conceivable; and each careful essay with the stone seemed to involve as much responsibility as taking a shot at a hostile wicket, in a crisis of the game, instead of returning the ball in the conventional manner.  When at last it was safely lodged, the height proved to be 27 feet.  I had hoped to find it much more than this, from the grandeur of the effect of the whole mass, and I took the trouble to measure the knotted string again with a tape, to make sure that there was no mistake.  The column formed upon the fir-tree was 3 or 4 feet lower.

I have since found many notices of this glaciere in the Memoirs of the French Academy and elsewhere, extracts from which will be found in a later chapter.  These accounts are spread over a period of 200 years, extending from 1590 to 1790, and almost all make mention of the columns or groups of columns I have described; but, without exception, the heights given or suggested in the various accounts are much less than those which I obtained as the result of careful measurement.  The latest description of a visit to the glaciere states a fact which probably will be held to explain, the present excess of height above that of earlier times.[37] The citizen Girod-Chantrans, who wrote this description, had procured the notes of a medical man living in the neighbourhood, from which it seemed that Dr. Oudot made the experiment, in 1779, of fixing stakes of wood in the heads of the columns, then from 4 to 5 feet high, and found that these stakes were the cause of a very large increase in the height of the columns, ice gathering round them in pillars a foot thick.  So that it is not improbable that the largest of the three masses of the present day owes its height, and its peculiar form, to a series of stakes fixed from time to time in the various heads formed under the fissures in the roof, though nothing but the most solid ice can now be seen.  It would be very interesting to try this experiment in one of the caves where, without any artificial help, such immense masses of ice are formed; and by this means columns might, in the course of a year or two, be raised to the very roof.  Further details on this subject will be given hereafter.

There was no perceptible draught of air in any part of the cave, and the candles burned steadily through the whole time of my visit, which occupied more than two hours.  The centre was sufficiently lighted by the day; but in the western corner, and behind the largest column, artificial light was necessary.  The ice itself did not generally show signs of thawing, but the whole cave was in a state of wetness, which made the process of measuring and investigating anything but pleasant.  I had placed two thermometers at different points on my first entrance—­one on a drawing-board on a large stone in the middle of the pond of water which has been mentioned, and the other on a bundle of pencils at the entrance of the end chapel, in a part of the cave where the ice-floor