switch and the current equaliser are to be used with the same series of cells, the switch portion of the equaliser may be mounted on the resistance frame; and thus a single resistance coil be made to do service for both instruments. This combination, of

FIG. 8.

which the connections are shown in fig. 9, while allowing the current equaliser switch to control the full range of the resistance, at the same time causes the whole coil to be thrown into the battery circuit upon the operation of either the overload or overdischarge switch. This, of course, diminishes the brilliancy of the lamps, and thus indicates the state of affairs to those inte rested.

FIG. 9.

In applying accumulators to plants operated on the three-wire system, a simple extension of the method already described is adopted, except in the case of small stations requiring only one series of accumulators, and where the working circuits can be multiplied during the light load, so that the ordinary method will suffice. In larger plants the batteries are installed in pairs, each series having the same apparatus as before; and as this has already been fully explained, it will only be necessary to add that,

P, E.

[NOVEMBER 29, 1889.

turned to their lower contacts, and the switches, s, to their uppe contacts, the current from the dynamos will divide at the swit S1, part going into the lamp circuits, and part through batteries, which thus receive a charge. As in the multiple s system, the potential of the dynamos being assumed to be rased above the normal to meet the requirements of the batteris, i order to suitably reduce the pressure on the lamp circuits, iti only necessary to adjust the resistance of the pressure equalize, In central stations where the feeders are already provided with pressure equalisers the connections are still more simple, t the batteries are merely connected to the omnibus wires, while i other connections of the ordinary three-wire station are unchanged In such cases, the increased pressure at which the dynamos an operated during the time the batteries are charging is reduced the lamps to the proper point by the usual adjustment of the feeder or pressure equalisers. During the charging period regulation of the accumulators may be dispensed with by opera the switch, s1, when both batteries will be connected in a sing series to the full pressure of both dynamos. In addition to apparatus shown in the diagram, each series of cells is providel with a double plug switch and sockets suitably disposed, by ma of which the relative positions of the batteries on the two sides d the system may be changed by simply transposing the plugs from one socket to the other. The object of this arrangement is to pro vide means for compensating for the unequal discharge of the batteries when the two sides of the lamp system are unbalanced, and the transposition is never made oftener than once each day, h all other respects the manipulation of accumulators when applia to three-wire systems is practically the same as when they are operated on the multiple arc plan. To increase the accumulater capacity of three-wire stations, double batteries are added n parallel to the first set, jnst as additional dynamos would be.

It was explained further back that the extra, or regulating cells, in each series often became charged sooner than the remainder d the cells, and required to be removed from the circuit before th others. This operation, of course, calls for some labour; and, little as it is, it would still be desirable to have the number of cells in series remain fixed both during charge and discharge. By the e of what are called counter electromotive force cells, this res may be effected, and at the same time the pressure equaliser be d pensed with. These very simple cells are made like an ordinary Planté accumulator of plain sheets of an inoxidisable lead all and without active material. When a current is passed throngi them they act as gas voltameters, and while they instantly oppos a counter electromotive force of about two volts, they are a capable of producing a current of any appreciable amount r duration on account of their inoxidisable property. The use counter electromotive force cells in place of equalisers is adva tageous in several ways, for not only may their internal resistan be made so small as to be practically negligible, but their count electromotive force is as effectual in reducing excessive press as a dead wire resistance, while possessing the unequalled adra tage that the fall in potential of the current passing through the is unaffected by any variation in the strength of such current The method of using these counter electromotive force cella shown in fig. 11, which represents a single series of accumulator installed in connection with a direct lighting plant operating an dynamo, as first described, and for simplicity and effectiveness the method cannot be well exceeded. The dynamo is connected t the lamp mains, as before, through an ammeter on one side, and

in adapting two series of cells to the three-wire system. the two batteries are connected in series in the same manner that the dynamos are connected. The general plan of this arrangement is shown in fig. 10, from which unimportant details have been omitted. At the top of this diagram the three horizontal lines M (+), M (+), and M (-) represents the omnibus wires in the dynamo room, from which the feeders, L, are led. The three vertical lines, o (+), 0 (±), and o (-), are extensions to the omnibus wires to which the dynamos are directly connected in the usual manner when the levers of the switches, s1, are on their upper contacts. The batteries are also connected with the omnibus wires in a sir way, so that when the switches, 82, are closed on their lower cuite

