217 



Thefo 

 mente. 



THERMO-ELECTRICITY. 



THERMO-ELECTRICITY. 



213 



allowing figure shows the general arrangement of these experi- 

 A rectangle is composed of antimony (shaded) and bismuth. 



t 



On heating one of the junctions a thermo-electric current will set in 

 in the direction of the arrows and the needle will be deflected. 



Thermo-electric circuits may be formed in a ring consisting of two 

 curved bars of different metals, as bismuth and copper, each being in 

 the form of a semicircle, and the two being attached together in the 

 direction of a diameter ; or they may be produced in a rectangle made 

 by placing in close contact four bars of metal, of two different kinds, 

 following one another alternately. M. Oersted formed a hexagonal 

 circuit with six pieces, three of bismuth and three of antimony, which 

 were disposed in alternate order : on heating, by means of a spirit- 

 lamp, one of the places of junction in the ring, or in the rectangle of 

 four pieces, a conipass.needle placed within or below the plane of 

 circuit waa found to deviate; and it deviated still more when the 

 opposite angles of the rectangle were heated. In the experiment with 

 the hexagonal circuit the deviation was greater in proportion to the 

 number of alternate joints which were heated. Similar effects were 

 produced when the alternate joints were artificially cooled ; but the 

 deviation waa the greatest when the alternate joints were heated and 

 the others were cooled. 



By doubling the lengths of the bars in a rectangle composed of four, 

 the deviation was legs than that which was produced by the smaller 

 rectangle ; but when the larger rectangle waa composed of eight pieces, 

 the deviation was greater. 



In this country the subject of thermo-electricity waa pursued by 

 Professor dimming of Cambridge, who appears to have entered upon 

 it without any other knowledge of the discovery of Seebeck than the 

 simple fact that electro-magnetical action was produced by heating one 

 end of a bar of antimony, to the extremities of which were made fast 

 those of a brass wire. The details of his researches are published in 

 the ' Cambridge Philosophical Transactions," for 1823. From these it 

 appears that all perfect conductors of electricity, on being heated or 

 cooled in any part, exhibit in general magnetical phenomena ; but the 

 intensity of the action, which is indicated by the amount of the 

 deviations produced in a magnetised needle, ! not the same in all 

 substances, and with some the direction of the current in contrary to 

 that which is produced in others. When a single bar, of symmetrical 

 form, is heated in the middle, it produces no effect on the needle, 

 probably because the opposing currents counteract each other ; and in 

 a ring formed of two metals, when heated at one of the points of 

 junction, the fluid seems to pass from one metal to the other ; so that 

 one loses positive electricity, or becomes negative, while the other 

 becomes positive. If, however, the ends of a sensitive galvanometer be 

 united by means of a platinum wire, no current will be perceived on 

 heating the wire until a portion of it be twisted into a loop; the 

 molecular tension thus produced will interfere with the conducting 

 power of the wire, and a current will flow through the apparatus 

 from right to left when heat is applied close to the loop and to the 

 right of it. Or if the wire be formed into two flat spirals, and one 

 be heated to redness, and brought into contact with the cold spiral, 

 a current will flow from the hotter to the colder portion. Or if portions 

 of a metallic wire be stretched by weights while other portions are 

 free, a current will flow from the former to the latter on applying 

 heat to the junctions. These last-named facts are due to Professor 

 W. Thomson. 



Professor dimming having ascertained from experiments on bars of 

 bismuth, which were made alternately hot and cold, and were placed in 

 contact with each other (each pair of the hot and cold parts, and also 

 the two extremities of the whole compound bar, being connected 

 together by wires), that the action of the whole bar on a needle was 

 greater than that of any two portions, one hot and the other cold, was 

 led to the discovery that electro-magnetism may be exhibited by the 

 mere juxtaposition of an indefinite number of small plates. He was 

 also enabled to determine the thermo-electric relations of different 

 metals by merely placing in contact with each other a small portion of 

 each of the two kinds of metal to be examined, and touching first one 

 of them, and then the other, with one end of a silver or copper wire 

 which was connected with the heated bar. When the metals were 

 bismuth and antimony, the former, on being touched, caused the 

 compass-needle to deviate so aa to indicate positive electricity, and the 

 latter so as to indicate negative electricity ; and in the memoir above 

 quoted there ia given a table of the electrical relations of metals 

 in several different combinations. In the same memoir there is also an 

 account ot*sevcral curious anomalies which were observed in the 



magnetic action : one of these is, that when iron wire is used to touch 

 the metals examined, of which one is iron, the needle deviates a certain 

 number of degrees in the positive direction ; then, as the heat of the 

 wire ia increased, the deviation in that direction gradually diminishes 

 iill it becomes zero ; after which the deviation takes place in a negative 

 direction, and it becomes a maximum in this direction when the wire 

 acquires a red heat. 



