after use, so that the door was only kept close by the difference of pressure on the two sides and in the levels. They opened of themselves as soon as the two sides were in equilibrium. The outer door had only a single lever worked from within the lock, while the inner one had one on each side, so that it could be worked either from within or without the chamber. The low-pressure air was introduced by four pipes placed close to one of the walls, and that at high-pressure, for the removal of spoil and water, by a fifth near the bottom. Two 16-inch pipes were passed through the upper arches, and two of 13-inch through the bottom of the wall, both series being provided with stop-valves at the ends. The larger pipes inclined outwards, and were turned up to a vertical position within the chamber. When required for use, the inside valve was opened, the tube filled with spoil and closed again. At a given signal the outer valve was opened, at the same time as that connecting the tube with the high-pressure main containing air of at least 4 atmospheres pressure, which cleared out the contents in a few seconds. The outer valve was then closed, and the filling repeated as before. In the same way the water accumulating in the bottom of the chamber was driven out by connecting the lower tubes with the high-pressure air service. The working chamber was lighted by Edison glow lamps, and a Gramme dynamo worked by a 15 HP. engine, and telephonic communication was established between the face and the office. The preparations having been completed, work was begun early in 1884, and from the first great difficulty was experienced in keeping the chamber air-tight, the compressed air finding its way out, not only through the penetrable strata above, but also along the extrados of the roof and the timber struts required for the support of the ground. It was only after building 20 inch buttress-rings on either side of the dam that the pressure on the chamber could be brought up to 1.8 and 2 atmospheres (absolute), under which condition the roof was completed to about 220 yards from the mouth, at the rate of 13 to 16 yards per month, when the work was stopped by an accident in August, 1884. This was caused by the compressed air getting into the overlying pyritic lignite-bearing strata, and after driving out the water oxidizing the pyrites, whereby sufficient heat was developed to fire the lignite, and the products of combustion penetrating to the chamber caused the death of seventeen men by suffocation. In order to render the workings accessible after this accident, six bore-holes were put down from the surface to the seat of most active combustion to give free issue for the gases, which were driven out of the chamber by projecting highly compressed air into it, the mean pressure being kept at 1.6 atmosphere. On the 4th of October the chamber was accessible, and after the ground had been properly shored up, the pressure was removed and the surfacewater, being no longer kept back, penetrated to the seat of the fire and extinguished it; but the heat developed was sufficient to keep the temperature of the water trickling through the roof at 90° for six months. After the accident the first dam was removed, and the second at 614 feet was erected as previously described. Provision was also made for the active ventilation of the part of the tunnel open to the air, by sinking a shaft a little on one side, and connecting it with the crown of the arch by a short inclined drift at 358 feet from the mouth. This shaft was closed at the top and provided with a Pelzer fan, 6 feet in diameter, with a minimum exhausting capacity of 500 cubic feet per second, a return air-way 6 feet wide being formed on the left side of the tunnel, by a brattice wall of masonry carried from the bottom of the pit to a point 13 feet behind the new dam. Working with compressed air was renewed in March, 1885, but the difficulty of keeping the chamber tight was found to be as great as before. It was therefore resolved to wall up the face at 820 feet, and increase the dam by leaving a breadth of 60 or 70 feet of ground unbroken, with the exception of three small galleries, two at the bottom and one at the crown of the arch, which were carefully lined with masonry. The upper one proved a failure on account of the large leakage at its mouth, and was blocked up by a 39-inch wall, after a length of 85 feet had been driven; but the lower ones being driven in compact clay, succeeded better, and remained perfectly tight when 33 feet had been lined on the left side and 79 feet on the right. A pressure of 2.3 to 2.4 atmospheres was then kept up in the workings by the use of two-thirds of the motive power during the remainder of the time that it was required. The lower galleries were carried forward to some distance beyond the dangerous ground at 983 feet from the mouth, their outer walls being constructed of the full thickness of the lower part of the finished side walls of which they formed part up to a height of 8 feet 8 inches, a further height of 8 feet 4 inches