ets the batteries will also discharge into the lamp cirllel with the dynamos. Now, if the switches, 81, are

through the upper contact of a two-way switch, s, on the other while the battery, B, is similarly connected to the lamp mains one side, but on the other the circuit is completed through severs counter electromotive force cells, K,, the number opposed bei governed by the position of the cell-regulating switch, c, s. Whet the lever of switch s is on its lower contact, the charging circuit completed in the now familiar manner. The action of a pls arranged in the above way is as follows:-assuming the lever the cell-regulating switch to be turned off, or open, as the expre sion goes, and that the dynamo switch is on its upper contact, the dynamo alone supplies current to the lamp circuits. If, now, lever of the cell-regulating switch is turned to its left ha contact plate, the battery will be similarly supplying current the lamp circuits in conjunction with the dynamo, or alone

th

NOVEMBER 29, 1889.]

The dynamo switch is opened. If the dynamo switch is now urned to its lower contact, the dynamo current will divide at the point where it connects between the battery and counter lectromotive force cells, part going through the battery and back o the dynamo through the (-) lamp main, while the remainder vill pass through the counter electromotive force cells to the lamps. Thus, while the full pressure of the dynamo current will be effective at the battery terminals, its pressure at the lamps will be less by 2, 4, 6, &c., volts, according as 1, 2, 3, &c., counter elecromotive force cells are opposed to the passage of the current nto the lamps by the position of the lever of the cell-regulating switch.

When the battery has been fully charged and it is desired to stop the dynamo, the latter may be disconnected by opening the switch, 8, when the battery alone will maintain the lamps. As previously explained, the potential of the battery will be a maximum immediately after it has received a charge, so that in this case if the battery has the usual allowance of extra cells its potential will be higher than the lamps require. More or less of the counter E.M.F. cells may now be inserted in the lamp circuit, however, until the potential is suitably adjusted. Although during the greater part of the discharge the E.M.F. of the battery will remain fairly constant, if discharged to its limit the potential will slowly fall towards the end, and this fall must be compensated for by removing one or more of the counter E.M.F. cells from the lamp circuit.

It is to be noted that by means of the above method, the number of cells in the battery proper is fixed and unalterable during both the charge and discharge, and that whenever the

623

dynamo is usually of high voltage and of moderate current capacity; machines of 500 volts and 15 ampères to 1,200 volts and 40 ampères may be employed in small stations, while in larger ones the latter machines and their circuits may be multiplied. For charging purposes shunt wound dynamos are preferable on account of their non-reversibility, but even series and compound-wound machines may be used if suitable precautions are taken. The arrangement of the accumulators and regulating apparatus in each sub-station on the half direct system is essentially the same as the method adopted in the last described multiple arc system. The connections of two sub-stations installed on this plan are shown in fig. 12, in which station one is shown with its switches in position for both its battery and lamp circuits to receive current from the charging circuit; while station 2 is disconnected from the charging line, which is closed outside of it, and the batteries alone supply the lamp circuits. The switch which is employed to throw the batteries into and out of the charging circuit is of the snap action type, and is sometimes called a consumer's switch. It maintains the integrity of the charging circuit when the batteries are removed from it, and provided with a spark coil which prevents the opening of the charging line and the consequent formation of the injurious arc which results on the interruption of high potential circuits. A modified practical form of this switch is illustrated in fig. 13. As shown diagrammatically in the sketch, it consists, in the main, of four terminal contact springs which bear against an insulated cylinder in which are imbedded two rows of metallic contact strips. In one position the two left hand and the two right hand springs are brought into electrical contact, but if the cylinder is rotated slightly the two middle springs will be

dynamo is supplying current to the lamps, its surplus current is always available for charging the battery; moreover, only one adjustment of the counter E.M.F. regulating cells opposed to the lamp current will be required. It is obvious that counter E.M.F. cells may be substituted with equal effectiveness for the equalisers and regulating cells employed in all of the previously described systems, and it seems unnecessary to go into any further particulars in order to illustrate the practicability of such systems when properly installed and operated.

