In our previous visit we had noticed a natural basin of some size and depth among the trees on the north side of the road, and we now found that the chaos was the result of a recent falling-in of this basin; so that from the bottom of the first cave, standing as it were under the road, we could see daylight through the newly-formed hole.
The total length of the floor of the inner cave, which lies north-east and south-west, is 51 feet; and of this floor a length of about 37 feet was more or less covered with ice, the greatest breadth of the ice being within an inch or two of 11 feet.

Excepting in the part of the cave already mentioned as being less than 3 feet high, we found the floor not nearly so dry, nor so completely covered with ice, as when we first saw the glaciere, three years before, in the middle of an exceptionally hot August.

Under the low roof all was very dry, though even there the ice had not an average thickness of more than 8 inches.

It may be as well to say, once for all, that the ice in these caves is never found in a sheet on a pool of water; it is always solid, forming the floor of the cave, filling up the interstices of the loose stones, and rising above them, in this case with a surface perfectly level.
[Illustration: ICE-COLUMNS IN THE GLACIERE OF LA GENOLLIERE.] We found four principal columns of ice, three of which, in the loftiest part of the cave, are represented in the accompanying engraving: I call them three, and not two, because the two which unite in a common base proceeded from different fissures.

The line of light at the foot of the rock-wall is the only entrance to the glaciere.