The atmosphere at this point was a sort of frozen vapour, most unpleasant in all respects, and the candles burned very dimly.

The thermometer stood at 32 deg., half-way down the slope of stones.
We were able to stretch a string in a straight line from the lowest point we reached, through the interstices of the blocks of stone, and up to the entrance-hole, and this measurement gave 50 feet.
Considering the inclination of the upper ice-floor, and the sharpness of the angle between the wall of ice and the line of our descent to this lowest point, I believe that 50 feet will fairly represent the height of the ice-wall from this point to the foot of the slope from the upper wall; so that 72 feet will be the whole depth of ice, from the top of the third ladder to the point where our further progress downwards was arrested.

The correctness of this calculation depends upon the honesty of Mignot, who had charge of the farther end of the string, and was proud of the wonders of his cave.

A dishonest man might easily, under the circumstances, have pulled up a few feet more of string than was necessary, but 50 feet seemed in no way an improbable result of the measurement.
[Illustration: SECTION OF THE LOWER GLACIERE OF THE PRE DE S.LIVRES.] The ice was as solid and firm as can well be conceived.

The horizontal bands would seem to prove conclusively that it was no coating of greater or less thickness on the face of a vertical wall of rock, an idea which might suggest itself to anyone who had not seen it, and I think it probable that the amount of ice represented in the section of the cave is not an exaggeration.