Before finally leaving the glacière, I examined the structure of the great stream of ice, at different points near the top of the limiting wall. From its outward appearance it might have been expected to be rough, but it was not so; it was knotty to the eye, but perfectly smooth to the foot, and, when cut, showed itself perfectly clear and limpid. It did not separate under the axe into misshapen pieces, with faces of every possible variation from regularity, that is, with what is called vitreous fracture, but rather separated into a number of nuts of limpid ice, each being of a prismatic form, and of much regularity in shape and size. It was smooth, dark-grey, and clear; free from air, and free from surface lines; very hard, and suggesting the idea of coarse internal granulation. In the large ice-streams of some darker glacières, this ice assumed a rather lighter colour by candle-light, but always presented the same granular appearance, and cut up into the same prismatic nuts, and was evidently free from constitutional opacity. FOOTNOTES: [Footnote 18: _Sancti Liberii locus_, the Swiss Dryasdust explains. There is nothing to connect any known S. Liberius with this neighbourhood, unless it be the Armenian prince who secretly left his father's court for Jerusalem, and was sought for throughout Burgundy and other countries. It seems that Saint Oliver is merely a corruption of S. Liberius, the Italian form of the latter, Santo Liverio, having become Sant-Oliverio, as S. Otho became in another country Sant Odo, and thence San Todo, thus creating a new Saint, S. Todus.--Act SS. May 27.] [Footnote 19: My sisters made a two-days' excursion from Arzier to this glacière in the autumn of 1862, and found no snow in the bottom of the pit. They took the route by Gimel to Bière, intending to defer the visit to the glacière to the morning of the second day; but being warned by the appearance known locally as _le sappeur qui fume_, a vaporous cloud at the mouth of a cavern near the Dent d'Oche, on the other side of the Lake of Geneva, they caught the communal forester at once, and put themselves under his guidance. The distance from Bière is two hours' good walking, and an hour and a half for the return. There was no ladder for the final descent, and the neighbouring châlet could provide nothing longer than 15 feet, the drop being 30 feet. Two Frenchmen had attempted to make their way to the cave a week before; but the old 30-foot ladder of the previous year broke under the foremost of them, and he fell into the pit, whence he was drawn up by means of a cord composed of rack-ropes from the châlet, tied together. However useful a string of cow-ties may be for rescuing a man from such a situation, A. and M. did not care to make use of that apparatus for a voluntary descent, so they were perforce contented with a distant view of the ice from the lower edge of the pit.] [Footnote 20: See the section of this cave and pit on page 41.]