198

We hope to obtain better results by employing a stronger motor, viz., one of the Gramme type, which gives off no sparks and has no dead point. For our purpose all that we require to do is to fit it with two attachments; one to ensure the starting, the other to regulate the speed.

A permanent current motor supplying the very variable power necessary for effecting the readings and giving at the same time the factor tirue, has the additional advantage, which is not without its importance, of a capacity of being introduced directly into the shunt circuit required for the wattmeter, so that a single shunt circuit is sufficient. These are the motives which guided us in the choice of the two principal parts of our meter; we have now to discuss the apparatus for registering and summing up the readings. But here we cannot start with general principles, all that we can say is, that it should be as simple as possible. We will explain this mechanism in the course of the full description of our apparatus, to which we will now proceed.

The Meylan-Rechniewski meter, as shown in fig. 1, consists of six principal parts.

[FEBRUARY 13, 1891.

watts the expenditure of energy of the circuit with which the meter is connected.

Constant Speed Motor.-The motor upon which we have finally decided is of a special type; fig. 2 shows an elevation and partial section explaining its working.

It is a disc motor, but one in which the two inductive poles are on the same side as the armature or above this latter; this arrangement produces, as we shall presently see, an actual raising of the induced core under the action of the magnetic field. We thus relieve the pressure on the stone on which the pivot rests.

The inductive system comprises two semi-circular polar pieces attached to two squares connected by two bolts which form the cores of the electro-magnets, E,, Eg.

The armature is a flat iron ring, insulated and wound round with 3,000 revolutions of fine wire, forming from five to six sections connected with an ordinary collector. This armature is set into an ebonite frame mounted on the axle, v, the endless screw of which sets in motion the system of wheels, r,, (figs. 1 and 3).

1. A permanent current electric motor, M (fig. 2), the speed of which is constant.

2. A system of gearing, reducing the speed of the motor and communicating to a final axle or to the principal axle, A (fig. 3), a movement which is continuous and strictly uniform, the speed of which is 3rd of a revolution per minute.

3. An electrodynamic balance, say a wattmeter consisting of one or two fixed bobbins, G1, G2, traversed by the principal current, and a movable bobbin, Gg, fixed to a beam, F, mounted on a pivot or on knife-edges; this balance gives rise, under the action of the principal current and the derived current, to a pressure proportional to E I, which pressure is brought into equilibrium by a fixed stop.

4. An elastic cam; this is the part which effects the readings of the balance.

5. A system of friction coupling, controlled by two ratchets, and communicating to the first motive portion of the counter, T, the movement of the principal axle during each reading of the balance. The action of the ratchets depends on the movements of the principal axle of the electric cam.

6. A counter, the various dials of which show in hecto

This arrangement is simple, and at the same time, easy to put together and take to pieces; all that is required for this is to unscrew a screw, take away the upper part of the column, and then remove the screw which keeps the hinder polar piece up against one arm of the support."

The axle of the armature fits into a socket, and a guide belonging to the framework of the mechanism (see fig. 3).

The system of regulation, of which we have spoken, consists in shunting periodically the bobbins of the armature. This result is very easily obtained by connecting one of the plates of the collector with the insulated slide which moves the conical pendulum that indicates the speed. One of the brushes, +, being connected with the ball of this pendulum, we can see that when the ball, B, touches the contact, C, part of the derived current passes out of the armature, for instance, when this plate is under ß, the current passes through X, C B, cß to the inductor 1, without passing I,

FEBRUARY 13, 1891.]

199

Thus, with a variation of 17.7 per cent. in volts, the variation in the speed was only 2.2 per cent.; within the practical limits of electric lighting, say with a variation in volts of 10 per cent., the maximum of error would only be about 1.5 per cent.

The sparks at C, B, are absolutely invisible, for they arise merely from the breaking of a short circuit traversed by 025 of an ampère; moreover the contact is always perfect, the spring, c, causing the revolution of the ball in one revolution of the motor; lastly, the current is neyer broken, as in the well-known apparatus of M. Marcel Deprez ; on the contrary, it may be considered as practically constant, if we are careful to give the armature a resistance which is slight in comparison with the entire circuit.

