coils, thus energizing the magnet sufficiently to enable it to release the catch. Although the winding of the magnet is in parallel with the neutral fuse, it does not in any degree affect the carrying capacity of the latter, as the resistance of the coils is great relatively to that of the fuse. The opening of the switch follows so closely upon the blowing of the fuse that it is said to be practically impossible to distinguish the interval which separates the two occurrences.

L. J. S.

Long-distance Transmission Experiment at Ogden.

(Electrical Review, New York, vol. xxxii., 1898, p. 287.)

This experiment was carried out at Ogden, Utah, in order to ascertain the limits within which high-voltage currents may be used commercially. The experiment was made by F. O. Blackwell over the lines which connect the power station at Ogden with the distributing circuits at Salt Lake City, the complete transmission circuit being 73 miles long over three No. 1 wires. The amount of power transmitted amounted to 1,000 HP., and the transmission voltage at times reached 30,000 volts. The current on the return was delivered to resistance-vats at the power-house, consisting of three wooden tanks. This power was transmitted with a loss of only 9 per cent., including 4 per cent. loss in the two sets of transformers. Continuing the experiment, part of the Salt Lake City station load was run from Ogden with current at 24,000 volts. This was supplied to about 500 HP. in synchronous motors and lights for two days under severe climatic conditions without the slightest hitch.

St. Anthony Falls Water-Power Plant.

(American Electrician, vol. x., 1898, pp. 185-192.)

W. G. R.

These works were recently constructed by the St. Anthony Falls Water Power Company, and on completion were leased to the Twin City Rapid Transit Company, to furnish power for street railways in Minneapolis and St. Paul, supplanting the steamdriven machines at first installed. The site of the generating works is in Minneapolis, just below the St. Anthony Falls on the Mississippi. A V-shaped dam, with a crest about 1,000 feet long, has been constructed of cut granite, and gives a maximum head of 22 feet; it is provided with sluices, waste weirs, etc., of which a full description is given. The power-house, which forms part of the dam, contains room for ten 700 kilowatt sets, seven of which are now installed, and two smaller sets for supplying exciting current. The main turbines consist of two pairs of 42-inch

horizontal water-wheels on the same shaft directly connected to a generator; the sluices for each set are regulated by a governor driven by belting off the generator shaft. In the generating station there are already installed five three-phase alternators yielding 700 kilowatts at 3,450 volts with frequency about 35; two 700-kilowatt direct-current dynamos feeding a 500-volt railway circuit direct; and two 100-kilowatt 6-pole exciters. There are three substations: to No. 1, about 1 mile distant, current is transmitted at 3,450 volts and actuates two 600-kilowatt rotary converters; these are of the 8-pole type, giving 580 volts on the direct-current side at 530 revolutions per minute. No. 2, situated about 4 miles from the main station, contains one rotary converter of the same type supplied at the same voltage. No. 3, ten miles distant, is supplied at 12,000 volts, and contains two similar converters. Twenty-one 233-kilowatt transformers are installed, six at the main station and fifteen in the substations. The transmission is by paper-insulated lead-armoured triple conductors laid underground, two leading to substation No. 1, and one each to the other substations. The Paper is illustrated by eighteen views of different parts of the plant.

M. G. W.

Booster System in Electric Railways. J. L. WOODBRIDGE,

(Journal of the Franklin Institute, vol. cxlv., 1898, pp. 374–385.)

Booster is any electro-magnetic generator whose armature is connected in series with a feeder with the object of compensating for drop of voltage in that conductor due to the current which it is transmitting. In lighting and traction work it has proved very useful, rendering it possible to take up loss in feeders too long to form part of the main system. In every case of heavy drop in feeders, boosting involves difficulties in keeping uniform voltage along the whole line. It is practically equivalent to a very strong over-compounding, reinforcing the system only in the needy branches. Any spare generator in the station can be used for the purpose. A "compound series booster" is simply an ordinary compound-wound constant potential machine, connected by a double-throw switch, to serve (1) as a booster when required, or (2) as an ordinary generator. The shunt-coils are connected in parallel with each other and with a certain portion of the feeder, so that a small but constant percentage of current circulates through them, and, like the series coils, the magnetizing force is proportional to the load, obtaining thus the series effect from both shunt and series windings. The variation of voltage is carefully gone into, and its causes inherent in the booster machine. (1) Saturation of field-magnets. When using a machine on hand with this fault, remedy it by limiting the load, or providing

sufficient copper in the feeder to limit the maximum voltage required and range of magnetization. (2) Hysteresis effects have little importance in railway work owing to the peculiar nature of the variations of load. (3) Sluggishness, due to inductance and Foucault currents. The effect of high inductance of shunt-fields is rather more serious, and this is minimized by limiting the extent of their magnetizing force by dividing them up in parallel, thus tending to check the portion which will lag. As a rule it is not advisable to operate the booster continually, but to put enough copper to carry ordinary loads satisfactorily, reserving the booster for times of excessive traffic. But with a low price of fuel it pays to use a booster. for as much as 18 hours a day. The booster system, if used with skilled judgment, may save a large expenditure in copper at the cost of an amount of wasted energy that is well within the limits of economy.