We have now to consider another branch of the electric lighting business, in which accumulators also claim a share of attention, viz. the supplying of current to lamps located at a considerable distance from the source of power.

The first method to be considered, is known as the half direct plan, in which one or more series of cells are installed in any number of sub-stations conveniently located near centres of lamp consumption. The charging station may be situated wherever economy dictates, without any special reference to the location of the batteries, for although the expense of the charging wire and the cost of the energy wasted in it cannot be neglected, both these factors must be offset by the saving effected from locating the power station in the most desirable situation. This half direct system is so called because during the period of maximum load half the current is supplied by the charging dynamo, and half by the batteries. At other times the dynamo may be simultaneously charging the batteries and supplying lamps, or the battery alone may be in operation. Each sub-battery station is, in fact, a small central station by itself, and it may contain one or several series of cells, the number, of course, depending upon the amount of lighting to be done in its vicinity. A common charging main, like an arc light circuit, passes through each sub-station, and each battery may be inserted in the main or withdrawn from it at will, exactly as arc lamps are cut in and out of circuit. The charging

624

reversal of the charging current, a principle which has already been applied.

While the method just described of locating the batteries in sub-stations near the lamps to be supplied, reduces the resistance of the supply wires to such an extent that the variation of potential with changes in current strength is unimportant compared with what it would be were all the lamps operated from one central point, under ordinary conditions, still a similar variation does occur from a different cause. We have already seen that the potential of a battery is higher while it is being charged than at other times, and that the greater the charging current the higher the potential of the battery becomes. It is evident that if charging commences when only a very few lamps are burning in a given battery, only a small part of the charging current will be required by the lamps, while the greater part will pass through the battery. This will raise the potential of the battery considerably, and as the lamp mains are permanently connected to its terminals, the lamps will receive an excessive pressure. As the number of lamps increases, however, more of the charging current will pass to them and less through the battery, the potential of

which thus becomes gradually reduced until it reaches a minimum, when the number of lamps burning becomes so great that the whole output from both the dynamo and battery is required to maintain them. The variation is in the opposite direction, of course, when the lights are diminishing in number. It is evident from the preceding facts that during the period when the charging dynamo is running and the lamp loads are varying, some regulation of the pressure at the lamp mains at each sub-station is required. This regulation is automatically effected by means of the apparatus shown in fig. 14, in which a solenoid magnet, A, R, and a polarised magnet, P, M, are inserted in the battery circuit. In all other respects the relative positions of the battery, lamps, and charging circuits remain unaltered from the arrangement shown in the last diagram. The operation of the apparatus in this sub-station will be as follows:-Assuming the charging dynamo to be delivering a current through the charging main, м,

AR

[NOVEMBER 29, 1889.

moved against its right hand contact, thus short circuiting the solenoid magnet and the counter electromotive force celle, as ind. cated by the broken line. For, as the sketch shows, the lever d the solenoid magnet is so balanced that when no current a traversing the magnet, its metallic contact rods all dip into the respective mercury cups, and so shunt around the counter elec motive force cells with which the latter are in electrical c nection. When the lamp load becomes reduced and the charging current is discontinued either by the stoppage of the dynamo, or by its being shunted past that particular battery by a movement of the switche s to the right, the battery alone works into the lap circuit, and the counter electromotive force cells are d circuited.

While the automatic regulator just described, when applied the half direct system, maintains the pressure at the lamps & ciently uniform for practical purposes, whether the charge dynamo is operating or not, and although the proximity of the battery to the lamps it supplies prevents undue variation of potas tial when the load changes; yet a perfectly automatic method d maintaining a constant potential at the lamp circuits, even whe the resistance of the leads is great, would be generally useful Such a method is illustrated in fig. 15, in which the battery, counter electromotive force cells, lamps and charging circuits occupy the same relative positions as before. The solenoid magnet of the regu lator is no longer in series with the battery, however, and is, in fact. entirely disconnected from it, while the polarised magnet is replaced by a second solenoid magnet, P, M, wound to high resistance. The armature lever of this second solenoid is so mounted that when à core is attracted or released this lever moves over a series of contac strips which are insulated from each other. Connected with these contact strips are coils of wire of suitable resistance, the whole corstituting a simple rheostat, more or less of the wire of which i included in a local circuit, according to the position of the leve which acts as a movable contact. The solenoid magnet of the regulator, A, R, and one or two cells of accumulators are also cluded in the local circuit referred to, the strength of the currect flowing in this circuit depending upon the amount of resistan inserted by the movement of the lever of the pressure magne P, M. The latter magnet is connected directly with the battery terminals, and as the potential of the battery rises and falls the current flowing through this magnet varies correspondingly. causing a similar variation in the current in the local circut Thus an increase of pressure at the lamps, which are connected t the battery terminals, indirectly causes the armature of the reg lating magnet, A, R, to be attracted, and this in turn opposes one or more counter electromotive force cells in the lamp circat until the pressure again becomes normal. If the pressure at the lamps falls below normal the regulator acts in the reverse way. cutting out the counter electromotive force cells until the normal pressure is again restored. If in connection with this method the polarised consumer's switch, previously mentioned, is substituted at s, the operation of such a sub-station will be entirely automati