The object of the regulator is to ensure the rapid disconnection of the motor, which gives rise to a great increase in the derived current, as long as the ball does not leave its central support. This latter consists of a contact, c', connected with the axle of the motor, and consequently of the mass, and the wire, f, with the one extremity of a resistance, x. The other extremity of this resistance being connected with the column, c, we see that as long as there is contact between B and c', the resistance, x, is shunted, the derived circuit containing only the motor and the rheostat, R1. The motor then quickly assumes its normal speed, and we need fear failure at starting.

never

any

It is true that we observe a spark on the breaking of the contact, C, B, but as in practice this only takes place once a day, it does not cause any inconvenience. Of course a reading taken during the disconnection would be falsified, but this contingency may also practically be disregarded. In cases where the meter is required to work with considerable variations in the differences of potential, we employ a double slide, i.e., a slide formed of two parts connected with two diametrically opposite plates.

The ball, B, having reached the limit of its course, then comes between the two contacts in such a manner that two of the plates of the collector are continually at the same potenas one of the brushes. The diminution in the motive effort must be double what it is in the preceding case, and the regulation is much more perfect. In this way we have obtained the following results :

Electro-dynamic balance.-This part of the apparatus is shown to the left of fig. 1. The fixed bobbins, G1, G2, consist of strips of copper insulated and wound on a central core of wood.

The two bobbins are connected sometimes in parallel, sometimes in series, according to the different types, but in such a manner that their action is combined, and that the force acts in an upward direction.

The current is introduced through two brass blocks, pierced with holes, in which the cables are fixed and held by means of a tightening screw. The movable part is formed of fine wire, G, attached to an ebonite support, which is itself fixed to the beam by a metal ring riveted to the latter. This part is shown in fig. 3. The current is brought to the bobbin on one side through the pivots, and on the other through a very flexible strip of foil, which is fixed to an insulating support near the axis of suspension.

In the little model meter, the beam, F, is simply pivoted at p, p; but to obtain greater sensibility, we must adopt a system of suspension by knife-edges.

The movable bobbin is brought into equilibrium by a counterpoise screwed to the end of the beam.

This counterpoise should be regulated so as to be slightly in excess, so that even in a case where the electromotive force is nil, there is still a certain force acting upon the beam so as to raise it, this, however, be it understood, constituting a false zero which is taken into account in registering as we shall see.

The pull on the beam by the elastic cam is made by means of a piece of steel, π, attached to this beam.

The Elastic Cam.-The elastic cam (fig. 3) comprises the disc, D, which bears the steel cam, 1, the spiral spring, R, the axle, A, and a rod, L, which connects it with the principal axle, A. The axle, A, is brought to the extremity of A, between points on pivots. As long as the cam, 71, does not meet the piece, 7, of the beam, F, the spring, R, is not twisted, and the axle-tree, A, shares the movement of the axles. the moment of contact, on the contrary, the cam is stopped and the spring is turned at an angle proportional to the pressure which the beam originally exercised on its stop. At the moment when the cam raises the beam, the reading is finished.

At

Friction coupling and disengaging mechanism.-The movement being uniform, the total of the torsions gives the total of the periods of the reading. In order to obtain it we must couple the axle, A, with the first wheel of the counter, T, during the periods of torsion. For this purpose we have constructed a friction coupling, regulated by ratchets, one of which is worked by the movement of the principal axle itself and determines the real commencement of the reading, whilst the other depends on a catch actuated by the active force of the elastic cam at the end of the reading, and disengages the eling at this moment. Thus this arrangement requires the value zero, we should have an active force in

T

[FEBRUARY 13, 1891.

reserve, i.e., a certain torsion of the spring, R; the counterpoise, P, must exert a certain excess of pressure on the fixed stop of the lever, F, as we have said.

