"Like those of the Vivarais [the volcanoes of Catalonia] are all, both cones and craters, subsequent to the existence of the actual hills and dales, or, in other words, no alteration of previously existing levels accompanied or has followed the introduction of the volcanic matter, except such as the matter erupted necessarily occasioned. The cones, at least fourteen of them mostly with craters, stand like Monpezat, and as perfect; the currents flow down where the rivers would be if not displaced. But here, as in the Vivarais, deep sections have been cut through the lava by streams much smaller in general, and at certain points the lava is fairly cut through, and even in two or three cases the subjacent rock. Thus at Castel Follet, a great current near its termination is cut through, and eighty or ninety feet of columnar basalt laid open, resting on an old alluvium, not containing volcanic pebbles; and below that, nummulitic limestone is eroded to the depth of twenty-five feet, the river now being about thirty-five feet lower than when the lava flowed, though most of the old valley is still occupied by the lava current. There are about fourteen or perhaps twenty points of eruption without craters. In all cases they burst through secondary limestone and sandstone, no altered rocks thrown up, as far as I could learn, not a dike exposed. A linear direction in the cones and points of eruption from north to south. Until some remains of quadrupeds are found, or other organic medals found, no guess can be made as to their geological date, unless anyone will undertake to say when the valleys of that district were excavated. As to historical dates, that is all a fudge ... I can assure you that there never was an eruption within memory of man." At Luchon Lyell rejoined Captain Cooke, and they visited one or two interesting spots in the more western part of the Pyrenees, such as the Cirque de Gavarnie and the Brèche de Roland. The former would afford object-lessons on the erosive action of cascades; the latter would set him speculating on the causes which could have fashioned that strange portal in the limestone crest of the mountain. They descended some distance on the Spanish side of the Brèche, in order to make a more complete investigation of the structure of the chain, sleeping at a shepherd's hut and returning across the snowfields next day. It is evident that whenever there was a hope of securing any geological information or of seeing some remarkable aspect of nature, Lyell was almost insensible either to heat or to fatigue.