90/0/0/0/0

M

FIG. 15.

and that the consumer's switch is in the position shown in the diagram, the charging current will pass from м (+) to the point q, where it will divide, part going through the lamps and the opposed counter electromotive force cells, K, to the point L, and thence to the line again, the remaining part of the current passing from point o through the magnets, P, M, and A, R, and out through the battery to point L and line. When the current is flowing in the direction of the arrow, as in the case just cited, the armature of the polarised magnet is moved to its left hand contact as shown, and the plunger of the solenoid magnet is more or less drawn down, according to the strength of the current. This action of the solenoid magnet causes one or more counter electromotive force cells to be inserted in the lamp circuit in such a way as to oppose their electromotive force to that of the charging current, while the full pressure of this current is available at the battery terminals. If now the number of lamps is increased to such an extent that not only all of the charging current passes through them, but also that more or less current from the battery joins in parallel with it; then the current from the battery will traverse

polarised magnet in the opposite direction from that pretaken by the charging current, and its armature will be

FIG. 16.

It is proper to mention that the pressure magnet must be very sensitive and requires a delicate adjustment, besides being som what costly on account of the excellence of the workmanship required. Although I have only seen the instrument used in s experimental way, its practical application to similar purposes ba given satisfactory revults.

Under certain circumstances, as, for example, when a lighting station is worked to its fuil capacity at night, but during the dayti has ample surplus power, the all-accumulator system may often be applied to increase the capacity of such a station with satisfactory

REVIEW

esults. For such cases the sub-stations of accumulators are ocated at distant points in the usual way, the cells being charged uring the day and discharged on the lamp circuits at night while he power plant is doing its regular work. The arrangement of he accumulators and regulating apparatus in the sub-stations of he all-accumulator system is practically the same as when the alf direct system is employed, except that a simple transfer witch which transfers the battery from the lamp circuit to the harging circuit, and vice versa, is used instead of the consumer's witch.

It is needless to say that all of the methods which have been escribed of employing accumulators in long-distance lighting may e adapted to existing lighting plants, and that when so adapted he earning capacity of such stations may be considerably inreased without extending the capacity of the power plant. For his purpose there will usually be required a special dynamo at he central station, a charging circuit taking in the territory outide of the regular lighting limits, and one or more sub-stations with accumulators.

There are, of course, numerous other ways of utilising accumuators in central station supply systems, such, for example, as the double battery method now operated in England by the Electrical Power Storage Company, in which the dynamos are kept running For 24 hours daily, duplicate sets of accumulators being alternately inserted in the charging and supply circuits at uniform intervals of time, by an automatic mechanical device. But I have limited his paper to a description of some of the simpler methods of accumulator regulation which may be readily adapted to ordinary lighting plants, and, in concluding, can only hope that I have succeeded to some extent, at least, in showing that when so applied in an intelligent manner, accumulators do occupy a useful place in the industry of electric lighting.