We effected in the following manner this periodical coupling of varying duration, the details of which are shown in fig. 3. A ratchet wheel of large diameter, r, is mounted, with slight friction, obtained by the pressure of grooved discs, at the extremity of the axle, A: it turns the first wheel of the counter, T, by means of a catch, c, which pushes a slide belonging to this first wheel. This wheel, r, is provided with two catches, q, 91, the first of which is always in position before the commencement of the reading and is liberated at the moment the reading begins; whereas the other, which is normally free, catches the wheel when the reading finishes.

With this object the first catch, q, which is attached to the axle, b, and the lever, 7, is worked by the cam, c, which is of such a form, and is held on A in such a manner that the catch is freed at the moment the reading begins, being brought back at the end of the revolution by the action of the spring, R2.

On the other hand, the ratchet, q,, is attached to the axle, b1, and to the lever, l, which is held in such a position as to restrain the action of the spring, R, by a notch in the lever, 12. When the reading is finished, the part 7, of the spring cam is released from the part of the beam. At this moment under the action of a previous torsion and the torsion corresponding to the reading, the disc, D, which is notched at one point of its circumference, raises the lever, 1, through m; the lever, 7, is released and the ratchet, q1, immediately catches under the action of the spring, R ̧.

At the end of the revolution a second cam, C, which is also held on A, and which presses on the lever, 7., fixed on b1, brings back the ratchet, q, and the lever, 73, stretching the spring, R1, again.

This last operation naturally takes place only after the first ratchet, q, has been brought into place by the cam, c, so that the ratchet wheel is not free until the next reading begins.

The regulation of the apparatus is effected simply by ensuring the simultaneity of the release of the ratchet, q, and the catching of the ratchet, 41, when the electro-dynamic pressure is nil; this regulation is performed by the scotching of the cam, c, and the adjusting of the weight, P.

Connections.-As we have said, the motor is wound in series and is included in the circuit of the fine wire of the wattmeter and a rheostat. The connections of the shunt circuit are therefore those of fig. 2, in which x represents a rheostat and R, the bobbin of the wattmeter. Each meter has four transmissions of current; the two first to the left which are insulated, correspond to the positive cable and to the positive wire of the lamps; they are connected by the large wire of the galvanometer.

The two others, forming a whole, serve to fix the negative cable and the negative wire of the lamps. Lastly the shunt circuit is connected by one end to one of the positive terminals and by the other to one of the negative terminals.

Calibration. The calibration of all our instruments is arranged that the first dial of the counter indicates a multiple or decimal sub-multiple of the unit of electrical work, the hecto-watt hour. Thus for the meters of the smallest calibre (100-110 volts, 12 ampères), one division of the first dial represents a deca-watt hour.

This calibration is effected, for a type once established, by modifying in the ratio required the total resistance of the shunt circuit which is about 2,000 w, the speed of all our motors being previously regulated so that the rotation of the principal axle takes three minutes.

Series of Types.-The calibre of a meter is determined by two conditions one of which depends on the galvanometric apparatus and the other on the extent of the measurements to be effected, i.e., on the number of lamps to be registered, if we admit that the meter commences to register at one lamp. In fact, if an apparatus is sufficiently sensitive to show within the required approximation from unity up to n times this value, all that we have to do is to adjust the galvanometer so that the minimum electro-dynamic effort corresponds to the energy consumed by each lamp, in order to obtain a meter for n lamps. If we wish to obtain types of smaller calibre we have only to modify the winding of the

FEBRUARY 13, 1891.]

galvanometer. But to come into practical use all th ́se galvanometers must cause only a very slight loss of

energy.

On first thoughts it seems that we should fix a certain limit to the loss of volts, 2 for instance, and give the apparatus a resistance in inverse proportion to the maximum current. Nevertheless, as the electro-dynamic effort is in proportion to the volume of copper and to the loss of energy, and as this increases, like the current, with a constant loss of volts, we may diminish this loss of volts in proportion as the calibre increases, at the same time obtaining more powerful efforts, a condition which is necessary in order to obtain great sensibility.

