Ch. XXIX.] AQUEOUS EROSIOX IX PALMA. 641 



rupted across the Cumbrecito, constituting the watershed ; but would 

 have sunk down and bare been replaced by the upper basaltic rocks. 

 If we could assume that the sea once entered the Caldera here as well 

 as by the great Barranco, it might have produced such a breach as e, 

 and such an extension of the line of cliffs as that now observable be- 

 tween e and a, map, p. 628, without any corresponding cliff to the 

 westward of e, a. 



Yet we could discover no elevated outliers of conglomerate to attest 

 the supposed erosion at the Cumbrecito, which is about 3500 feet 

 above the level of the sea. It might also be objected to the hypothesis 

 of marine denudation in Palma, that there are no ranges of ancient 

 sea-cliffs on the external slopes of the island. The flanks of the 

 mountain, except where it is furrowed by ravines or broken by lateral 

 cones, descend to the sea with a uniform inclination. In reply to 

 such a remark, I may observe that we do not require the submergence 

 of the uppermost 3000 feet of the old cone in order to allow the sea 

 to enter both the great Barranco and the Cumbrecito and to flow into 

 the Caldera. It would be enough to suppose the land to sink down 

 so as to permit the waves to wash the base of the basaltic cliffs in the 

 interior of the Caldera, and to wear a passage through the Cumbrecito 

 where there may have been alwa) T s a considerable depression in the 

 outline of the upper formation. But would not the same waves which 

 had power to form in the Barranco a mass of conglomerate 800 feet 

 thick have left memorials of their beach-action on the external slope 

 of the island? IS"o such monuments are to be seen, and their absence 

 raises an objection of no small weight against the supposition of the 

 sea having ever entered the Caldera. It may, however, be said, in 

 explanation, first, that cliffs are not so easily cut on the side of an 

 island towards which the beds dip as on the side from which they 

 dip ; secondly, if some small cliffs and sea-beaches had existed, they 

 may have been subsequently buried under showers of ashes and 

 currents of lava proceeding from lateral cones during eruptions of the 

 same date as those which were certainly contemporaneous with the 

 conglomerate of the great Barranco. 



On the eastern coast of Palma, about half a mile from the sea, in 

 the ravine of Las Xieves, not far from Santa Cruz, we observed a con- 

 glomerate of well-rounded pebbles having a thickness of 100 feet, cov- 

 ered by successive beds of lava, also about 100 feet thick. In this 

 instance the ancient gravel beds occupy a position very analogous to 

 the buried cone, s, p, fig. 698, p. 630. When in Palma, I conceived 

 them to be of fluviatile origin ; but, whether marine or freshwater, it 

 must be admitted that the superposition of so dense an accumulation 

 of lavas to a mass of conglomerate 100 feet thick shows how easily 

 the outer slopes of the island may have been denuded by the sea and 

 yet display no superficial signs of marine denudation, every old beach 

 or delta once at the mouth of a torrent being concealed under newer 

 volcanic outpourings. At the same time I should state that M. 

 41 