Through the courtesy of the Electrical Accumulator Company. I am enabled to present to the Institute this evening the first proofs of some cuts of a new line of standard accumulators recently brought out by that company. One of the cells referred to is shown in fig. 16, and the dimensions and capacities are descibed in the following table ::

16471. "Improved means or apparatus for reversing and regulating the brushes of dynamo-electric machines, partly applicable to electric motors." J. S. RAWORTH. Dated November 13. 11d. One or more pairs of brushes are carried by spindles or shafts projecting from a rocking frame or lever. Each of these spindles or shafts is provided with a small crank or lever arm having a projecting pin, and springs or weights are provided to keep the rocking frame and the spindles carrying the brushes in a middle or central position. A fan driven by the dynamo spindle, and provided with radial vanes or blades, is arranged to revolve in a casing

17930

"

TUCKER, 17974.

NEW PATENTS-1889.

Improvements in incandescent electric lamps." J. H.
Dated November 11.

"Improvements in apparatus for welding by aid of electricity, which improvements are also applicable for other purposes." H. H. LAKE. (Communicated by H. Lemp, United States.) Dated November 11.

18005. "Improvements in electro-magnets for electric signalling apparatus. G. BowRON. Dated November 12.

18010. "Improvements in lamp fittings for electric lighting." I. A. TIMMIS. Dated November 12. (Complete.)

18018. "Improvements in electro-magnetic brake and traction increasing system.' A. J. BOULT. (Communicated by E. E. Ries and A. H. Henderson, United States.) Dated November 12. (Complete.)

18028. 66

Improvements in moulding plates for electric accumulators." E. DE PASS. (Communicated by La Société Jacquet Frères, France.) Dated November 12.

also provided with vanes or blades, and filled with liquid. The fan, in revolving, tends to carry with it the casing which is supported in a bearing that permits of circular motion. A double ended lever in connection with the casing engages the pins carried by the herein before mentioned crank or lever arms. When the dynamo commences to revolve in either direction, the fan, acting upon the liquid in the casing, causes the double ended lever in connection with such casing to overcome the action of the springs or weights which hold the brushes in the middle position, and to force the brushes into contact with the commutator and in proper position according to the direction of rotation of the dynamo. As the speed continues to increase, the double ended lever overcomes the action of the springs or weights, and carries the rocking frame or lever forward sufficiently to adjust the lead of the brushes corresponding to the increase of speed. 1 claim.

CORRESPONDENCE.

The Zipernowsky Electric Tramway.

We have read in No. 625 of your esteemed Journal some observations relating to our electric tramway with vertical track, which induce as to reply as follows:

You remark, in the first place, that the manufacturing expenses of a tram line made on this system must be very high, especially with regard to the extremely strong construction of the channel, owing to the strong side thrust. We beg to remark in this point that this

626

side thrust does not exclusively and directly act on the masonry channel; it acts also on the strong iron sleepers in the channel, that will oppose, at all events, a sufficient resistance to the thrust.

As relates to the supposed expensiveness of such a tram line, we will soon be in a state to dispose of concrete facts, as we actually made, on our own territory, a proof line on this system, for a distance of about one kilometre, with ascensions, curves and automatic points. We may already mention at present, however, that we have made very particular calculations on the probable manufacturing cost of this line, and found that it will be relatively less than those of a line with horizontal track, having two rails laying on the pavement, and, besides, a channel with underground conduits.

As a further deficiency you mention that an obstacle on the rails will cause the car to tumble on its side, or, at least, to jam itself tight. To this objection we beg to reply, that these dangers, arising from the now mentioned eventualities, are by no means greater in our tram line than in any other electric line with subterranean conduit; on the contrary, it is supposable that with tram lines on our system such hindrances can be much easier laid aside, as the construction of the fixed connection extending into the channel is much stronger than that of the conduits of other electric tramways with underground leads. By no means, however, can this be considered as a defect that would make the system unpractical, or that could not be prevented by precautionary measures of a sufficient security; and we may note, that by means of clearers and brooms, fixed on the cars, accidental obstacles on the lines will easily be removed.

You noted furthermore that the faculty of making the cars very narrow performs no peculiar advantage of this system, as this can be done with ordinary cars too. This we concede within certain limits. But with tram lines with vertical rails there is soon reached a track-width that will be impracticable, as an unequal inclination of the rails would cause a very unquiet movement.

In regard to the pavement we may note that the advantages, mentioned in the description, are effectively peculiar to our system, it being indicatable that by the falling out of the horizontal track there must be spared a respective part of pavement work.

