where it is in use, and where the dampers have to be closed each evening until the following morning, the heat is maintained and the furnace regains its ordinary temperature in half the usual time; there was also an economy of 16 per cent. in the fuel in the month of March, with intermittent working; but it will be easily understood that in winter, when the furnace is working regularly day and night, the economy would be considerably greater, the more so because the apparatus gives better results when the outer temperature is low. This was proved at the same works during a twelve days' trial in November last, when the saving in fuel was 28 per cent., as compared with the corresponding period of the preceding year. The Wery apparatus is of very simple construction and requires only occasional cleaning when coal or tar are used for fuel. It is destined to become generally used in gasworks, as it will effect considerable economies which have not hitherto been realized by any other system. C. G. Storage of Petroleum used as an Illuminant for Prussian Lighthouses. (Zeitschrift für Bauwesen, 1889, p. 397.) Petroleum is almost exclusively used as the illuminant for lighthouses on the Prussian seaboard, and a description of the various special store-tanks is given, and their efficiency is considered; the principal ends kept in view being the avoidance of waste by leakage, and in drawing supplies, and maintaining a low temperature. The tanks are of wrought-iron, and generally of a size sufficient for storing a year's supply. Their shape is cylindrical, the larger ones of 3 feet 3 inches diameter, 5 feet 0 inch high, constructed of-inch to inch plates for the sides, inch to inch for the cover, and inch to inch for the base, that which is of the main importance being an efficient closing of the joints. 7 32 They are made in four different manners, viz., 1st, of single plates with double-riveted joints; 2nd, of the same, but with caulked joints; 3rd, of the same, but with brazed joints; and 4th, with a double skin, forming a water-jacket, with single riveting and caulked or brazed joints. The third mentioned method is that found to be the most efficient, although expensive. One objection to the water-jacketed tanks is that in frosty weather the water has to be let off. Diagrams are given showing the various forms of tanks, and of the special brickwork cellars in which they are placed, these being generally isolated from the other buildings. The tanks are fitted with glass gauges, and are filled by a hose passed through a trap in the roof of the cellar. For comparison with the present improved method of storage, a return is given relating to various lighthouse stations under the old system when the oil was drawn direct from the casks, and varied from 5.4 to 13.7 per cent. A tabular statement is given of the number of tanks and their form of construction at each of the seventeen lighthouse stations, the names of the makers, the cost, cubical contents, cost per litre, and percentage of loss per annum now and formerly, the greatest loss quoted being reduced from 13.7 to 1.01 per cent. Particulars are also given of the cost of construction of the cellars at six stations. D. G. The Construction of Railway Embankments in Soft Soils. (Tijdschrift van het Koninklijk Instituut van Ingenieurs, 1888-89, p. 110.) Where it is necessary to construct embankments on soils of a compressible and treacherous nature, several methods may be employed to ensure a stable foundation. In the early times of railway construction in the Netherlands, when the weights of rolling stock and loads were considerably less than at present, and the velocities of the moving loads comparatively low, the object was obtained by covering the surface of the ground with fascine mattresses, upon which the embankment was subsequently raised. In this way the existing topsoil was rendered tougher and the soft subsoil was gradually compressed and rendered more solid. With the mechanical appliances existing at that time, no other system could have been followed, and where the ground was of an exceptionally untrustworthy quality refuge had to be sought in rows of piles forming the foundation for the dams, as well as for the other works required, such as buildings, sluices, bridges, or aqueducts. These embankments, resting on the surface, have the great drawback of never settling completely and requiring constant raising, as the subsoil, under the influence of the weight of earth and the vibration of traffic, continually sinks or is pressed out sideways. In recent years it has been practicable, through the improvements in mechanical means for