current both in the copper plate and in the secondary spiral, but the current in the secondary spiral cannot induce a current in the main spiral, because the latter is not closed by metal, the two coatings of the jars being separated by glass. The only possible influence of the current in the secondary spiral upon that in the main spiral is some retardation of the discharge.

Now, if the closure of the secondary spiral is more perfect than that of the main spiral, the current of the former will pass more rapidly than that of the latter, and on that account no reaction of the secondary spiral can take place upon the main spiral; hence, with a more perfect closure of the secondary spiral, the temperature in the conducting circuit is found very little less than when no secondary spiral is present.

With an imperfect metallic closure of the secondary spiral the secondary current has a longer duration, and then the discharge current in the main wire finds, during its whole course, the secondary wire traversed by a current passing in the same direction, and we must assume that this is the cause of the retardation of the main current, which is indicated by the diminished temperature; by imperfect closure of the secondary spiral the température in the main current was reduced in the proportion of 0.43 to 0.17.

By inserting a tube of water into the secondary spiral the temperature again increases almost as much as though no secondary spiral had been present, which is well explained by the fact that, with very imperfect closure of the spiral, no sensible secondary current is generated.

The circumstance that, with quite perfect as well as with very imperfect closure of the secondary spiral, the influence on the main wire is less than for a moderately good closure, leads us to expect that, when the secondary spiral is closed by constantly increasing lengths of thin wire, at first the temperature of the main circuit will decrease, that, with a given length of the introduced wire, the influence of the secondary spiral will become a maximum, and then decrease again, and that, therefore, the elevation of temperature of the conducting circuit of the main spiral will again increase when the wire by which the secondary spiral is closed is lengthened.

This was verified by experiments which Riess made.—(Pog. Ann., LI, 177.)

Representing by 100 the temperature observed in the thermometre introduced in the conducting circuit of the main spiral, the secondary spiral being closed by a short thick copper wire, the results given by the insertion of a German silver wire 0.1517 line diameter and of different lengths, are as follows:

It is seen from this table how very rapidly at first the temperature of the circuit of the main spiral decreases with increasing length of German silver wire inserted in the circuit of the secondary spiral, and that a minimum is reached when the length of the introduced wire is 29.6 Paris feet, in which case the heating effect is only 48 per cent. of that which is observed with perfect closure of the secondary spiral. When the length of the wire exceeds 29.6 feet the temperature gradually increases again; and by lengthening the wire to 582 feet the temperature rises to 87 per cent. of that originally obtained.

A metallic closed circuit near the conducting wire of an electrical battery acts retardingly on the discharge of the battery in proportion to the length of its closing wire. The circuit of the secondary wire being progressively prolonged its action successively increases, attains a maximum, and then decreases.

The changes which the temperature in the main wire undergoes by lengthening the secondary wire, obey the law indicated by the last table, whether the charge of the battery be stronger or weaker; with stronger charges, as well as with weaker, the retarding effect of the secondary wire attains a maximum when the secondary spiral is closed by 29.6 feet of the above-mentioned German silver wire; and then the temperature in the main wire is 48 per cent. of that which would have been observed with an equal charge if the secondary spiral had a perfect metallic closure; but as soon as the conducting circuit of the main wire is lengthened by the introduction of a thin wire the course of the retarding effect of the lateral wire changes.

In the main conductor a platinum wire 7 inches 5 lines long and 0.023 line radius was introduced, and the results in the following table were obtained; the lateral spiral being closed by German silver wire of different lengths:

We see here that, on prolonging the main conductor, the maximum effect of the secondary wire is not reached until a greater length of wire has been introduced into the secondary spiral, and moreover that the retarding effect of the secondary wire is now much less. During the previous experiments the temperature of the main wire was reduced by the maximum effect of the secondary spiral to 48 per cent.; now, the maximum effect of the secondary spiral produces only a reduction to 78 per cent. of the temperature, which would have been observed either without the secondary spiral or by one perfectly closed.

This is easy to explain. The secondary current is stronger in proportion as the part of the main wire acting on the secondary wire is greater, and to the stronger secondary current we must also attribute a greater reaction upon the discharge. The length of the main wire was the same in both series of experiments, namely, 13 feet of copper wire, which acted upon the same length of the secondary wire. In the first series these 13 feet made by far the greatest part of the circuit of the battery; in the second a platinum wire was introduced, whose retarding power was equal to a copper wire 568 feet long and 0.55 line thick; consequently, in the last case, only about one-forty-fourth part of the virtual length of the main wire acted upon the secondary spiral. Riess caused two other spiral disks to be made, each containing 531 feet of copper wire two-thirds of a line in diameter. The large and small spirals were introduced into the main circuit.

The small main spiral being now placed opposite the small secondary spiral at a distance of 2 lines, the maximum retarding action of the secondary spiral took place when it was closed with 29.6 feet of German silver wire. With this maximum effect the temperature of the main circuit was 76 per cent. of that which was observed without the lateral spiral, or when it was perfectly closed.

