limit to the number of plates which may thus be multiplied, without the least deterioration or damage to the original. The copper which is precipitated is found to be of such superior quality that plates prepared from it are preferred by artists for their original work to common copper, which it is difficult to obtain of uniform quality, and consequently is very expensive.

A process of volta-etching may also be carried on by placing a plate, etched upon a resinous ground in the usual way, at the zincode of the battery instead of the platinode, when the biting in will be performed by the oxysulphion disengaged upon it more sharply and much more conveniently than by the usual action of nitric acid.

§ 821. It is impossible indeed to describe all the ingenious applications which have already been made of the philosophic principles which have been lately developed in this branch of science; most of which have been brought to a degree of perfection in two or three years which is truly admirable. The last which we can mention may perhaps be called voltaembossing, by which the substance of copper plates is partially thickened, according to the designs of a pattern stopped out upon the plate with resinous grounds, and which are afterwards used for printing. In this manner an artist draws his design upon a plate with prepared liquids, submits it to the battery, and it is afterwards printed upon paper, and cylinders so prepared are now used with great advantage in establishments for printing cottons.

XV. THERMO-ELECTRICITY.

§ 822. WE have seen that, when the electric current meets with obstruction to its course, the equilibrium of heat is disturbed even in good conductors; it might, therefore, have been à priori expected that a disturbance in the equal flow of heat would produce an electric current, in such forms of matter as are capable of transmitting it.

We are indebted to Professor Seebeck of Berlin, for the experimental confirmation of this conclusion in 1822, and the discovery of the phenomena of THERMO-ELECTRICITY.

§ 823. If a platinum wire be carefully soldered to the two extremities of a delicate galvanometer, and it be heated at

any point remote from the junctions, no disturbance of the electric equilibrium will be produced; from the homogeneous structure of the wire, the heat will flow equally to the right and left of the heated point. But it will be very different if a knot, or a spiral turn, be made in the wire without breaking it; for if the focus of heat be applied to the right of such obstruction, an electrical current will be established from right to left and will be indicated by the needle. This must arise from the unequal rate at which the heat will obviously be propagated on the two sides of the obstructing mass. Wires of copper and silver will act in the same way, only in a very inferior degree. The same effect will be produced if the wire, instead of being continuous, be divided, and each end being twisted into a spiral to increase the surfaces, one be heated red in a spirit lamp, and brought into contact with the other. The deviation of the magnetic needle will indicate as before, that a current is passing from the hotter to the colder point. That these effects do not depend upon any chemical action of the air, is proved by performing the experiment under the surface of well purified oil, under which circumstances the same results will be obtained.

Those metals which possess a decidedly crystalline texture present even more marked electrical phenomena from the unequal propagation of heat in their masses. If a ring, or rectangle, be cast of antimony or bismuth, of the diameter of three or four inches, and of the substance of about the third of an inch, and one half of its surface be kept cool by ice and the other heated, a current of electricity will be immediately established of sufficient power to affect the magnetic needle without the assistance of a coil.

§ 824. These thermo-electric effects, again, are very much increased by combinations of two metals. If a bar of antimony have a copper wire soldered to it, or merely twisted round one of its ends and attached to the other in the form of a loop, when heated at the contact of the metals at one extremity, it will strongly deflect a magnetic needle placed above or below it (149).

It is found that similar circuits may be formed of combi

(149) a represents a bar of antimony with a piece of copper wire twisted round one end of it, b, and looped over the other end, c.

a

When heated by the flame of a spirit lamp

at the contact of the metals b, a magnetic needle placed at d, will be deflected.

nations of other metals, and that they may be ranged in the following order according to their thermo-electric efficiency; the most powerful combination being formed by those metals which are the most distant from each other in the series:bismuth; platinum; lead; tin; copper, or silver; zinc; iron; antimony. When heated together, the current is found to proceed from those which stand last to the first.