We

You doubt whether specialists and tramway engineers will look with favour upon this system. have pleasure in informing you that numerous specialists, tramway engineers, and delegates of local boards have visited and thoroughly studied the model of this tram line, shown at our factory, and have convinced themselves of the really important and practical advantages of our system.

We request you kindly to publish this reply in your esteemed Journal.

Ganz & Co.,

Eisengiesserei und Maschinen-Fabriks-Actien-Gesellschaft in Ofen Electrotechnische Abtheilung. Budapest, 22nd November, 1889.

Underground Mains.

Conduits for underground electrical conductors are generally supposed to facilitate the construction and repairs of an underground system, but such is not the case, as I will endeavour to show.

Nothing that is drawn in can long withstand the damp and confined atmosphere of a conduit, mixed as this air is with sewer and illuminating gases. If the conduits are made of wood, the acid coming from the wood itself contributes to the process of disintegration and decay.

When a cable or conductor in a conduits fails, in nine cases out of ten there are no cables or conductors ready to supply the place of the one pulled out, and it is necessary to wait for one to be manufactured.

This is the " draw in and draw out" system.
The Edison is different, and the conductor cannot be

NOVEMBER 29, 1889.

pulled out and replaced by another; but the fan when it occurs, is local, and can be located to an inch and remedied in as short or less time than it takes so get ready to pull a cable of 500 feet length out of a conduit.

The Edison Company in this city have located and removed or repaired over 300 faults, due to origina bad construction rather than to the system itself.

By the laws governing the currents the defective spot is ascertained, the earth above it removed, and the cable or conductor repaired.

The manner of locating faults with such exactness is known as the "loop-test," and the process is well described in page 226 in Culley's hand book, eight edition, and also by Kempe under the head of “Murray': loop-test," by which is found the sum of two resistances and their difference. Subtracting the difference froc the sum, half of the remainder is the resistance of the shorter length; which again, deducted from the sum gives the resistance of the longer length. The res tances are proportional to the lengths.

Faults in ocean cables are located by this mesra within a very few miles or fractions of a mile.

In ordinary city limits they are located within a few feet or fraction of a foot. When it comes to the dis tances between splice boxes or hand holes, the defect can be located to an inch.

A portable battery, galvanometer, and a modified form of Wheatstone bridge are used for this purpose.

The system with which I am identified is a draw in but not a draw out system; for this reason there ha been no occasion to draw out. If the insulation is inpaired or destroyed it must be from mechanical means, such as breaking or injuring the wrought iron pipes. which is proof against an accidental blow of the picks.

The pipe is placed far enough below the surface of the ground not to be disturbed in ordinary street repairs.

The material within the pipe is hermetically sealed from such agencies as destroy gutta percha, india rubber or anything else that is of the vegetable kingdom.

The oil we use in a pipe is a pure distillate, and no product of distillation undergoes change except for the better, when confined in air-tight vessels. Thus there is no deleterious change.

As before stated, the Brooks method is not a draw out system, for the reason that there is no necessity. Br a thousand feet of cable can be drawn out easier than five hundred feet could be drawn from any conda system, because the former is lubricated inside the pipe.

If perchance the insulation becomes impaired, a stated, it must be from some damage done to the pipe, enough to expose the conductors, as a mere hole in the pipe will not admit moisture, because the oil is much heavier than water. The defect must be local, and is position can be readily determined as before explained. and a pipe cutter, with a few feet of cotton covered wire, can easily repair the damage.

On the other hand, when a fault is detected in a cable that has been drawn into a conduit, every portion of it shows serious disintegration.

If a cable of our system is injured by mechanical means, the surface of the ground must be disturbed. which would be enough to indicate the exact spot of the "trouble" without the aid of the "loop-test."

The effect of confined air, such as contained in a conduit, has been known in England for at least 3. years.

A patent was taken out to preserve the gutta-percha covering on conductors by drawing the cables into iro pipes filled with water.

The English Post Office Department do this yet wherever it is practicable.

Johnson and Phillips adopted this plan in laying their mains for electric light installations until latterly: now they insulate with oil.

Telephone exchanges require large numbers of con ductors of small diameter. Five hundred of these conductors, properly insulated, can be drawn into a 3-inch diameter pipe, when oil is used. These conductors