Thus the relative loss of energy diminishes with the calibre. On the other hand, in all meters in which the readings are not continuous, the sensibility or the range is always much greater in proportion as the dimensions of the registering apparatus and the efforts of the measuring apparatus are more considerable; we are thus led to increase the dimensions with the calibre of the meters.

These are the principles by which we were guided in the establishment of our series of types. Thus our first type of meter, the ratchet wheel of which has 300 teeth, enables us to measure from unity to 70 or 80; when furnished with a galvanometer the resistance of which is 01", it constitutes a 12 ampère type, and can measure from one lamp of 8 candle-power (30 watts) up to 40 lamps of 8 candle-power, or 20 lamps of 16 candle-power.

The principal constants of this type are :

It is easy to alter this type so that it will measure as much as 25 ampères by modifying the galvanometer, so that the maximum ampère turns remain constant, with the same weight of couple.

Our second type is capable of registering from one lamp up to 80 8-candle lamps or 40 16-candle lamps.

For higher calibres the dimensions of certain parts must be increased.

Fig. 4 gives the results obtained from a meter of the second type, the ordinates corresponding to the readings taken on the meter, while the abscissæ represent the actual power.

By examining the curve we see that with this meter which is intended to show the expenditure of 40 16-candle lamps, we can register the expenditure of the first lamp, but with a considerable error. After five lamps the indications are correct within about 2 or 3 per cent.

Thus the error in the total consumption will amount to about this percentage on condition that there is on an average a certain number of lamps lighted at once, which will always be the case in practice.

Measurements have been performed by the meter during periods of several hours.

Our meter can be adapted to various combinations which enable it to be transformed into a volt hour meter, and the principle of its construction enables it to be applied to alternating currents. In short, although the presence of the motor causes a considerable delay in disengaging the shunt current in comparison with the principal current, this does not cause the electric balance to give less accurate results as long as the frequency of the reversals is not changed; we have simply to modify the rheostat in each given case.

The only difficulty that presents itself is that on account. of this delay we require a very high potential in order to work our motors, although they do not appreciably absorb any more energy than with a continuous current.

We shall probably have occasion to return to this point when this difficulty has been surmounted.

201

THE GALVANIC BATTERY IN INDUSTRY. UNDER this somewhat sensational title, M. de Meritens, on the 3rd inst., made a communication to the Société Internationale des Electriciens, which was listened to by an audience not as a rule greatly interested in the proceedings of the Society. M. de Meritens's address produced the effect which might have been anticipated on the listeners who filled the hall, and on Thursday morning the most important political journals vied with each other in singing his praises, speaking of his paper as containing an important discovery, and implying that it solved the problem of supplying electricity at a cheap rate.

The figures given by M. de Meritens are too incomplete and have been presented in too vague a form for it to be possible to draw any conclusions whatever from them; we must therefore wait until we have a technical report drawn up by some authority, who, while being no less reliable, would at least be less interested than M. de Meritens, before we can give any opinion whatever as to the value of his idea, of which we can at present only sketch the bare outlines. In the first place, M. de Meritens claims to have succeeded in rendering the galvanic pile absolutely constant as a source of electricity, by employing, on the one hand, pure zinc, ard on the other, as the positive electrode, a plate of carbon perforated and covered on the side, facing the zinc, with a sheet of platinised lead, the exciting liquid being water mixed with sulphuric acid.

He placed before the Society a battery of small dimensions which, being short circuited, gave, during about an hour and a half, a constant current of 33 ampères. It is impossible for us to know what would be the output of this battery under normal conditions, or at its maximum power, over a given external circuit, or what would be the consumption of zinc and acid per effective horse-power hour or kilowatt hour in the external circuit, when the cell is under conditions of practical work.