transporting large quantities of material, to adopt other methods. In constructing the line Sliedrecht-Gorcum, the sections where treacherous soils had to be dealt with, the silt and the mixed strata of bog, running sand and slippery clay were removed by dredging, sometimes to considerable depths, and the trenches then filled up with sand, forming in this way a sufficiently firm and stable foundation for superstructure. On the line now being constructed between Schiedam and Maassluis it was, however, not in every case possible to revert to this method, as the line running close to existing river embankments, and polder dams, and canals, these would have been Abstracts.] CONSTRUCTION OF RAILWAY EMBANKMENTS IN SOFT SOILS.455 weakened and deflected. In these cases the top layer only of soil and turf is removed, and the sand tipped as much as possible on the centre line. The sand, of greater weight than the subsoil, sinks down, displacing the underlying strata sideways. In consequence of this the ground at the sides bulges up, and has to be levelled again, which only occasions inconvenience where canals are concerned, easily corrected by dredging. Subsequent borings show very irregular movements of soil, the sand in many cases finding its way, not always vertically down, but often sliding sideways according to the varying nature of the morass through which it sank, occasioning great differences in the time of settling. Sometimes this settling was slow and gradual, at other times intermittingly or suddenly going down at irregular intervals. The cost price of this section was about £30 per lineal yard, an amount for which a viaduct on piles could not have been constructed. This communication is accompanied by several drawings. H. S. The Shafts of the Braye Tunnel. (Notice sur les Modèles, Dessins, etc., des Ponts et Chaussées et des Mines à l'Exposition Universelle, 1889, p. 270.) The middle section of the Braye Tunnel is reached by two shafts known as Nos. 2 and 3, 301 feet and 378 feet deep, and distant 1,181 yards and 1,722 yards respectively from the mouth on the Aisne side, which necessitated passing through from 40 to 70 feet of the Soissons sands below the water-level. The method adopted for sinking was similar to that formerly applied in making the Rilly la Montagne tunnel in the Reims and Epernay Railway, namely, the use of cast-iron tubbing through the water-bearing bed, which was afterwards secured by under-pinning with oak cribs, the ground being kept dry during the erection of the latter by the use of compressed air. The shafts, rectangular in section, 5 feet 3 inches broad, and divided into three compartments, of the respective lengths of 5 feet 7 inches, 5 feet and 3 feet 6 inches, were put down to about 30 feet below the water-level by the ordinary method of timber frames, 40 inches apart, and close boarded sides, when further progress by this means became impossible, and cast-iron tubbing was substituted. A separate tubbing was used for each compartment. The rings, 12 inch thick, and 39.4 inches. high, ribbed and flanged inside, are united by bolts through the flanges, the joint being kept tight by an india-rubber ring filling seats turned into the adjacent flanges. The bottom ring was provided with a cutting edge. The erection of the tubbingcolumn was effected at the bottom of the shaft in a depth of 25 to 30 feet of water, and when completed pumping was stopped, and the water allowed to rise to the natural level, in order to prevent irregular movements of the ground during the working, which, by causing dangerous hollows, were likely to endanger the timbering of the upper part of the shaft. The sand was at first excavated by a bucket-dredger, then shell augers were used, but finally divers filling the sand directly into tubs were found to be most expeditious. The diver removed the sand at the bottom of the compartment, but without uncovering the cutting edge of the tubbing; the column was then forced down by hydraulic jacks of 20 tons power, until a depth of 20 to 40 inches was attained, when the sand was again cleared out, and so on, the work being done alternately in the different compartments. It was hoped that a tight joint would have been obtained when the cuttingring had penetrated the clay below, but, owing to the small breadth of the ribs dividing the different compartments, they gave way, and allowed the sand to penetrate from above. It therefore became necessary to have recourse to compressed air in order to render the junction of the tubbing with the ordinary timbering impermeable. For this purpose air-locks were established at the mouth of each compartment; one