When the large secondary spiral was opposed to the large main spiral at a distance of 2 lines, the maximum retarding action of the secondary wire occurred when the latter was closed by 79 feet of German silver wire, and in this case the temperature in the main wire was reduced by the retarding action of the secondary spiral to 25 per cent. Finally, the two secondary spirals, properly connected, being placed opposite the two main spirals, then 138 feet of German silver wire had to be introduced into the secondary circuit to obtain the maximum retarding effect, and the temperature in the main wire was thereby reduced to 20 per cent. of that which would have been observed without a lateral spiral. From these experiments it follows that—

The maximum effect of a secondary wire upon the electrical discharge attained by lengthening the secondary circuit is as much greater as the length of the main wire acting on the secondary wire is greater. But, at the same time, to attain this maximum, a proportionately longer circuit is required for the secondary wire.

The length of the platinum wire in the air thermometer in these experiments was 143.5 lines. This wire, which is very long in proportion to the whole circuit, can never act inductively on the secondary wire; to make the longest possible part of the main wire act on the secondary spiral, the wire in the thermometer must be shortened, by which means the action of the main wire is, indeed, increased, but on the other hand the sensibility of the thermometer is diminished.

Riess, in order to shorten the platinum wire which closed the main spiral, used Berguet's metallic thermometer instead of the air thermometer.

Fig. 63.

A straight platinum wire 61.5 lines long and 0.04 line radius was fastened immovably in the axis of a sensitive thermometric spiral, similar to that represented in fig. 63, and introduced into the circuit in a suitable manner. The instrument was of course placed under a bellglass. The platinum wire in the axis, on being heated by a discharge of the battery, communicated its heat to the spiral; the index then traversed a number of degrees, but soon returned to its first position, in consequence of the rapid cooling caused by the large volume of air in the bell-glass.

The experiments with the metallic thermometer

teach nothing new, on which account no further mention need be made of them, though I could not leave this method of observing unnoticed.

§ 66. Direction of the secondary current.-To investigate whether the direction of the lateral current changes with the distance of the secondary wire from the main wire, Riess used the following method. (Pog. Ann., LXXI, 351.)

Au insulator, which cannot be pierced by electricity, being placed between the free ends of the secondary spiral, no secondary current occurs. Nevertheless the electrical equilibrium of the secondary wire is destroyed by the act which would have produced the current, as the following experiment shows:

If we place between the free ends of the secondary spiral a thin cake of resin, so that the two ends of the wire are opposed to each other, after the discharge of the battery by the main line, the two surfaces of the cake of resin may be distinguished from each other in the most decided manner. Peculiar electrical figures are produced, which, in most cases, are brought out by slightly breathing upon them. If it be desired to fix the figures, it is done, as shown by Lichtenberg, by strewing the surfaces with a mixture of flowers of sulphur and minium. On one of the surfaces of the resin treated in this way there

Fig. 65.

Fig 64.

Fig. 66.

appears a red disk, with a red border, and beyond it a dark (unpowdered) ring, surrounded by yellow rays. On the other surface yellow and red segments of circles are visible, embraced by a wide red ring. The rays and the ring increase with the strength of the electrical excitation; with very feeble excitation the rays of the first figure are wanting, and a simple red disk remains, which, however, is sufficiently distinct from the second figure, in which the red ring may always be recognized.

Each of these figures is composed of the two elementary forms which Lichtenberg has distinguished as positive and negative, and for this reason the direction of the secondary current cannot be deduced from these figures.

Fig. 67.

In the following experiments the ends of the secondary spiral were lengthened by copper wires, and a part of one formed a short, close coil, wound to the right. In fig. 67 let x and y indicate the ends of the secondary spiral to which the above-mentioned wires are attached. To magnetize a steel needle the ends a and were put in contact, and the needle was placed in the coil, with its point toward m. To obtain the figures on the resin it was introduced between a and B. The results contained in the following table were obtained with the small main and secondary spirals, consisting of 13 feet of copper wire, already mentioned.

m

In the main spiral the discharge current passed in the direction indicated by the arrow. The following table shows the polarity indicated by the needle when it lay in the coil pointing towards m.

A glass plate was interposed between the two spirals.

It is seen that for the same direction of the main current the magnetism of the needle varies with the other circumstances, whence a difference in the directions of the secondary current might be deduced; but the resin plate being interposed between a and ẞ, and the battery discharged through the main spiral under all the circumstances given in the table, fig. 65 was constantly formed on the side of the resin plate turned toward the end of the wire ẞ-a proof that the direction of the secondary current remained the same, though the magnetism of the needle was reversed.

Riess used for producing the figure a small glass or copper plate, both sides having been covered with a thin coating of pitch or resin. A surface of resin once used must be heated over the flame of a spirit lamp to melting before it can be employed again.

The direction of the secondary current, which, as already remarked, could not be directly determined from the figures of the resin plate, was ascertained in the following way: Two three-inch condensers were separated by a thin plate of mica; the lower one touched the