THERMO-ELECTRICITY.

(1.) Introduction.—I take a hoop of copper wire and hold one part of it in the flame of a candle, or any other source of heat. By degrees the whole hoop becomes heated, the heat travelling round it on either side. I now insert a piece of antimony in the hoop, by cutting out a piece of the copper wire, which I replace by the antimony. Heating the point of junction, I find I have done more than send a current of heat round the wire. If I place a magnetic needle near the hoop (or connect it with a galvanometer) I get evidence of a galvanic current that did not previously appear to exist when the hoop was of one metal only. What are the different conditions introduced by the insertion of a piece of antimony into the circle of copper? Why should the passage of heat develop electricity in one case and not in the other?

The answer to this is, like the answers to most questions in natural science, of a twofold character. First, the reply is, that the difference in the conducting powers of the two metals and in their specific heats causes the electric current. I heat one part of a copper hoop, and place a delicate galvanometer at the opposite side. No deflection of the needle takes place; not because there is no force acting on it, but because there are two equal forces. Whatever effect the heat transmits round the hoop in one direction, it transmits also an equal and opposing force in the other. How then do we know that any force whatever is transmitted, since there is no evidence of its existence? That there is such a force may be shown by disturbing the equality of the two forces, and allowing one to act more than the other upon the measurer of force, the galvanometer. I take away a portion of copper on one side of the lamp, and replace it by an equal amount of bismuth. The conducting power and specific heat of bismuth are not the same as those of copper, therefore the heat passing from the lamp to the galvanometer on one side through copper only, is not exactly counterbalanced by that passing on the opposite

ELECTRICITY DERIVED FROM HEAT.

263

side through copper and bismuth, and the result is, the greater of the two forces deflects the needle of the galvanometer.

But if the influence of the bismuth be merely a mechanical one, might not the same result be attained by other means? If the bismuth merely retards the passage of the heat, might not the same result be attained by inserting an additional piece of copper, as by replacing copper by bismuth? If the bismuth accelerates the passage of the heat, will not the same result be attained by removing a part of the copper? To prove this I might simply move the lamp towards the galvanometer either on the right or left, so that the heat has to traverse a less distance on one side than on the other. I do this, and a faint evidence of galvanism is the result. I twist the wire a few times on the longer side, so as to interpose more resistance to the passage of the heat, and the movement of the needle is increased.

(2.) Derived from Heat.-From this it would appear that two unequal currents of heat opposed to each other produce the result to which we give the name of an electric current; and that the bismuth inserted in the copper ring causes electricity by its interference with the equality of the currents, and not as a distinct effect. So that electricity would seem clearly to be connected with heat.

But if electricity be thus produced by two unequal currents of heat, it should be stronger as this inequality is increased. Might it not also be reasonably expected to be strongest when this inequality was greatest-i.e., when the weaker current of heat was altogether suppressed?

Thus I place the source of heat (say a gas flame) on one side of a copper hoop, and the galvanometer on the other side. But little, if any, electricity is evident. I interpose some substance on one side, so as to make the two currents acting on the galvanometer unequal, and I find that the needle moves more and more as I increase this inequality. But suppose I remove one side of the hoop altogether; ought not the current to be still stronger? But it apparently ceases to exist, and the result is simply conduction of heat from the gas flame to the needle.

This brings us back to the question-What effect can be produced by heat passing round a metal hoop which could not be produced by its passage along a simple rod of metal? Thus, I attach one extremity of a length of copper wire to a galvanometer and heat the other by a flame. At the same time I take another length of wire, and fasten both ends to another galvanometer, so as to make a continuous hoop of copper, which I heat (at the point opposite the needle) by a flame of equal power to the first one.

The single copper wire conducts heat to the galvanometer, but does not produce any signs of an electric current; neither does the double wire of the hoop, so long as the light and galvanometer are opposite to each other, having equal lengths of wire on either

side. But when I move the flame on one side nearer to the needle, or if I interpose any comparatively bad conductor, the needle is at once deflected. Assuming heat to be vibration, and electricity to be arrangement, how can this be explained?

The heat travels from the lamp along the wire towards the galvanometer until it comes into contact with the foreign substance interposed. This is supposed to be a bad conductor of heat. What does this mean? Is it not that the vibration does not travel so rapidly along it? This is probably because of the greater difficulty in setting the atoms in motion. The heat will, therefore, accumulate at the point of junction. What will be the effect of the accumulation? Partly to raise the temperature, but partly also it will tend to arrange the atoms of the body.

The ordinary apparatus, fig. 157, used for the production of thermo-electricity is the "thermo-electric pile," composed of a

number of pieces of bismuth and antimony, so that a current of heat passing through the battery has to pass from bismuth to antimony, and from antimony to bismuth again and again. This repeated interruption, accumulation, and conversion of heat produces an electric current, and this may be made evident by the use of a magnetised needle or galvanometer. If a wire be fastened to either extremity of the battery, and connected with the galvanometer, the needle is deflected whenever the junctions of the bismuth and antimony are heated or cooled.

