maintenance would be out of all proportion to the amount of traffic.

In the case of a line destined specially for merchandise traffic, and on which the speed can without inconvenience be reduced to 6 miles per hour, M. Le Chatelier, referring to the fact that rack rails have been used in England, and in one instance at least in the United States, considers that such a rail laid between the ordinary rails, and into which should gear a toothed wheel driven by the engine, could with advantage be employed. With engines of the general dimensions above stated, driving a pinion 133 inches in diameter, a tractive power of 8.3 tons would be developed by each locomotive, so that one engine could take a load of thirteen wagons up an incline of 1 in 20. For passenger traffic he proposes that the ordinary engines should remain attached to the train, and that a rack rail engine should be employed as a bank engine, and placed at the rear of an ascending or at the head of a descending train. In the latter position it would form a perfect break. Such an engine should be fitted with small auxiliary cylinders, acting on the ordinary wheels, so as to enable it to be readily moved about on those parts of the line where the rack rail is not laid down.

M. Le Chatelier remarks that the principles he has enunciated do not appear to have been understood by the engineers who settled the terms of the Semmering competition, nor by the constructors who furnished the engines; the locomotive Bavaria' and its rivals having been made with wheels of too great diameter, and the whole of the working gear having to be proportionately heavy in consequence. The Bavaria' possesses less tractive power than a pair of such engines as have been described, although weighing half as much again. The two engines could be coupled back to back, and driven by one engine driver, one assistant driver, and one fireman.

As regards cost of working, the Author states that in 1857 the expense of locomotive traction on French railways was given by the formula P 0.475 + in which P is the cost in francs per

=

F

100'

kilomètre, and F the price of coke in francs per ton. Considering that in the case of the engines proposed, for working heavy gradients, the wheels will make twice as many revolutions per mile as those of ordinary engines, M. Le Chatelier assumes their working expenses to be double, and, with coke costing 40s. per ton, he calculates the expenses at 32d. per mile, or a maximum of 40d. per mile, including interest on the cost of the engines, and the exceptional repairs which bank locomotives require when working under unfavourable conditions.

Altogether, M. Le Chatelier considers that whatever system is ultimately adopted for working a mountain railway, locomotive haulage should first be tried. The locomotives thus provided are

1 If P be taken as the cost in pence per mile and F the price of coke in shillings F per ton, the expression becomes P = 7.6 + 5.27'

useful during the formation of the line; they will serve for carrying on the traffic in the event of any stoppage of the stationary engines, supposing these to be ultimately provided; while, besides this, when not required on the mountain section, they can be turned to useful account on other parts of the line. For the traffic over a line between Modane and Susa (with a tunnel through the Mont Cenis), as laid out by the Chevalier Maus, with gradients varying from 1 in 100 to 1 in 28, over a length of 22 miles, and a gradient of 1 in 52 6 for a length of 7 miles in the great tunnel, it is considered by M. Le Chatelier that three passenger trains and four goods trains each way per day will suffice, and that these trains could be so arranged that in the case of only one of the passenger trains each way per day would it be necessary to run a "light" engine over the line. For working such traffic the locomotive expenses are calculated by M. Le Chatelier at £33,490 per annum, and it is considered that the work could be done by twenty-four locomotives, of which the cost is estimated at £50,000. If the line were worked by fixed engines and ropes on M. Maus's system, ten fixed engines, costing with the ropes £20,000 each, would be required, while the cost of maintenance of the cables is estimated at £320 per mile per annum, and the cost of repairs of engines, enginemen's and breaksmen's wages, &c., at £1,600 to £2,000 per annum. If to these expenses be added the interest on the capital expended on fixed plant, it will be seen on which side the advantage lies, even if the fixed engines be worked by water, and the motive power be thus furnished gratuitously. The atmospheric system being even more costly, has a still more doubtful application.

In conclusion, M. Le Chatelier considers the Piedmontese Government should build a pair of locomotives of moderate weight, with very small wheels, such as he has recommended, and also give a fair trial to the railway with a rack.

W. H. M.

Common Error in ascertaining Locomotive Adhesion available for the Traction of Trains.

By J. MOSCHELL, Engineer in Chief of the District Railway of the Jura. (Annales du Génie Civil, March 1874, pp. 145-149.)

The Author states that ordinarily, after having determined the total tractive force which can be produced by the adhesion of any particular locomotive on the rails, engineers deduct therefrom a certain proportion, as being required for the locomotive itself, and treat the residue only as available for overcoming the resistance of the tender and of the other parts of the train. The Author is of opinion that engineers make this deduction on the basis that the friction between the wheels and the rails has to overcome the resistance of all the moving parts of the

engine, and he enters into elaborate arguments with illustrations to show, that it is the steam which has to overcome this resistance, and that the adhesion of the engine is not called upon to play any part in the matter.

Further, he directs attention to the fact that by coupling a second pair of wheels in a locomotive, so as to turn them into drivers, not only is the adhesion available for traction increased by the effect of the weight upon the other pair of driving wheels thus brought into play, but that the adhesion formerly employed to overcome the journal friction of these wheels is no longer necessary. By assuming a proportion between the diameter of the wheel and that of the journal of 7 to 1, and a load of 10 tons upon the pair of wheels, with a friction on the journals of 2, he proves that 160 lbs. out of the whole adhesion, required to overcome the friction of the journals, are now set at liberty for the purpose of assisting in drawing the train; and he attributes to this fact the explanation of a matter observed by M. Flachat, viz., that the adhesion of two pairs of coupled wheels was a greater percentage of the insistent weight than that afforded by the adhesion when one pair of driving wheels only was employed. M. Flachat explained this discrepancy by assuming that the wheels were not truly of the same diameter, and that thus there was a slight grinding action which increased the adhesion. The Author, however, believes that he has found the solution in dispensing with the journal friction of the one pair of wheels.

