sidings. At the end of 1876 more than 90 per cent. of the whole was laid with steel, the substitution for iron, having been in progress since 1864. The effect of this on the maintenance of the line is seen in the following table :

The rate at which the actual substitution of steel for iron proceeded is given in the following table, representing the number of Bessemer steel rails laid down and removed in each

year since

Out of the total of 1,935 rails rendered unserviceable, 1,204 broke through the full section, 227 through the fish-bolt holes, and 504 were otherwise damaged. That the number of removals does not increase, but has substantially diminished since 1874, is accounted for by the fact that these removals are necessitated not so much by wear as by defects in manufacture, which are usually discovered within a short time after the rail has been laid.

In order to obtain accurate data as to the comparative efficiency of different classes of rails, a number of samples from different makers were laid on a part of the line having the heaviest traffic, near the Oberhausen station. The experiment commenced in 1864, and the results obtained up to the end of 1876 were as follows.

The rails were all of the same section, called Calibre IV., and 5,650 millimètres area :

The average wear of the experimental Bessemer rails is 4.86, which represents the effect produced by the passage of 6,500,000 axles of passenger and goods trains, or about 1,340,000 axles for each millimètre of wear.

H. B.

Experiments on Steel Rails. By J. Van Hamel.

(Organ für die Fortschritte des Eisenbahnwesens, vol. xiv., pp. 231-233.) These experiments were made on steel rails ordered for the Transvaal Republic, and manufactured by John Cockerill and Co., of Seraing. They were of the Vignoles type, weighing 56 lbs. per yard, and of the following dimensions: height, 3.9 inches; breadth of foot, 3.9 inches; breadth of head, 2.2 inches; thickness of web, 0.4 inch. Sixteen rails were experimented on, and the tests and results were as follows::

(A) The rail was to be placed on bearings 3 feet 7 inches, and receive a blow from a weight of 1,102 lbs., falling freely from a height of 19 feet 6 inches, without showing a set of more than 1.8 inch; and was then to be turned over and straightened back again under similar blows without breaking. The whole of the rails bore this test well, the largest set under the first blow being 2.1 inches. They were subsequently nicked in the foot, and then broken by blows of the same weight; the number of blows required varied from one to eight, the last number occurring with a rail which had been a long time under a hot sun, and may thus have been rendered more ductile.

(B) The rail, placed on the same bearings, was to support a weight of 9.8 tons at the centre for five minutes without showing any permanent set; and subsequently a weight of 27.5 tons under the same conditions without breaking. The whole of the rails bore both these tests satisfactorily, the permanent sets in the second case varying from 1.97 inch to 4.9 inches.

(c) From each charge two small ingots were taken, and forged into bars 0.8 inch square, which when cold were bent double without breaking: the object of this test being to show that phosphorus, sulphur, and silica, were not present in an inordinate degree.

(D) Four pieces were cut off finished rails, forged into square bars, and then turned down to four different diameters, 0.59, 0.63, 0.61, and 0.73 inch respectively. These specimens, each 4 inches long, were then tested separately in a hydraulic press for tensional strength. The three first specimens behaved nearly alike, beginning to stretch sensibly at about 22.2 tons per square inch, and breaking at about 38 tons per square inch, with a final extension of about 18.5 per cent., and contraction at the point of fracture of about 16 per cent. The fourth specimen began to lengthen at about 19 tons, and broke at 34.3 tons: but from the fracture it appeared to have been somewhat overheated, and thus not to give a fair test.

The above results show sufficiently the nature of the steel; which belonged to the category called in Belgium "Très-tendre or "Tendre," not capable of hardening in water. These qualities have from 0.18 to 0.20, and from 0.20 to 0.28 per centage of carbon respectively, and were used in this case on account of local circumstances, which did not allow of the quality" demi-dur,” which is preferred in Europe. This quality has 0.28 to 0.30 per centage of carbon. The charge of raw material used by Messrs. Cockerill when making this steel is given in detail, and also its chemical composition when melted.

W. R. B.

Mallet's Locomotives with Compound Cylinders.
By M. LESTANG.

(Revue industrielle, vol. viii., pp. 393, 394.)