P. D.

Accumulator Traction.

(Electrician, vol. xli., 1898, pp. 9, 10.)

An abstract is given of a report drawn up by a deputation of the Blackpool Corporation after visiting various Continental systems of electric traction. The estimated cost is given of supplanting the present conduit system by accumulator traction. The rolling-stock consists partly of double-bogie cars weighing 11 tons and partly of smaller cars weighing 7 tons. The relative costs of maintenance work out to the following figures:—

(Mechanical Engineer, vol. i., 1898, pp. 520-531, 566–568, 607–609, 628-629.)

A variety of forms of storage-cells-both obsolete and in present use-are described, and some elementary data given concerning the calculation of internal resistance, efficiency, and so forth. Many particulars as to the construction and working of the electric cabs now running in London and in New York are also given. The former carry forty Faure-King cells having a total weight of 14 cwt. and an output of 170 ampere-hours when discharging at a 30-ampere rate. One motor of the Johnson-Lundell type is used, capable of normally developing 3 HP., and provided with

double windings on both fields and armature for the purpose of obtaining the various speeds. The latter contain forty-four threeplate cells of the chloride type, whose total weight is 900 lbs. and capacity 100 ampere-hours at a 21-ampere rate. Two motors, each of 1.5 HP. at 800 revolutions, are employed.

E. J. W.

Electric Traction. G. PELLISSIER.

(L'Éclairage Electrique, vol. xiv., 1898, pp. 449–456; vol. xv. pp. 63–67, 140–14 3, 187-189.)

The Author gives an account of various patents relating to electric traction. Among the number are included the Blackburne and Spence system of traction, the Priest and Merrick pneumatic series-parallel controller, the MacElroy truck, the arc-trolley of the Industrie Électrique Company, and that of Siemens. Some account is given of the New York conduit-lines. The Author then describes the conduit systems of Griffin and Small, Hecker, Allen, Nigel, Harington, Balfour, and Smith, in which a small trolley within the conduit is magnetically attracted by the car and thus caused to run with it, making alive the various sections of a rail on the surface; of Siemens, and of Arno and Caramagna.

W. R. C.

Counterweight System in Electric Cable Tramways.

(Street Railway Review, vol. viii., 1898, pp. 286-288.)

The grade of the tramway begins at 5 per cent., increasing gradually to 16 per cent., then falling to 4 per cent. The vertical height attained is more than 100 feet. In going up- or down-hill, each passenger-car is controlled by a grip-car immediately below it on the grade; the grip-car is attached by a gripping device to the cable, and a counterweight running in a tunnel half the length of the tramway balances the weight of the cars. There are thus no counterweight cable fastenings and draw-bar connections. A number of drawings are given illustrating the mechanical construction of the line.

A. S.

Electric Street-Railways in Baltimore. C. B. FAIRCHILD.

(Electrical Engineer, New York, vol. xxv., 1898, pp. 569–571.)

This is a description of the Baltimore City Passenger Railway, which is about to be transformed into an electric line. Formerly the cable cars were operated in trains, consisting of an open gripcar and trailer. The train system still remains; but the open grip-cars have been converted into trailers, and the closed cars used for the grip-car.

W. G. R.

Best Arrangement of Return Feeders for Traction.

F. NATALIS.

(Street Railway Journal, vol. xiv., 1898, pp. 277-283.)

To tramway engineers the arrangement of the return feeders is an important question. The rails alone are seldom of sufficient carrying capacity for returning the current to the power-station. Large potential differences between two points on the line are not admissible, on account of so-called "vagabond currents," which disturb telephones and scientific instruments, ruin pipes, etc. The local authorities usually insist upon a definite maximum fall of potential on the return circuit. Return feeders are at best expensive; and if their number and locations are unwisely chosen, their cost of construction is considerably increased. The Author gives a number of formulas and examples for settling this important question, and concludes this article by stating that, considering what great expenses are often involved, it is well worth the trouble to make use of such formulas. In these calculations, as well as in many similar ones, many things must of course be left to the judgment of the engineer.

L. J. S.

Rail-Bonding. J. R. CHAPMAN.

(Street Railway Review, vol. viii., 1898, p. 347.)

A cast-welded joint in a 56-lb. rail, which had been in use for 2 years and was worn out, was tested, with the results tabulated below, showing that the conductivity of the joint was practically equal to that of the rail. The joint was afterwards sawn through; a view of the section is given.