The thermopile is an instrument consisting of one or more pairs of dissimilar metals, in the form of wires, oars or blocks, joined together at one extremity. On applying heat to the junctions, an electric current is set up, the pressure and volume of which depend partly on the size of the surfaces in contact and partly on the temperature to which the junctions are raised.

Such instruments are extremely useful for the production of such small, steady currents as are required for grading delicate ammeters and volt meters; for registering minute variations of temperature, etc., in which case they are used in conjunction with a delicate galvanometer, the readings of which are known to correspond with certain amounts of current, and

An Experimental Thermopile 221

Fig. 1.

consequently to given rise9 in temperature attained in muffles and similar furnaces, wherein no ordinary thermometer could be employed, since the heat would melt the glass and dissipate the mercury.

The two metals which have been more generally employed in the construction of thermopiles are antimony and bismuth, but as there are some little technical difficulties in making good junctions between the metals, he advises the student to commence his attempts in this direction by procuring two 6 in. lengths of No. 18 wire, one being of copper, the other of nickel. Having cleaned them nicely by rubbing with fine emery paper, he will hold them parallel to each other between the jaws of a vise. In order not to crush the wires it is advisable to wrap a piece of paper around them before screwing up the vise. About 1 in. of the wires should be allowed to project above the vise, when the projecting ends should be tightly twisted together by the aid of a pair of pincers.

The wires should then be removed from the vise, the twisted extremity soldered together with a little tin-man's solder. To prevent after rusting it is well not to use acid or soldering fluid as a flux, but to employ rosin for this purpose. When this has been done the two free extremities of the wires are opened out to about 4 in. apart, as shown in Fig. 1, a, b. If now the two " poles " or extremities a and b of this simple thermopile, be connected to the terminals of a fairly delicate galvanometer, the pointer of which should be brought to 0° it will be found that on applying the flame of a lighted match to the junction c, the sufficient current will be set up to deflect the needle or pointer, through 10 or 15°. A rather more powerful

An Experimental Thermopile 222

Fig. 2.

arrangement can be easily put together by making up a dozen such combinations, with the individual wires about 4 in. long only, and twisted up for a length of 3/4 in. instead of 1 in. It will not be necessary to open the wires out to more than § in. for this purpose. A common wooden curtain ring of about 3 in. internal and 4 in. external diameter is now chosen, and around its outer edge are drilled 24 equi-distant holes about 1-18 in. in diameter, just sufficient to allow the single wires to be pushed through. Then taking one of the pairs of wires with the copper to the left, the operator

An Experimental Thermopile 223

Fig. 3.

pushes the free ends thereof through the first pair of holes from the inside of the ring, until he clears the center by about 1/2 in. In like manner he inserts the remaining 11 pairs, taking care that each copper wire enters the left hand, and each nickel wire the right-hand hole of each pair, so that the ring, viewed from the outside, should show the free wire ends alternately copper, nickel; copper, nickel, and so on all round; and never two coppers or two nickels adjacent. Starting now on the outside of the ring, we twist and solder together the wire proceeding from one pair joined inside to the dissimilar wire of its next neighbor; that is to say, the nickel to its neighbor's copper, and so on all round, except only the first copper and the last nickel wires, which are left free for connection to the outer circuit. See Fig. 2.

On applying heat from the flame of a spirit lamp to the junction inside the ring, a current of electricity can be taken off the two free wires Cu and Ni. Having understood the principle of the thermopile, the operator will find no difficulty in making up the more powerful instrument generally used for measuring the temperature of different radiating bodies. We will

An Experimental Thermopile 224

Fig. 4.

begin by making a little pattern of any hard wood 2 in. long, 1/4 in. square in, the sides. From this, as a pattern, he will, after oiling it all over, to prevent adhesion, make half a dozen or more moulds in plaster of paris, as shown in Fig. 3. These plaster moulds must be allowed to dry thoroughly before use, otherwise in attempting to cast in them, the molten metal might sputter and fly into the operator's face. He will then cut 48 strips, 2 1/2 in. long, 1/4- in. wide, out of thin tinned iron sheet. Taking one of the moulds in hand, he will insert one of these strips at each end, projecting downwards into the cavity for about 1/4 in. He will do likewise with all the other moulds he has prepared. He will then melt in an iron ladle about 2 ozs. of clean zinc, with 4 ozs. of metallic antimony, with which, when well melted and mixed, he will fill the moulds he has got ready. As soon as he sees the metal has set, he will pull out the castings by the side strips; since, as they expand considerably in cooling, they would break the moulds if allowed to get cold therein. The moulds must be again fitted with tinned iron strips, and the casting operation repeated until at least 24 little square bricks of zinc antimony have been satisfactorily prepared. These are then gone over with a file to remove any excresences, care being taken not to break these in doing this, as the alloy is as brittle as glass. The operator then joins these blocks together in five separate sets of five, one of which is shown at Fig. 4; the lower tin strip of the first block being bent upwards nearly parallel to the second block where it is turned at right angles, folded upon, and soldered to the upper strip of the second block, and so on until each set of five blocks is joined. Then each tin strip is separated from its neighboring blocks on either side, by the insertion of a thin sheet of mica l 7/8-in. long by 1/4 in. wide. It will be noticed that there will be two free strips of tin to each set thus formed; one at the left hand top corner, and one at the right hand bottom corner. Each set of five blocks, when thus completed, could be temporarily tied together with a bit of fine twine.

The operator then cuts out six pieces of thin asbestos board, or failing that, of stout common pasteboard, 1 3/4 in. wide, long enough to cover the bundles of blocks-from side to side, but to have the ends protruding about 1/8 in. at each extremity. Placing a length of 1 in. wide tape on the table, he lays one of the asbestos-squares on it, puts one set of blocks squarely on it, removes the twine, then covers the first set with another square of asbestos, over which he places a second set in the reverse position, that is to say, that the free tin strips on the left hand top corner should be placed

An Experimental Thermopile 225

Fig. 5.

over the right hand top soldered junction of the first set. In like manner, with the same precautions of placing a layer of asbestos between each set, and of reversing the position of the free tin strip on the top corner with each succeeding set of blocks, the five seta are laid squarely one over the other, surmounted by a final square of asbestos or pasteboard. A piece of the same material is now cut to fit over the two sides; the tape bound tightly round the compound block, stitched firmly together and served out with a thin coat of shellac varnish. When this is quite dry all the projecting tin strips at the two ends, except only the first top corner left-hand strip of bottom row, and right-hand lower corner strip of top row, which must be left free for connection to terminals, are bent slightly diagonally over the side, so as to meet the free strip of the layer above alternately on the one side and on the other; to which they are neatly soldered,, and any excess of tin strip cut off.

The finished thermopile should then be encased in a four-sided wooden case, about 1 3/4 in. wide, tightly packed in with sheet asbestos. The two free strips are then bent up to the sides of the containing case, and fitted with small terminals. The ends of the 25 blocks should project about 1-8 in. at both ends of the case, which may for convenience be itself mounted on a rather heavy column and base. In order that the temperature of the thermopile may be maintained fairly equal at both extremities when not in use, it is usual to fit the case with little covering caps at both ends, one of which can be removed when it is desired to expose one end to the effect of the source of heat to be tested. In Fig. 5 we give a sketch of the completed thermopile - " Hobbies" London.