The locomotive stock of the Bayonne and Biarritz railway is constructed with compound cylinders on M. Mallet's system: the first or smaller cylinder to the left side, and the second or larger cylinder to the right side. The engines are constructed like ordinary tank-locomotives, with six wheels, of which four are coupled; they have horizontal outside cylinders, with the addition of a starting valve (tiroir de démarrage), by means of which steam

LUBRICATING FLANGES OF LEADING WHEELS OF LOCOMOTIVES. 315

may be admitted direct from the boiler to both of the cylinders, for starting a train, or when full power is required for ascending an incline. In ordinary conditions, steam is admitted from the boiler to the first cylinder only, whence it is exhausted into the second cylinder, and thence into the chimney.

The railway is 5 miles long, and the gauge 4 feet 8 inches; there is one incline of 1 in 66 for one-third of the length, and there are curves of 440 yards radius. The locomotives, three in number, were constructed by Messrs. Schneider and Co.

The dimensions

in working order

There are fifty-eight trains daily, making a total of 290 miles run per day. On the total mileage run, which amounted to upwards of 30,000 miles at the end of September, the total fuel consumed did not exceed 14 lbs. per mile. The weight of the train was from 40 to 50 tons, rising occasionally to 60 or 70 tons, taken at an average speed of 20 miles per hour. The steam is expanded four times, and neither the stability of the engine nor the force of the draught appears to be affected by the compound arrangement.

D. K. C.

On Lubricating the Flanges of the Leading Wheels of Locomotives. (Organ für die Fortschritte des Eisenbahnwesens, vol. xiv., pp. 183, 184.)

The rapid wear of the flanges of the leading wheels is an important element in the cost of repairs of a locomotive engine. Various attempts have been made to lessen this wear, especially by giving a more decided coning to the tires; but this device has several disadvantages, and cannot be applied where the leading wheels are coupled to the engine. It has lately been proposed to lubricate the flanges at the point where the greatest wear takes place, and the State railways of Bavaria have experimented extensively on a system of this kind, designed by Herr Fischer von Rösslerstamm, of the Kaiserin-Elisabeth-Railway. The first attempts were made with fixed tablets of hard grease, inserted in casings, which were fastened to the bearing springs of the wheels in such a way that the grease tablets were close to that

part of the flanges where the chief wear takes place. Tablets of three different degrees of hardness were provided, to suit the temperature at different times of the year. Better results were obtained, however, by using tablets of felt, stiffened by thin layers of wood, and saturated with either rape, or "Vulcan" oil, thus obtaining the advantage of being independent of temperature. Experiments were made both with passenger and with six-coupled goods engines, care being taken that wherever two engines were tried against each other they should be fitted with tires of the same make and date, and should as far as possible have the same running duty. The general result after allowing for accidental discrepancies, was to establish that the duration of lubricated as against unlubricated flanges may be taken to be in the proportion of 3 to 2; an advantage so large as to put the cost of lubrication quite out of account. In addition, the noisy grinding of the wheels round sharp curves was quite done away with, and the motion rendered smooth and pleasant. The corresponding saving of wear and tear in the rails is also not to be lost sight of. The lubrication is not applied in wet weather, as the rain forms a natural lubricant, and has also to be suspended in time of snow, which is apt to hang on the tires and clog the lubricating boxes. The apparatus requires no attention beyond the occasional dropping on it of a little oil by the driver; and the fear that the face of the tires might get greasy and lose adhesion has proved groundless, no case of wheels slipping from this cause having been reported. The cost of lubrication with felt and oil has been found to average 38. to 58. per 10,000 kilomètres, or say 128. to 20s. per engine per annum.

W. R. B.

Warming Railway Carriages. By F. CLÉRAULT.

(Annales des Mines, vol. xi., 1877, pp. 129-182.)

M. Clérault gives an abstract of a Report by M. L. Regray, addressed to the Administration of the Chemins de Fer de l'Est,1 containing a detailed account of the systems of warming carriages on the principal European railways, comprising upwards of 60,000 miles, with the results of experiments made on the Eastern Railway of France. A great diversity of opinion exists amongst railway officials as to what constitutes the best system of warming, though it is acknowledged that in England, the common foot-warmer is, by general consent, accepted as satisfactory.

The following is a résumé of the cost of the different systems

1 Vide "Le Chauffage des Voitures de toutes classes sur les chemins de fer." par L. Regray, Ingénieur en chef du matériel et de la traction de la compagnie des chemins de fer de l'Est. Publié par ordre du conseil d'Administration avec un atlas de 31 planches. Paris, 1876.