new machines for the compression of air; and, in perfecting the mechanical perforators, unexpected difficulties were encountered, which often threatened to prove insurmountable. The total inexperience and unskilfulness of the workmen, and the necessity of giving to them the most tedious instruction; accidents of most disheartening and discouraging kinds--all tended to delay the successful application of the new system. The first important work to be undertaken was the tracing or setting out of the centre line of the proposed tunnel. It was necessary first to fix on the summit of the mountain a number of points, in a direct line, which should pass through the two points chosen, or rather necessitated by the conditions of the locality, for the two ends of the tunnel in the respective valleys of the Arc and of Rochemolles; secondly, to determine the exact distance between these two ends; and thirdly, to know the precise difference of level between the same points. These operations commenced toward the end of August, 1857. Starting from the Northern entrance at Fourneaux, a line was set out roughly in the direction of Bardonnêche, which line was found to cut the valley of Rochemolles at a point considerably above the proposed Southern entrance of the tunnel. On measuring this distance, however, a second and corrected line could be traced, which was found to be very nearly correct. Correcting this second line in the same manner, always departing from the North end, a third line was found to pass exactly through the two proposed and given points. The highest point of this line was found to be very nearly at an equal distance from each end of the tunnel, and at but a short distance below the true summit of the mountain-point, called the "Grand Vallon." The line thus approximately determined, it was necessary to fix definitely and exactly three principal stations or observatories--one on the highest or culminating point of the mountain, perpendicularly over the axis of the tunnel; and the other two in a line with each entrance, in such a manner that, from the centre observatory, both the others could be observed. At the Southern end, owing to the convenient conformation of the mountain, the observatory could be established at a point not very far from the mouth of the tunnel; but toward the North, several projecting points or counterforts on the mountain necessitated the carrying of the Northern observatory to a very considerable distance beyond the entrance of the gallery--not, however, so far as not to be discerned clearly and distinctly, and without oscillation, by the very powerful and excellent instrument employed. These three points permanently established, remain as a check for those intervening, and serve as the base of the operations for the periodical testing of the accuracy of the line of excavation.