Metals with a marked crystalline structure, and an inferior power for 

 conducting electricity, display thermo-electric effects most perfectly. 

 Thus bismuth and antimony are favourable metals, and even the 

 warmth of the hand applied to one of the junctions of a pair will 

 deflect the galvanometer needle. Indeed for small differences in 

 temperature, no instrument is so sensitive as a thermo-electric pile or 

 thermo-multiplier. An instrument of this kind was invented by 

 Gumming, and improved by Nobili, who introduced the astatic needle 

 into it. 



If a thermo-electric pair have one of its junctions kept at 32, whilst 

 the other junction is gradually raised in temperature, the current will 

 gradually increase in intensity up to a certain point, then decline, and 

 become reversed. If, for example, the metals be zinc and silver, the 

 current will decline at 248", then cease, and set in, in the opposite 

 direction as the temperature continues to rise. This probably arises 

 from the effect of heat on the crystalline structure of zinc. Iron and 

 antimony produce the same effect to a less extent. The thermo-electric 

 order of the metals is quite different from their voltaic order. It is 

 for a few of the principal metals as follows : 



Bismuth. 

 Platinum. 

 Lead. 

 Tin. 



Copper and Silver. 



Zinc. 



Iron. 



Antimony. 



but for further details, see Matthiessen, ' Phil. Trans.', 1858. 



Thermo-electric circuits have been formed between non-metallic 

 bodies. Thus the point of a heated cone of porcelain clay, brought 

 into contact with a cold cylinder of the same material, and connected 

 with the galvanometer by means of moistened cotton, caused a 

 deflection of the needle. This effect has been ascribed to the mutual 

 re-action of two portions of water of different temperatures. 



Tkermo-multiplier. Bars of antimony and bismuth A and B may be 

 arranged into a thermo-multiplier or thermo-electric pile, so that a 

 high temperature may be applied to one series of junctions and a low 

 one to the other. The current is then measured by its effect on the 

 galvanometer needle, as shown in the figure. By means of 

 a thermo-electric pile chemical effects may be produced, such aa 

 decomposing a solution of iodide of potassium, and with one hundred 

 pairs of platinum and iron wire, each 1 inch long and jjgth of 

 an inch in diameter, water acidulated with sulphuric acid has been 

 decomposed. Wires of iron and German silver make a very good 

 thermo-electric pile, but in all cases the current is feeble, and the size 

 of the elements do not add to the effect, except by increasing the 

 conducting power. The thermo-electric pile used by Nobili and 



Hot 



lira 



Cold, 



Melloni formed a very delicate measure of temperature. It consisted 

 of thirty-six pairs of bars of bismuth and antimony packed into a 

 small space, connected with a delicate astatic galvanometer. When 

 the alternate junctions of the bars at each end of the pile were covered 

 with lamp-black the pile was so sensitive to radiant heat as to be 

 affected by the warmth of the hand at a distance of 30 feet. Even 

 the amount of heat radiated by insects could be estimated by means of 

 this apparatus. Mr. Joule placed the instrument in vacuum, with a 

 steel magnet near it to diminish the action of the earth's magnetism, 

 and so far increased its sensibility, as to indicate by means of a single 

 pair of bismuth and antimony, a change of temperature amounting to 

 only the ^.th part of a centigrade degree. By means of this instru- 

 ment Melloni was able to prove the instantaneous transmission of heat 

 through glass and other bodies, and the thermo-transparency of rock- 

 salt. By its means Forbes was able to show the polarisation of heat. 

 The fineness of the wires used for the junction, together with its great 

 sensibility, allow the use of this instrument, where a common thermo- 

 meter could not be applied, as in measuring the temperature of the 