Why bismuth and antimony are used will at once be apparent from the following table, which gives the order of power in a battery of this kind.

Bismuth.
Platinum.

Tin.

Lead.

Gold.

Copper.
Silver.

Zinc.

Iron.
Antimony.

Any two of these metals being used as elements of a thermo-electric battery, the one standing higher in the list here given would be positive to the lower one. Thus, tin is positive to gold, gold to copper, silver to iron, &c., so that necessarily bismuth and antimony, being the extremes of the list, will, when coupled, give the strongest current.

The similarity of the reasoning here and in the case of the elements of a galvanic battery is at once apparent. In the case of the development of electricity by chemical affinity, the affinity of a metal for oxygen determines its place in the table. In the case of development of electricity by heat, what is it that determines the order? Why is bismuth at the top, and antimony at the bottom of the list? I will try to answer this question in the final chapter.

RADIATION.

(1.) Introduction.-Radiation is the giving off, in all directions, of heat, light, sound, &c., and depends upon the surface rather than nature of the radiating body. In a perfectly dark room I bring together the carbon points of an electric lamp, and then slightly separate them. Instantly a bright light is produced, which spreads throughout the room, reaching to every point of the walls, ceiling, or floor that is not hidden by any nontransparent body; and even if any such substance prevent the light from reaching directly any part of the wall or ceiling, yet the part so hidden from the light is not perfectly dark, for the light bends round the screen more or less. If I enclose this electric light in a box having an opening (say a foot square) on one side, the light will pass out through this, radiating as before. If I place this box near the wall, I produce on the wall a light of a foot square; if I withdraw the box from the wall, the illuminated space enlarges, but the light becomes less bright. When the distance from the aperture to the wall is equal to the distance from the light (ie., the carbon points of the lamp) to the aperture, the lighted space on the

wall will be 2 feet each way, or 4 square feet in size, but the light will be much less bright than before. The surface illuminated will be four times as great, but it will be lit up with only one fourth the degree of brightness. As I draw the box still farther from the wall, I light up a still larger surface, but less brightly.

Fig. 158 shows a source of light shining on a screen at 1 foot distance, and on another at 2 feet. space 6 inches square on the first screen, a space 12 inches square on the second.

(2.) Law of Inverse Squares.-Just as any given quantity of water will cover a small surface more deeply than a large one, so a given quantity of light will light up more brightly a small than a large surface. And this is true whether light be a vibration or a real substance. Assuming the truth of the theory that it is but a vibration, the more vibration the more light; but the larger the body of ether to be set in vibration by any given source of vibration the more feeble the vibration will be, and consequently the less bright the light-i.e., the less effect will it have upon the nerves of the eye.

Thus I construct a hollow pyramid 3 feet in height, and having a base of 9 square feet.e., 3 feet in length and in breadth. I place the electric light at the apex, so that the rays of light fill the pyramid. If I put a screen across this at 1 foot from the light, it will measure just 1 foot each way— i.e., 1 square foot. The space thus enclosed is filled with light -.e., the ether filling the interstices of the air is set in vibration with a force corresponding with the intensity of the light. This light falling upon the screen will be more or less refracted, reflected, and absorbed, according to the nature and surface of the screen. The reflected light will make the screen visible, and it may be seen if the eye be placed at any small opening in the side of the enclosure. If now the screen be drawn farther from the light, it will have to be enlarged to cover the base of the pyramid. At 2 feet from the light this base will be 2 feet each way, or 4 square feet. The light now fills the whole of the larger pyramid, there is a larger body of ether to be set in vibration, and the vibration is consequently less, just as any given motive power will move a large weight a less distance than a small one. The farther the screen be drawn back, the greater will be the body of ether to be set in vibration between the light and the

screen.

It is not, however, merely because of its greater bulk of ether to be set in motion that the intensity of the light or vibration is diminished, but because of its increased lateral dimensions. Thus a certain area-say a square foot-of air is set in vibration, so as to produce a sound. As this vibration continues onwards its lateral area also increases, and the sound decreases; but if the vibration be confined, as in a tube, so that it moves only longitudinally, and not laterally, the sound may be conveyed with but little diminution for considerable distances. Thus, by using a speaking-tube, a whisper from a room at the top of a house may be made quite audible in the kitchen, or vice versa. speaking-tube is nothing more than a pipe which allows the longitudinal extension of the vibration, but prevents any lateral spreading, and consequent decrease in intensity. This may be illustrated by noticing the passage of any liquid, such as water, through a narrow pipe that becomes gradually wider and wider. If it be poured rapidly in at the narrow end, it will preserve its

This