of these, allowing the passage of timber, was 16 feet high, and the others 7 feet 6 inches each. The aircompressor was driven by a 15-HP. portable engine, and maintained a pressure in the bottom varying from 2.35 to 2.8 atmospheres. The joint was formed of six oak rings, from 9 to 10 inches square, and from 4 feet 4 inches to 6 feet in inside diameter, built into a pile 5 feet high. The broadest rings at the bottom resting. on a ledge cut in the impermeable strata, and the narrowest one bears against the bottom of the tubbing, the cutting edge of the shoe resting in a groove in the upper surface containing an indiarubber washer. The seat for these rings was made by excavating a bell-mouthed chamber, which was filled up with concrete carefully rammed after the rings had been placed in position, and packed with moss and wedged in a similar manner to that employed in coal-pit sinking. The order of placing the rings, which is somewhat peculiar, is described in full detail with illustration. The top ring, No. 1, under the tubbing, was first placed provisionally, then the broader ones, Nos. 6 to 3, were built up, and the hollow behind concreted, and finally No. 2 was driven in between No. 3 and No. 1. The joint between the tubbing and No. 1 ring was further secured by a sheet-iron ring backed by cement. Afterwards the sinking was resumed at the ordinary pressure, and secured by close cribs of oak for a depth of 6 feet, below which ordinary frames 40 inches apart are used as in the upper part. The cost of these appliances to the two pits was £11,240 for a depth of 161 feet 9 inches, or about £210 108. per yard. H. B. The Tunnel of Braye en Laonnois. (Notice sur les Modèles, Dessins, &c., des Ponts et Chaussées et des Mines à l'Exposition Universelle, 1889, p. 255.) The summit dividing the basins of the Oise and Aisne on the navigable canal in course of construction between these rivers is passed by a tunnel 2,580 yards long, at a depth of 400 feet below the crest of a ridge, which is made up of alternations of sands and clays capped by the Calcaire Grossier, the whole being of eocene tertiary age. The stratification, which is regular in the higher parts, is subjected to a disturbance at the base, so that the tunnel, which should be entirely in the lower sand, Sable de Bracheux, with about 40 feet of cover, consisting of clays and lignites between it and the overlying Soissons sands, passes through a short fold of these clays, which at 300 yards from the mouth on the side of the Oise brings the crown of the roof into contact with the upper sands, causing a great flow of sand and liquid clay into the heading, so that it became necessary to carry out this part of the work by compressed-air. The plant required was erected in 1883; it comprised seven portable steam-engines, altogether of 220 HP., driving eight compressors capable of supplying 180,000 cubic feet of air at double the atmospheric pressure, to the working chamber in twenty-four hours. A series of reservoirs, of about 3,000 cubic feet capacity, were also provided for air at 4 to 6 atmospheres absolute pressure, which was used for the removal of the excavated material. The working chamber at the face of the tunnel was formed by a wall of masonry, perforated by air-locks, giving admission and exit to and from the chamber. At first this wall, 11 feet thick, was placed about 400 feet from the mouth of the tunnel, but subsequently a second one, improved in many particulars, was placed at 614 feet, and the first was removed. This dam, built of concrete between retaining walls of dressed stone, was 22 feet thick in centre, and 261 feet at the bottom, where the two air-locks were placed; a passage for a third lock was also provided, but was walled up. In the upper part of the dam two other passages were left for the tubes serving for the removal of the débris, and these were similarly walled up when the tubes were erected. The air-locks were of the section of ordinary mining-galleries, 26 feet long, 5 feet 5 inches broad, and 7 feet 3 inches high, or sufficiently large to allow the passage of mine-wagons. They were provided with a sheet-iron lining, built up in rings, bolted together with india-rubber washers, and air-tight doors closing against seats faced with india-rubber, opening inwards, i.e., from the outer air to the lock, and from the latter into the working chamber. At first the doors were closed by screws, but these were inconvenient on account of the slowness in manipulation, and rack and pinion movements, governed by levers, were substituted. The pressure of the air released the lever, which was then lowered shortly |