However, this may be, we should obtain, according to M. de Meritens, electrical energy and hydrogen which could be collected and utilised by the expenditure of pure zinc and sulphuric acid for the formation of sulphate of zinc; the composite electrode of platinised lead and carbon remaining unattacked. The sulphate of zinc would then be treated electrolytically, plates of iron being employed as the anode, so as to reduce to a minimum the counter-electromotive force of decomposition of the electrolyte. The products of the operation, for which motive power is of course employed, would be on the one hand pure zinc, which would serve again for other elements, and on the other hand, sulphate of iron, the marketable value of which of course varies with the demand. We will complete this very vague description when we have before us the official text of M. de Meritens's communication, the industrial importance of which it is at present impossible for us to estimate. On the face of it, however, it would appear that the French electrician has been taking a leaf out of the book of his fellow-countryman Perreur-Lloyd.

TELEGRAPHS IN INDIA.

A SUPPLEMENT to the Gazette of India, reviewing the Administration report of the Indo-European Telegraph Department gives the history of the work and an account of the company's finances for the year just ended. The capital account of the department, it appears, is something over two lakhs in excess of what it was at the commencement of the year. Quite two-thirds of the amount are accounted for by the adjustment of loss on exchange on English stores for this and previous years. The message revenue is over ten and a half lakhs, a diminution of little less than 70,000 rupees as compared with the previous year, in which for thirty-two days the whole of the traffic along the East Company's lines was diverted to the Indo-European wires, owing to an interruption in the Red Sea. No such diversion took place this year; on the contrary, it was necessary for the company to repair 227 knots, as against 201 in the previous year. Ten faults have occurred during the twelve months, of which one is stated to have repaired itself, five were cut out, leaving four

202

still in existence at the close of the year. It is interesting to read that of the total number of messages despatched to England, 2,403 were State, 10,175 commercial and private, 3,104 Press; while of this class not less than 120, containing on an average 663 words each, were addressed to the London Times. The total is less than that of the previous year, and the percentage of errors, as might be expected, is slighter. The accuracy of transmission between Karachi and Teheran was great, there being only 60 complaints of error, delay, or non-delivering, as compared with 84 the preceding year. There was no delay so serious as to interfere with the receipts, no period under 24 hours being taken into account. The most prolonged interruption was the outcome of a flood in the Kinarigaird hills; the others by wilful damage. It would scarcely be believed by electricians in Europe that the substitution of porcelain insulators for those of iron has produced such unsatisfactory results that their abandonment is under serious consideration. Yet the fact remains that the more extensively the former are adopted the more frequent are the interruptions. Damage is of almost daily occurrence. This is on the Persian section, where the natives use them as targets "for testing the efficiency of their guns or their skill in throwing stones." To such an extent has this description of wilful damage been done, that not less than 10,500 kraus have been paid by the Persian Government and district authorities as compensation. Birds and camels continue to be sources of annoyance and causes of interference with communication. There may be no means of preventing this, but there seems to be some of minimising wilful damage. In November, 1889, on the Teheran coast, the wires were cut, and the extent of mischief reported was lamentable. The year shows that in average speed of transmission the Teheran route still leads, with a record of one hour and five minutes as against Suez, two hours and 21 minutes, while the Turkey route is a bad third with three hours and six minutes. The most important paragraph of the report is that which refers to duplex working on the gutta-percha cable. It is believed to be the first instance of the duplex method being applied with ordinary Morse instruments to a long sea cable. This experiment was made by Mr. Porsman, the electrician of the Persian Gulf section. Some experiments in automatic signalling, considered likely to have useful results, were also carried out by Mr. Preece, one of the superintendents of the same section, and two sets of Wheatstone's instruments have been indented for, with a view of giving practical effect to the system. The Government of India notices with satisfaction both of these experiments, which it considers do credit to the Department.

THE ELECTRICAL CONDUCTIVITY OF SALTS IN THE BUNSEN-BURNER GAS FLAME.

SOME experiments have recently been conducted by S. Arrhenius upon the electrical conductivity of salts in the flame of a Bunsen gas burner.

A burner of the ordinary type was used in these experiments, and the flame was fed with a mixture of atmospheric air, and gas, both being maintained at a constant pressure.

