At this point the breadth was 25 yards, and, a few feet above the line in which I crossed, all traces of rock or stone disappeared, and there was nothing but unbroken ice. I had of course abundant opportunities for examining the structure of the ice, and I found in all parts of the fall the same large-grained material, breaking up, when cut, into the usual prismatic nuts. I now rejoined Christian, and we worked our way upwards to the mouth of the cave, penitently desisting from stoning a remaining raven. We observed at the very mouth, by watching the flame of the candles, a slight current outwards, extremely feeble, and on our first arrival I had fancied there was a current, equally slight, inwards, but neither was perceptible beyond the entrance of the cave. M. Soret was fortunate enough to witness a curious phenomenon, at the time of his visit to the Schafloch, in September 1860, which throws some light upon the atmospheric state of the cave. The day was externally very foggy, and the fog had penetrated into the cavern; but as soon as M. Soret began to descend to the glacière itself, properly so called, he passed down out of the fog, and found the air for the rest of the way perfectly clear.[63] M. Soret states that he has not absolute confidence in his thermometrical observations, but as he had more time than I to devote to such details, inasmuch as he did not pass down into the lowest part of the cave, I give his results rather than my own, which were carelessly made on this occasion:--On a stone near the first column of ice, 0°·37 C.; on a stick propped against the column on the edge of the great ice-fall, 2°·37 C.; in a hole in the ice, filled with water by drops from the roof, 0° C. approximately.[64] The second result is sufficiently remarkable. My own observations would give nearer 33° F. than 32° as the general temperature of the cave.