The thermo-electrical current has a prodigious quantity of electricity in comparison with the hy dro-electrical of silver, zinc, and water, but the in tensity of the electricity is much greater in the lat ter; the electricity of the former is impaired by the resistance of the long multiplying wire, the electricity of the latter surmounting this resistance is on the contrary increased by the multiplying wire.
The complex thermo-electric circuit produces much more effect upon the multiplier, not only when the increased number of elements heightens the effect upon the compass needle, but still also when this increase does not augment the direct ef fect upon the needle. We must therefore conclude that the intensity increases with the number of the elements in the thermo-electrical as well as in the hydro-electrical current. It must therefore be possible to attain an intensity of the thermo-elec trical current great enough for penetrating the liquid conductors, and producing the most consi derable chemical effects. Still the construction of a thermo-electrical circuit of a great number of elements is very difficult, because the elements must be as short as possible in order to preserve the conducting faculty; but even the smallness of the distance between the heated and cooled parts must give way to a very speedy re-establishment of equilibrium. The best way seems to be, to pro duce the heating and cooling of the junctions by some continual current of hot and cold liquids.
A very easy manner of constructing thermo-elec tric batteries deserves to be mentioned. Fig. 18 represents it. The parts indicated by the odd numbers 1, 3, 5, represent copper slips, and those indicated by the even numbers 2, 4, 6, small bars of bismuth. All the junctions situated on one side of the dotted line ed, are to be heated, those on the other side are to be cooled. The extremities a and b are to be connected by a conductor. The num ber of elements may here be tolerably great.
That the intensity of the electro-magnetic cur rent must increase with the temperature was to be presumed; but this is not a general law. Dr. See beck had already found some exceptions, and also Professor Cumming at Cambridge, who made his experiments without knowing those of Dr. Seebeck upon this subject. We shall not stop here to de tail these experiments, as another philosopher, Mr. Becquerel, availing himself of the imposed instru ments of research, and making a very ingenious ap plication of them, has given us exact measures of the quantities here occurring.
It was supposed that the declination of the needle, produced by the electrical current is in the ratio of the sine of the angle of deviation. Though this is a consequence of the resolution of powers, he thought that, in a matter so little known as the magnetical effects of the electrical current, it might be advisable to examine the law of this measure by experiment, particularly with regard to the multi plier, where the current makes so many windings round the needle. In order to execute this plan, he formed his multiplier with four parallel and equal wires, covered with silk, and each making an equally great number of windings. Thus he had four multiplying windings about one frame. To the ends of each multiplying wire he soldered the ends of an iron wire, so that four thermoelectrical circuits, consisting of the copper wires of the mul tiplier and the iron wires were formed. When he
wished to put one of these currents in activity, he cooled one of the junctions with ice, and heated the other in mercury. The junction was included in a thin bent glass tube, in order to guard it against the dissolving power of mercury. The mercury was heated by means of a lamp, somewhat above the tem perature required, and when heated the lamp was taken away; thus the temperature remains for a short time stationary. In this manner he tried first the effect of one, then of two, three, or four of the multiplying circuits, and noted down-the deviations produced, one of the junctions still being kept at the freezing point. Thus he found that one of the circuits gave, by 5° Centigrade or 9° Fahr. above the freezing point (41° Fahr.) a magnetic deviation of 0.65° French division, or of common di vision of the arc. Two circuits gave by the same temperature twice three gave thrice, and four gave four times this quantity; whence he con cluded, that when one circuit produces it has four times the power of that producing only It is easily understood that the greater angles of deviation could not be in the same ratio as the action; but this does not hinder us from drawing analogous conclusions. Thus by a differ ence of 180° Fah. one circuit gave the deviation 10,71° of the circle; but two circuits gave nearly the same (10.575) by a difference of 90° Fah. But it is not in all temperatures that this proportion of the effect and temperature takes place; in very high degrees of heat he found that the effect of circuits of copper and iron did not increase so fast as the temperature. From the freezing point (32° Fah.) up to 284° Fah. the magnetical effect increases with the temperature. From this degree to the magnetic power, though increasing with the tern perature, still proceeds in a decreasing progression; and exposed to the immediate action of a lamp, the current is inverted. When none of the junctions is at the freezing point, the effect of the circuit is equal to the difference of the effect, which each of the two temperatures applied to one of the junctions, the other being at the freezing point, should give; thus, for instance, a circuit cf iron and copper, when one junction is heated to 392° F., the other being at 32° F. has an intensity expressed by 37; but when the heat is only at 212', the intensity is ex pressed by 22. The difference of these two numbers is 15, which is found by experiment to be the effect of the circuit, in which one junction is heated to and the other to 212°. He found that a com plex circuit of copper and iron produced an effect proportional to the number of elements, which is not the case, when the whole power of the circuit can be exerted, but is only so, when a very small part of the whole effect can be transmitted through a conductor, of such a length or feeble conducting faculty; that it requires much intensity of electri city, for being penetrated. Thus the observation of 1Ir. Becquerel proves, what had already been shown by less perfect experiments, that the intensi ty of thermo•eleetricity increases as the number of the elements.