ume of my treatise, (Lehrbuch der Physik.) If the current of a battery be passed through a voltameter, and then, directly after breaking the circuit, each of the voltameter plates be brought into contact with the terminating wire of a multiplier, the latter will indicate a current traversing the voltameter in the direction opposite to that of the original current of the battery. This experiment, made as early as 1827, by De la Rive, merely shows that an electro-motive opposing force is generated in the voltameter by the primary current; but it gives us no clue to the cause.

Becquerel maintained that the secondary current appeared only in the case when the poles were immersed in the solution of a salt. Under these circumstances, says Becquerel, the salt is decomposed, the base collects at the negative pole, the acid at the positive; and if the wires be put in conducting connexion after the removal of the battery, a current is generated in consequence of the re-combination of the acid and base.

Schönbein now shows that a solution of a salt is not at all necessary for bringing about a secondary current; that the experiment succeeds perfectly with pure water very slightly acidified with pure sulphuric acid, even if the platinum electrodes communicates but for an instant with the battery.

These secondary currents are by no means of only momentary duration; they last, according to circumstances, a longer or shorter time. In an instance in which the original deflection of the galvanometer needle by the secondary current amounted to 80°, four minutes elapsed before it altogether disappeared; in another, when the deflection was 160°, it lasted thirty minutes.

Schönbein produced secondary currents as well with electrodes of gold as with those of platinum. Iron wires being used instead of platinum, and a solution of potash for sulphuric acid, the secondary current also appeared. Experiments with silvered copper wire, zinc, and other metals, gave similar results; so that it is in the highest degree probable that all metallic conductors have the property of being electrically polarized.

In the second of the above-mentioned memoirs (P. A. XLVII, 101) Schönbein arrives at an explanation of the phenomenon. The most important facts which lead to it are the following:

1. Platinum wires or plates which, being placed for a greater or less length of time in pure water, or in water with sulphuric or nitric acid, have served as electrodes, and are then heated to redness in a spirit flame, lose entirely all their electro-motive power.

2. If the positively polarized electrode, or that which has served as a negative pole, be exposed but for a few moments to an atmosphere of chlorine or bromine, the electro-motive force will be completely destroyed; the same result is also obtained by a longer immersion in oxygen gas.

3. A negatively polarized platinum wire loses its electro-motive force if it be exposed a few seconds to an atmosphere of hydrogen.

4. By exposing positively or negatively polarized platinum plates to a gas which has no chemical action either on oxygen or hydrogen

in the presence of platinum, the electro-motive force of the plates will not be destroyed.

5. A platinum plate exposed for only a few seconds in an atmosphere of hydrogen, is polarized positively.

6. Gold and silver wire do not acquire electro-motive power by immersion in hydrogen gas.

7. A platinum wire placed in oxygen does not become negatively polarized, nor do gold and silver.

8. Platinum, gold and silver, exposed for a few seconds in chlorine gas, become polarized negatively. Bromine gas produces the same effect on these metals.

Before passing to a further elucidation of these facts, we will consider the most advantageous way of showing the electrical polarization of a metallic plate.

Fig. 28.

In a small cup of mercury a, Fig. 28, connected with the terminal wire of a multiplier, the end of a wire of a platinum plate p is immersed. The plate must be first perfectly cleaned, and then suspended in a glass of acidified water. In the cup b the wire of a second and exactly similar platinum plate is placed the plate being in like manner cleaned and suspended in the acidified water. The needle will, of course, remain at rest, since both plates act exactly alike electro-motively. But if the second plate, which we will denote by p', should be polarized in any of the above ways, a deflection of the galvanometer needle would follow, from which the direction of the current could be ascertained.

n

For example, if the platinum plate p' were immersed in hydrogen gas, it would act electro-positively towards the other; that is, the galvanometer would indicate the current passing from p' through the liquid to p. The plate p' being immersed in chlorine gas, the deflection of the needle would show p' electro-negative to p.

If the platinum plate p' should have served as the negative pole in the decomposition of water, it will act exactly as though it had been plunged into a jar of hydrogen; that is, if used for closing in the apparatus of Fig. 28, it would generate a current passing from p through the liquid to

p.

