in- 

 clined layers dipping outwards from the central axis of eruption. It 

 is in perfect accordance with this hypothesis that we should behold 

 agglomerate, lapilli, and scoriae predominating in the walls of the Cal- 

 dera ; whereas in the ravines nearer the sea, where the inclination of 

 the beds has diminished to 10 and even to 5 degrees, the proportion 

 of stony as compared to fragmentary materials is precisely reversed. 

 It is also natural that the dikes should be most numerous where the 

 ejectamenta are to the more solid beds in the proportion of 3 to 1, as 

 at 6, fig. 698, p. 630 ; while the dikes are few in number where the 

 stony lavas predominate (as at c, ibid.). Many of the scoriaceous beds 

 at b may be the upper extremities of currents which became stony 

 and compact when they reached c, and flowed over a more level coun- 

 try ; but this suggestion cannot be assented to by the advocates of the 

 upheaval theory, for it assumes the existence of a cone long before the 

 time had arrived for the catastrophe which according to their views 

 gave rise to a conical mountain. 



If, however, we reject the doctrine that the beds were tilted by a 

 movement posterior to the accumulation of all the compact and frag- 

 mentary rocks, how are we to account for the steepness of the clip of 

 some stony lavas high up in the walls of the Caldera ? These masses 

 are occasionally 50 or 100 feet thick, of lenticular shape, as seen in 

 the cliffs from below, and to ah appearance parallel to the associated 

 layers of scoriae and lapilli. But unfortunately no one can climb up 

 and determine how far the supposed parallelism may be deceptive. 

 The solid beds extend in general over small horizontal spaces, and 

 some of them may possibly be no other than intrusive lavas, in the 

 nature of dikes, more or less parallel to the layers of ejectamenta. 

 Such lavas, when the crater was full, may have forced their way 

 between highly inclined beds of scoriae and lapilli. We know that 

 lava often breaks out from the side or base of a cone, instead of rising 



