Ch. XXIX] PEAK OF TENERIFFE. 511 



in 1854, -will show the manner in which that lofty cone is encircled on 

 more than two sides by what I consider as the ruins of an older cone, 

 chiefly formed by eruptions from a summit which has disappeared. That 

 ancient culminating point from which one or more craters probably 

 poured forth their lavas and ejectamenta may not have been placed pre- 

 cisely where the present peak now rises, and may not have had the same 

 form, but its position was probably not materially different. The great 

 wall or semicircular range of precipices, c c, surrounding the atrium, b b, 

 is obviously analogous to the walls of a Caldera like that of Palma ; but 

 here the cliffs are insignificant in dimensions when compared to those in 

 Palma, being in general no more than 500 feet high, and rarely exceeding 

 1000 feet. The plain or atrium, b b, figs. 651 and 652, lying at the base 

 of the cliffs, is here called Las Canadas, and is covered with sand and 

 pumice thrown out from the Peak or from craters on its flanks. Copious 

 streams of lava, d d, have also flowed down from lateral openings, es- 

 pecially from a crater called the Chahorra,/, fig. 652, which is not seen 

 in the view, fig. 651, as it is hidden by the Peak. The last eruption was 

 as late as the year 1*798. 



s. w. • n. E. 



Section through part of Teneriffe, from N. E. to S. W. On a true scale ; as given in 

 Von Buch's " Canary Islands." 

 a. Peak of Teneriffe. l>. The Canadas or atrium. 



c. Cliff bounding the atrium. d. Modern lavas. 



f. Cone and crater of Chahorra. 



To what extent the lavas, d d, figs. 651, and 652, may have narrowed 

 the circus or atrium, b, or taken away from the height of the cliff c, no 

 geologist can determine for want of sections ; but should the Peak and 

 the Chahorra continue to be active volcanoes for ages, the new cone, a, 

 might become united with the old one, and the lava might flow first from 

 e to c and then from a to c, fig. 652, so that the slope might begin to 

 resemble that formed by lavas and ejectamenta from the summit a to 

 Guia, on the southwestern side of the cone. 



Madeira. — Every volcanic island, so far as I have examined them, 

 varies from every other one in the details of its geographical and geo- 

 logical structure so greatly, that I have no expectation of finding any 

 simple hypothesis, like that of " elevation craters," applicable to all or 

 capable of explaining their origin and mode of growth. Few islands, 

 for example, resemble each other more than Madeira and Palma, inas- 

 much as both consist mainly of basaltic rocks of subaerial origin, but, 

 when we compare them closely together, there is no end of the points in 

 which they differ. 



The oldest formation known in Madeira is of submarine volcanic origin, 