All the phenomena we have just considered, appear to indicate that the stratum of gas which escapes at the electrodes during electrolysis is really the cause of galvanic polarization. If such be the case, it is perfectly evident that the stratum of gas will be destroyed by heating the metal plates to redness. This circumstance alone, however, would prove nothing, because such a heat must act destructively upon the polarity, even if it should depend upon other causes than upon a stratum of gas. The second experiment is decisive. The instantaneous destruction of the positive polarity of a platinum wire, by chlorine, can hardly take place otherwise than by the chemical action of the chlorine on the oxygen, by which every trace of hydrogen disappears

in the formation of hydrochloric acid. On immersion in oxygen, the hydrogen adhering to the platinum plate is caused by the action of the latter to combine with the oxygen, and thus the cause of the polarization is removed. That oxygen does not destroy the positive polarity so quickly as chlorine, is owing merely to the slow action of the oxygen.

The fact mentioned under No. 4 is also favorable to the view, that the cause of the polar condition of the electrodes exists in the hydrogen and oxygen which adhere to them. The certainty of this supposition is established by the fact stated in No. 5; at least this appears to prove incontestably that the positive polarity of the negative electrode is due to hydrogen.

A platinum wire which has not been used as a negative pole, and has not been subjected in any way to the influence of a current, presented all the voltaic properties of a positively galvanized wire, merely from the fact of having been exposed a few seconds to hydrogen.

Schönbein has, in fact, by these experiments, removed the vail which has hitherto concealed the nature of galvanic polarization.

Only two of the facts stated above, namely, those under 6 and 7, appear to oppose the explanation he has given.

While a platinum plate, which has been used as a positive electrode, is negatively polarized, the polarization cannot be produced by exposure to oxygen; this seems to show that the negative polarity of the positive pole is not to be ascribed to oxygen.

The circumstance that gold and silver wire do not become electiopositive in hydrogen, while the same metals, if they have played the part of negative electrodes but for a few seconds, become sensibly positively polarized, excites some doubt as to the correctness of the view that the positive polarization of the negative electrodes is to be attributed to hydrogen.

But before passing to a closer examination of this subject, we will first consider the polarization of liquids which Schönbein also discusses in the above mentioned memoirs.

§ 40. Polarization of liquids.-If dilute hydrochloric or dilute sulphuric acid be placed in a U-shaped tube, and connected a few seconds by platinum electrodes with the poles of a battery, the current of which causes a sensible development of gas in the acidified liquid, and if then the wires thus used be replaced by new ones, or such as have not served as poles, and these wires be connected with the galvanometer, the needle of this instrument will deviate, and in a direction which shows that the positive current of the liquid column in which the negative pole was immersed passes in the direction of that in which the positive electrode was, or, in other words, the secondary current is in the opposite direction to the current of the battery. Thus liquid columns indicate galvanic polarization.

The cause and nature of this polarization are explained by the following experiments:

1. Water, made conducting by a little sulphuric acid, being agitated with hydrogen and placed in a tube closed below with a bladder, and the tube put in a vessel which also contains some acidified water, but free from hydrogen, and both liquids then connected with the

galvanometer by platinum wires, a current is obtained which passes from the hydrogen solution to the other liquid. The former acts relative to the latter as zinc to copper. When gold or silver wires were used in this experiment, no current was obtained.

2. The experiment having been made under exactly the same circumstances, excepting that the liquid in the tube contained oxygen in solution instead of hydrogen, there was no current with connecting wires of platinum, gold, or silver.

3. When the liquid in the tube contained a small quantity of chlorine or bromine instead of hydrogen, a current was obtained, which passed from the wide vessel into the tube, whether the experiment was made with platinum, gold, or silver wire.

4. If the current of a battery be passed through water containing sulphuric acid placed in a U-shaped tube, this liquid will yield a secondary current only in case the connexion with the galvanometer be made with a platinum wire. By using gold or silver wire the needle of the multiplier does not show the least deflection.

5. If the experiment be made as in 4, using, however, dilute hydrochloric acid instead of dilute sulphuric acid, a secondary current will be obtained even when the closing has been made with gold or silver wire.