§ 825. There can be little doubt that the specific heat and conducting power of the metals must be concerned in these effects; but in comparing the tables of each, the connexion does not immediately appear. Structural crystalline arrangement has also much influence upon them. With some combinations, as for example, zinc and silver, the current will go on increasing with the temperature to a certain point, 248°; will then become null, and, by increasing the heat, will be reestablished in the contrary direction. This singular phenomenon is most probably referrible to a change in the arrangement of the particles of the zinc.

§ 826. The different powers of the currents from different couples of metals, for the same differences of temperature, may be ascertained by forming a compound circuit of all those which it is wished to compare. All the junctions are kept at the temperature of melting ice, except that which is to be thrown into activity, which may be raised by being plunged into hot oil. The mere conducting power of the circuit thus remains the same in every experiment, and the results will be strictly comparable with each other.

§ 827. Similar circuits may also be formed, according to the experiments of M. Nobili, with substances whose conducting power is lower than that of the metals. He made small cylinders of porcelain clay, of the length of two or three inches and three or four lines in diameter, and wrapped round the ends of each some cotton steeped in a conducting liquid which served to place them in direct communication with the galvanometer. One of the ends was reduced to a point, and after heating it red at a spirit lamp, he pressed it upon the cold extremity of the other cylinder, and a current was established from the hot part of the arrangement towards the cold. This effect is owing to the mutual reaction of two portions of water of different temperatures.

The following table exhibits the quantities of the currents for a difference of 36°, of pairs of eight metals differently arranged; the lengths of which were 7.88 inches, and their diameters about two hundredths. The sign + indicates the metal from which the current proceeds :—

TABLE XLV. Thermo-electric Powers of different Metallic

§ 828. The thermo-electric current may be increased by forming a compound circuit, and arranging the pairs of metals in a series of alternations analogous to those of the voltaic pile. If three bars of bismuth with three of antimony placed alternately, so as to form the sides of a hexagon, be soldered together and placed upon two supports in a horizontal position, one of the sides being in the magnetic meridian with a compass-needle below it, upon heating one of the junctions, the needle will be sensibly affected. The deviation will be considerably increased on heating two of the alternate angles of the hexagon; and a still greater deviation will be produced when the heat is applied to the three alternate angles.

In all these combinations, similar effects may be produced by reducing the temperature of one or more of the junctions by

the application of ice or other means; and when the action of ice, on one set of the alternate angles, is combined with that of flame on the other set, the effect is still more considerable (150). Upon multiplying still further the number of alternations, it is, however, found that the total effect is inferior to the sum of the effects which the same elements could produce when employed in the formation of simple circuits. This is owing to the low state of intensity of the current, which occasions great loss of power, whenever it has to traverse any considerable line of conductors even of the most perfect kind.

§ 829. The thermo-electric pile has lately been applied by Messrs. Nobili and Melloni, as a most delicate and accurate measure of temperature. Thirty-six pairs of bars of bismuth and antimony, packed into a small compass, and having a very delicate galvanometer with two needles attached to them, were found to be so sensible as to be affected by the warmth of the hand at the distance of thirty feet, and an instrument constructed upon this principle, with all the necessary precautions, is applicable to a variety of delicate investigations for which any of the common thermometers would be totally insufficient.

§ 830. The thermo-electric current will occasion convulsions in the limbs of a frog, but is inadequate in its primary state to effect any kind of chemical decomposition.

§ 831. M. Pouillet has compared together the thermoelectric and the hydro-electric currents, by passing the latter through a platinum wire of sufficient length to reduce it to an intensity just sufficient to balance the former. In one of his experiments he found that 590 feet of platinum wire '006 inches in diameter, comprising the resistance of the battery previously

b

(150) a a a a represent four bars of antimony soldered to four bars of bismuth, b b b b. If the extremities at c be placed in perfect metallic communication, as by the wires of a galvanometer, and the temperature of the junctions be alternately raised and depressed, an electric current will be determined in the direction of the arrows; which will cause the deflection of a magnetic needle placed immediately over, or under, any part of the circuit.