Ch. XXIX.] CALDEEA' OF PALM A. 499 



channel of discliarge subsequently buried under the products of newer 

 eruptions. Countless dikes, more or less vertical, consisting chiefly 

 of basaltic lava, traverse the walls of the Caldera, some of them ter- 

 minating upwards, but a great number reaching the very crest of 

 the ridge, and therefore having been posterior in origin to the whole 

 precipice. 



We could not discover in any one of the fallen masses of agglomerate 

 which strewed the base of the cliffs a single pebble or waterworn 

 fragment. Each imbedded stone is either angular, or, if globular, consists 

 of scoriae more or less spongy, and evidently not owing its shape to attri- 

 tion. It would be impossible to account for the absence of waterworn 

 pebbles if the coarse breccia in question had been spread by aqueous 

 agency over a horizontal area coextensive with the Caldera and the vol- 

 canic rocks which surround it. The only cause known to us capable of 

 dispersing such heavy fragments, some of them 3, 4, or 6 feet in diame- 

 ter, Avithout blunting their edges, is the power of steam, unless indeed we 

 could suppose that ice had co-operated with water in motion ; and the 

 interference of ice cannot be suspected in this latitude (28° 40'), espe- 

 cially as I looked in vain for signs of glacial action here and in the other 

 mountainous regions of the Canary Islands. 



The lower formation of the Caldera is, as before stated, equally of 

 igneous origin. It differs in its prevailing color from the upper, exhibit- 

 ing a tea-green and in parts a light yellow tint, instead of the usual 

 brown, lead-colored, or reddish hues of basalt and its associated scoi"ia3. 

 Beds of a light greenish tuff are common, together with trachytic and 

 greenstone rocks, the whole so reticulated by dikes, some vertical, others 

 obhque, others tortuous, that we found it impossible to determine the 

 general dip of the beds, although at the head of the great gorge or 

 Barranco they certainly dip outwards, or to the south, as stated by Von 

 Buch. But in following the tsection down the same ravine, where the 

 mountain called Alejanado ((?, figs. pp. 494 and 497) is cut through, 

 and where the rocks of the lower formation are very crystalline, we 

 found what is not alluded to by the Prussian geologist, that the beds 

 exposed to view in cliffs 1500 feet high have an anticlinal arrange- 

 ment, exhibiting first a southerly and then a northerly dip at angles 

 varying from 20 to 40 degrees (see section, fig. 646 at k). Hence 

 we may presume that the older strata must have undergone great 

 movements before the upper formation was superimposed. No or- 

 ganic remains having been discovered in the older series, we cannot 

 positively decide whether it was of subaerial or submarine origin. 

 We can only affirm that it has been produced by successive erup- 

 tions, chiefly of felspathic lavas and tuff's. Many beds which probably 

 consisted at first of soft tuffs have been much hardened by the contact 

 of dikes and apparently much altered by other plutonic influences, so 

 that they have acquired a semi-crystalline and almost metamorphic 

 character. 



The existence of so great a mass of volcanic rocks of ancient date 