to the opening of the arch being put in by galleries driven above the lower ones in lengths of about 70 feet at a time. This work proceeded at the rate of 5 feet per day during the working period, which was, however, interrupted on four different occasions, owing to fresh fires in the lignite, causing the abandonment of the work for a time. When the walls were finished to 1,302 feet, a rise was put at 1,286 feet, and a length of 13 feet of the arch was completed. From this point the work was carried on in both directions, but principally backwards, at the rate of about 40 feet per month of actual work. The crown of the arch, although partly in running sand, was kept sufficiently tight by timbering and closely-driven packing laths, when care was taken to insert the latter singly, and to drive each one perfectly tight before placing the next. The sand was rendered sufficiently coherent by the compressed air not to run as long as only a small surface was exposed at a time. The arch was joined up to the point where it was first stopped, in September, 1887, and the excavation of the full section and construction of the invert, which was done in face, was resumed after the "decompression" of the chamber, on September 25th, and completed to 1,476 feet at the end of October, 1888. The cost per yard of this part of the work was £248. The full section of the finished tunnel inside is as follows: The further extension towards the Aisne valley was carried on from working shafts, sunk from the surface by means of metal tubbing and compressed air, which are described in a separate memoir. H. B. Some Cases of Wear of Steel Rails. By C. WALCKENAER. There are nine cases of exceptional and abnormal wear of Bessemer steel rails, from various manufactories, on the Western Railway of France. They were double-headed rails, 78 lbs. per yard, 5.1 inches in height, 2.44 inches wide, and 0.71 inch thick at the web. They were supported in cast-iron chairs, and joined with fishplates. They were 19 68 feet (6 metres) in length, and supported each upon eight transverse sleepers in sand ballast. The situation and characteristics of the rails are given in a tabular statement, showing the service of each rail, lasting from ten to sixteen years. The sections of the rails are illustrated. The particulars of wear of four of the rails, from the Beauvoisine tunnel, are stated, and are reproduced in the following Table :— 8 9 Straight line; incline, up 1 in 187 A little damp 0.55 0.16 0.71 0.43 0.16 0.59 The wear or indentation of the rails by the action of the fishplates was very irregular in depth, varying from nothing or nearly nothing to about inch, dependent apparently upon the degree of precision of fit. D. K. C. Wear of Rails of Different Degrees of Hardness. (Revue Générale des Chemins de fer, August 1889, p. 156.) Sixteen rails of different degrees of hardness from four different charges were submitted to observation for wear. The charges A and B contained 0.40 and 0.36 per cent., or a mean of 0.38 per cent. of carbon, and the tensile-resistance of the steel of the rails was 41 48 and 41 42 tons, or a mean of 41.45 tons per square inch. The wear of these rails was 0.76 lbs. per lineal yard. The charges E and H contained only 0.23 and 0.19 per cent., or a mean of 0.21 per cent. of carbon, and the tensile-resistance did not exceed 32.96 and 30 10 tons per square inch, or a mean of 31.53 tons. The wear of these rails was 0.97 lb. per lineal yard. From these results it appears that the wear of the two mild-steel rails was 27 per cent. more than that of the hard rails, and that the amount of wear was nearly in the inverse ratio of the tensileresistance of the steel. The time during which the respective wears took place was one thousand eight hundred and thirtythree days, in the course of which, at the average rate of fourteen and a quarter trains per day, a total of twenty-six thousand one hundred and twenty trains passed over the rails. Allowing a final maximum wear of 7.20 lbs. per yard, it is estimated that the hard rails would have lasted ten years longer than the mild-steel rails. D. K. C. Cost of Maintenance of Trial Lengths of Line laid with Metal Sleepers on the Netherlands State Railways. (Bulletin de la Commission Internationale du Congrès des Chemins de fer, 1889, p. 1138.) The systems of sleepers referred to in column 8 of Table (see next page) are as follows: I. Wrought-iron, weight 88 lbs., Vautherin cross-section, uniform throughout, 7 feet 8 inches long, 9 inches wide. The ends are bent upwards so as to give the rail-bed an inclination of 1 in 20. Ends closed. II. Wrought-iron, weight 104 lbs., Vautherin cross-section. uniform throughout, 8 feet 2 inches long, 83 inches wide. Railbed inclined 1 in 20. Outside the rail the sleeper is bent down again to a radius of 2 feet 5 inches. Ends closed by angle-irons. Two angle-irons across underneath sleeper, 10 inches on each side of centre. III. Mild steel, weight 110 lbs., Prussian State Railways section, uniform throughout, length 8 feet 2 inches, width 93 inches, bent |