B.

Locomotive without Fire. By M. S. PICHAULt.

(Annales Industrielles, June 14, 1874.)

Among the applications of mechanical force to the traction of vehicles on tramways or roads, a fireless locomotive has been employed by Mr. Lamm on the New Orleans tramway, since the spring of 1872, to which he gives the name of thermospecific engine. It consists of an ordinary steam-engine mounted on the tramcar, or on a separate truck, with a boiler having no furnace, and therefore smokeless and less liable to explode. This locomotive is supplied with water from certain stationary boilers along the route, heated to a temperature corresponding to 12 or more atmospheres of steam pressure. As this heated water gives off steam to the engine, its temperature, and the corresponding pressure of the steam, continually diminish, until a new station is reached and a fresh supply of hot water taken in. In order to judge of the quantity of work which such a boiler can give out, the Author obtains from the principles of thermodynamics :

T = 90,000 V (t) in French units

V being the capacity of the boiler in cubic mètres or feet;

T, the work produced during the falling of the pressure in kilogrammètres or foot-pounds;

to and t1, the initial and the final temperatures in degrees centigrade or Fahrenheit.

Modified, to allow for losses by radiation, conduction, leakage, &c., the formula is given as

=

As an application of the above, let to 190° cent. or 374° Fahr., which corresponds to a pressure of 11 atmospheres, and t1 = 153° cent. or 3070-4 Fahr., which corresponds to 4 atmospheres, then

Tu = 22,500 V x 37 =

830,000 V,

or Tu = 2,500 V × 66·6 = 166,500 V.

That is to say, each cubic mètre of water under these conditions can furnish 830,000 kilogrammètres of work, or each cubic foot can furnish 166,500 foot-pounds. If the journey lasts for an hour, this is equivalent to about 0.08 HP. (English) per cubic foot, or about 3 HP. (French) per cubic mètre of water. The experiments at New Orleans are examined in the original Paper by this formula, and are found to agree with it.

Though the application of this source of power is comparatively easy on tramways, it is less so on railways, because a boiler, large enough to hold the water usually carried by a locomotive and tender, would scarcely contain sufficient to produce a motive power of 60 HP. (French); while ordinary locomotives attain almost 600 HP. To diminish the size of the boiler, the distance between the replenishing stations may be shortened; but there will remain the inconveniences inherent to the variation of pressure within great limits, and to the variation of adhesion.

For the weight of steam p, formed while the temperature passes from to to t1, the Author obtains the formula :

-

Adopting the same temperatures as before, μ = 64 V or 4 V. That is to say, each cubic mètre of water has given off 64 kilogrammes of steam; or each cubic foot has given off 4 lbs., or about onefifteenth of its weight. The number of thermal units to be taken from a stationary boiler, to recharge the locomotive when the temperature has fallen from 190 centigrade to 153° centigrade by giving off steam, is the diference of thermal units in the boiler under these two conditions of temperature. It must be

remembered that, besides the fall of temperature, 64 kilogrammes of water in the form of steam have been abstracted per cubic mètre. It is shown in the original Paper that about 48,000 French thermal units have to be supplied per cubic mètre of capacity of the thermospecific boiler, or 5,270 English thermal units per cubic foot.

The restoration of lost thermal units is effected in practice by connecting the stationary and locomotive boilers, and allowing the heat to pass from one to the other, taking care to have a proper arrangement of level between them, and to have the stationary boiler much larger than the locomotive boiler, as the temperature of the former will otherwise be considerably lowered by the abstraction of so much heat.

It is calculated, in the original Paper, that if compressed air be stored up instead of hot water, each cubic mètre of compressed air between 11 and 4 atmospheres of pressure can only give out 24,500 kilogrammètres of work; while, as shown above, each cubic mètre of water gives out 830,000 kilogrammètres, or thirty-eight times as much. With fifteen thermospecific engines an economy was effected of 50 per cent. over the use of horses.

S. D.

On the Tendency of the Reversing Lever of Locomotives to “return suddenly" when being pulled over. By A. BALGUERIE.

(Bulletin de la Société d'Encouragement, Feb. 1874, pp. 73-85, 1 pl.)

gear, or

Referring to the well-known tendency of the reversing lever of a locomotive, when disengaged, to fly into full gear, more particularly when the engine is reversed with full steam on,' M. Balguerie investigates the reaction caused by the obliquity of the link with respect to the valve-spindle, as the cause of such tendency, and calculates approximately the force with which the reversing lever is urged to move. He selects for investigation the valvedistribution,' of three classes of locomotives on the Midi railway, fitted respectively with Stephenson's link, Gooch's link, and Allan's link. These are well-known types, of which, in the first, the expansion-link is shifted vertically, whilst the valveblock is maintained in a fixed centre-line; in the second, the link is suspended from a fixed point, or is stationary, whilst the valverod link is shifted vertically; and in the third, both the expansionlink and the valve-rod sling are shifted vertically, in contrary directions, being suspended from the ends of a double lever worked by the reversing handle. M. Balguerie assumes, for simplicity, that

6

This relates to the practice on many continental lines of reversing the engine with full steam on, to obtain, on the Le Chatelier contre-vapeur' system, the arresting of the motion of the tain. To absorb the heat developed in the cylinders by this reversal an injection of water is provided.—SEC. INST. C.E.