The experiments under 1, 3, and 5 indicate that the course of the polarization is to be found in the gases which are dissolved in the

water.

The cases in which the liquids treated with gases yield no current from polarization, (Nos. 2 and 4,) exactly correspond with the cases above described, where metallic wires or plates immersed in gases produced no such current, (Nos. 6 and 7.) In order to prove that the stratum of gas adhering to the metallic plates or the gases dissolved in the liquids are the cause of galvanic polarization, it must be explained why the same effect is not also produced in these cases. The view of Schönbein on this subject we give in the following paragraph:

§ 41. Schönbein's theory of galvanic polarization.-If two like metallic plates be immersed in a liquid, one clean and the other coated with a stratum of gas; or if two such plates be placed in the two branches of a U-shaped tube filled with the same liquid, except that in the liquid in one of the branches a gas is held in solution, and not in the other, the dissimilarity between the two parts is a sufficient cause for the appearance of electrical tension. This tension will cause an electrical current as soon as a metallic connexion is made between the two plates. But in order that such a current may traverse the wire of a multiplier, it must pass through the liquid, which cannot transmit the feeblest current without electrolysis. The appearance of the polarization current therefore is inseparably connected with the beginning of the electrolysis of the liquid; the current cannot exist in any case when the electrical difference in the two surfaces in contact is not sufficient to bring about electrolysis.

For example, if the water acidified by sulphuric acid on one side be terminated by a pure gold or silver plate, and on the other side by one coated with a stratum of hydrogen, no current appears on con

necting the two plates, because the hydrogen of the gold plate is not in the condition to attract the oxygen of the nearest particle of water, and thus to produce electrolysis throughout the whole stratum of liquid; but if platinum be used instead of the gold plate, the peculiar relation of this metal to hydrogen and oxygen induces electrolysis by causing the hydrogen nearest the platinum plate to attract the oxygen from its neighboring particle of water, and thus the decomposition and recomposition of water extends to the other plate. Thus it is shown why the experiment No. 1 succeeds with platinum plates, but not with gold or silver.

Schönbein considers it not improbable that this action is produced by a sub-oxide of hydrogen, the hydrogen of which has a greater deoxidizing force than pure hydrogen, as the third atom of oxygen of the super-oxide shows a greater affinity for oxidable bodies than pure

oxygen.

A platinum plate immersed in chlorine gas, combined voltaically with a clean one in dilute sulphuric acid, yields a current, because, in this combination, the affinity of the chlorine is sufficiently strong to attract the hydrogen from the nearest molecule of water, and form hydrochloric acid; hence the electrolysis of the water is induced all the way to the other plate. Even if gold and silver plates be used in this experiment, the chlorine has the power of decomposing water; hence, in this case, the current which passes, of course, in the direction in which the particles of hydrogen go, continues until the chlorine on one of the plates disappears.

The formation of the current in experiment No. 3 is to be explained in a manner entirely analogous to this.

But, by using pure oxygen instead of chlorine or bromine, in the above combination, it is not found in such a state of activity, to use Schonbein's language, as to cause the decomposition of the nearest particle of water; hence the absence of the current in the experiments No. 7 (in section 39,) and No. 2 (in section 40.)

But if pure oxygen in this case cannot excite a current of polarization, how is the negative polarization of a platinum plate, which has served as a positive electrode, to be explained? Certainly not by the oxygen evolved at its surface. Ozone, as well as oxygen, escapes at the positive electrode, and that this remarkable body can polarize platinum plates negatively has been stated.

According to Schönbein, ozone is a super-oxide of hydrogen; a view which is strongly supported by the fact that the super-oxides of metals have a precisely similar voltaic action. The third atom of oxygen has a greater affinity for oxidable bodies than free oxygen, and thus the strong electro-negative action of these substances is explained.

§ 42. Hyper-oxide batteries.—A platinum plate, covered with a hyper-oxide, as, for instance, hyper-oxide of lead, acts electro-negatively towards a clean platinum plate. On immersing the two plates connected with the terminal wires of a multiplier, in dilute sulphuric acid, a powerful current arises, passing from the clean plate to the one covered with the hyper-oxide. The third atom of oxygen in the