Before the current of air was allowed to mingle with the gas it was caused to pass through a fine spray of an aqueous solution of the salt whose properties it was desired to examine. The degree of concentration of this solution determined the amount of the salt which was carried by the air into the flame. Electrodes were formed by placing two small plates of platinum in the flame, at a distance from each other of 0.560 mm. A battery of forty Latimer Clark cells and a galvanometer of the Wiedemann type were employed in the circuit.

The conductivity of the flame was measured by observing the deflection of the galvanometer, since it was found that the resistance of the flame as compared with that of the remainder of the circuit was very great.

Arrhenius found that the following relation between the intensity of the current and the electromotive force, held good, namely, i = kf (e), in which equation = the elecFromotive force, i = the intensity, k = a constant, ƒ = = a

[FEBRUARY 13, 1891.

function which remains the same for all salts at all degrees of concentration. Ohm's law was not found to hold good in all cases for conduction in the Bunsen flame, the exceptions being at values below 0:2 Dan.

The conductivity of the salts in the flame was found to be proportional to the square root of the concentration of the solution; by the conductivity of the salt in the flame is understood the total conductivity, minus that of the flame itself. This law was shown to be practically true for no less than sixteen different salts. When the solutions were extremely dilute the conductivity was observed to increase much more rapidly.

All the salts of potassium conduct equally well, and this statement is also true, for the salts of sodium and lithium ; salts of hydrogen, ammonium, magnesium, zinc, cadmium, copper, iron, nickel, cobalt, aluminium, chromium, magnesium and tin do not appear to possess any conductivity. The salts of barium, calcium, and strontium conduct by the convection of small solid particles, and there is some reason for believing that they conduct electrolytically. The conductivity of the chlorides of the alkali metals increases with their molecular might.

Arrhenius also observed that the difference of potential between two metals in a Bunsen gas-flame which contains saline vapours is of the same sign and order of magnitude as when the same metals are placed for examination in some electrolyte. He concludes, that on the whole, when salts in the state of vapour above described, are caused to pass into a Bunsen flame, they behave like a feebly dissociated electrolyte, the chief point of difference being that Ohm's law does not hold good when the electromotive force has a large value. The original paper may be read in the Sitzungsberichte der K. Akademie zu Wien, No. 99.

ELECTRIC LIGHTING IN MADRID.

LAST week we briefly alluded to two events: the issue, or attempted issue, of £75,000 of 6 per cent. first mortgage debentures by the Electricity Supply Company of Spain, and the withdrawal of the contract from, and consequent stoppage of the Matritense Company, which has been for some time purveying electric light in Madrid. There is no connection between these two companies in any way that we are aware of, except that they have been operating in the same city, but that they have been such near neighbours leads us naturally to bracket them, for having the same locale the surrounding circumstances must necessarily be to some extent similar. And though we sincerely hope that history may not repeat itself in failure, the narrative of the birth, life, and demise of its unfortunate brother may not be without interest, but be a useful warning to those embarked in or contemplating joining Mr. Hammond's Spanish Company. As when we see a friend who has recently met with a family bereavement, and are in darkness as to whether expectations have been realised, and affliction thereby mitigated, so now we hardly know how much sympathy to offer the last-mentioned undertaking, or whether it should be of a condoling or congratulatory nature, for it is possible that where others sowed it may be reaping, the ill-wind which wrecked the Matritense may have blown it good, and as the latter is dead, we sincerely trust that the English Company may have a share of its inheritance.

The Matritense was originally established by a firm of electrical engineers in Barcelona, to whom were allotted for promotion money and for the plant and material which they supplied by far the greater proportion of the shares. The plant was old type and uneconomical in working, and the over capitalisation, the weight of dead shares the company had to carry, crushed it in its cradle-for the knowledge that profit, if obtained, must go nearly all into the pockets of others, is not conducive to the attraction of extra funds, and, as usual, sufficient were not provided in the first instance. Obtaining the first concession to canalise the streets of Madrid, they laid 500 metres of mains in the principal thoroughfare, the Calle d'Alcala, running from the Puerta del Sol to the Prado and the Parc de Buen Retiro. This effort appears to have exhausted their canalising capacity, for