The sea of ice on the floor was in a very wet state at the surface, being at a lower level than the stones on to which the rain from the open hole fell; and here the prismatic structure was not apparent to the eye, nor do I know whether it existed at all. In the Glacière of La Genollière I carried a large block of perfectly prismatic ice into the outer cave, where it was exposed to the free currents of air passing from the pit of entrance to the hole newly opened by the falling in of the ground; and, two days after, the external lines were scarcely perceptible, while on the occasion of our third visit I found that they had entirely disappeared, and the whole block was rapidly following their example. This disappearance of the surface-lines under the action of atmospheric thaw is probably the same thing as their absence when the flooring of ice is thinly covered with water. Wherever the flooring rose slightly towards the edges of the sea of ice, the usual structure appeared again. There were no currents of air in the cave, the candles burning steadily through the whole time of our visit. Excepting for the purpose of detecting disturbance in the air, there is no need of candles, as the two holes in the roof supply sufficient light. Some account of the careful observations made here by M. Thury, at different seasons of the year, will be found in other parts of this book. We passed, on our return, by the source of water which springs from the foot of a rock at some distance from the glacière, and is supposed to form the outlet for the drainage of the cave; but it is difficult to understand how this can be the case, considering the form and character of the intervening ground. The two ice-caves so far described are the least interesting of all that I have visited; but a peasant informed me, a day or two after, that if we had penetrated to the back of the pyramid of snow which lay half under the open hole, being the remains of the large collection which is formed there in the winter, we might have found a deep pit which is sometimes exposed by the melting of the snow. He had some idea that its depth was 30 feet a few years ago, and that its sides were solid ice. I shall have occasion to mention such pits in another glacière; if one does exist here, it has probably been quarried in the ice by the drops from the hole in the roof, and there might be some interest attached to an attempt to investigate it.[17]