[Illustration:  DESTRUCTION OF THE TARDES VIADUCT.]

Upon taking as a basis the horizontal displacement of the superstructure, which was 45 meters to the right of the pier, and upon combining the horizontal stress that produced it with that of the loads, the stress exerted upon the body may he deduced.  But this hypothesis seems to us scarcely tenable, especially by reason of the great stress that it would have taken to lift the superstructure.  On another hand, it was possible for the latter to slide over one edge of the pier, and this explains the horizontal distance of 45 feet by which its center of gravity was displaced.  It is probable, moreover, that the superstructure, before going over, moved laterally upon its temporary supports.

The girders were, in fact, resting upon rollers, and the roller apparatus themselves were renting upon wedges, and there was no anchorage to prevent a transverse sliding.

Under the prolonged thrust of a very high wind, the superstructure, by reason of its considerable projection, must have begun to swing like a pendulum.  These oscillations acquired sufficient amplitude to cause the superstructure to gradually move upon its rollers until the latter no longer bore beneath the webs.  The flanges therefore finally bent upward where they rested upon the rollers, through the action of the weight which they had to support, and the entire superstructure slid off into space.

An examination of the bent pieces seems to give great value to this hypothesis.—­Le Genie Civil.

* * * * *

JOY’S REVERSING AND EXPANSION VALVE GEAR.

[Footnote:  A paper read before the Mechanical Section of the British Association, at Montreal, August, 1884.]

Four years ago, in August, 1880, a paper was read on this subject before the Annual Summer Meeting of the Mechanical Engineers’ Society of Great Britain, then held in Barrow-in-Furness, describing this valve motion and its functions, which was then comparatively new.  It was, however, illustrated by its application to a large express goods (freight) engine, built by the London and North-Western Railway Company (England) specially to test the advantages and the endurance of the gear.  This engine had cylinders of 18 inches in diameter and 24 inch stroke, and six wheels coupled 5 feet 1 inch diameter, and was designed by Mr. Webb, the Company’s chief engineer, for their heavy fast goods traffic on the main line.  The engine has been running this class of traffic ever since.  In January, 1884, it was passed through the repair shops for a general overhauling, when it was found that the valve motion was in such good condition as to be put back on the engine without any repairs.

The main object of this present paper is to deal with the advantages of the valve gear and its application to various classes of engines both on land and at sea, and with the results of such applications, rather than treating it as a novelty, to give an exhaustive description of its construction and functions, which was done in the paper above referred to.  A very short description of its action and main features will, however, be necessary to the completeness of the paper, and as a basis from which the improved results to be recorded should necessarily be shown to spring.