It has already been observed, that three bodies are requisite to form a galvanic circuit, one of which must be fluid; but in 1822, Professor Seebeck of Berlin, discovered that electric currents may be produced by the partial application of heat to a circuit formed of two solid conductors. For example, when a semicircle of bismuth, joined to a semicircle of antimony, so as to form a ring, is heated at one of the junctions by a lamp, a current of electricity flows through the circuit from the antimony to the bismuth, and such thermo-electric currents produce all the electro-magnetic effects. A compass needle placed either within or without the circuit, and at a small distance from it, is deflected from its natural position, in a direction corresponding to the way in which the electricity is flowing. If such a ring be suspended so as to move easily in any direction, it will obey the action of a magnet brought near it, and may even be made to revolve. According to the researches of M. Nobili, the same substance unequally heated, exhibits electrical currents. The experiments of Professor Cumming show that the mutual action of a magnet and a thermo-electric current, is subject to the same laws as those of magnets and galvanic currents, consequently all the phenomena of repulsion, attraction, and rotation may be exhibited by a thermo-electric current. It is, however, so feeble, that neither heat, the spark, nor chemical action have been observed, nor can repulsion, attraction of light substances at sensible distances, or any other effects of tension, be perceived.

In all the experiments hitherto described, artificial magnets alone were used, but it is obvious that the magnetism of the terrestrial spheroid which has so powerful an influence on the mariner's compass, must also effect electrical currents. It consequently appears that a piece of copper wire bent into a rec

tangle, and free to revolve on a vertical axis, arranges itself with its plane at right angles to the magnetic meridian, as soon as a stream of electricity is sent through it. Under the same circumstances a similiar rectangle, suspended on a horizontal axis at right angles to the magnetic meridian, assumes the same inclination with the dipping needle. So that terrestrial magnetism has the same influence on electrical currents as an artificial magnet. But the magnetic action of the earth also induces electric currents. When a hollow helix of copper wire, whose extremities are connected with the galvanometer, is placed in the magnetic dip, and suddenly inverted several times, accommodating the motion to the oscillations of the needle, the latter is soon made to vibrate through an arc of 80° or 90°. Hence it is evident, that whatever may be the cause of terrestrial magnetism, it produces currents of electricity by its direct inductive power upon a metal not capable of exhibiting any of the ordinary magnetic properties. The action on the galvanometer is much greater when a cylinder of soft iron is inserted into the helix, and the same results follow the simple introduction of the iron cylinder into, or removal out of the helix. These effects arise from the iron being made a temporary magnet by the inductive action of terrestrial magnetism, for a piece of iron, such as a poker, becomes a magnet for the time, when placed in the line of the magnetic dip.

M. Biot has formed a theory of terrestrial magnetism upon the observations of M. de Humboldt as data. Assuming that the action of the two opposite magnetic poles of the earth upon any point is inversely as the square of the distance, he obtains a general expression for the direction of the magnetic needle, depending upon the distance between the north and south magnetic poles; so that if one of these quantities varies, the corresponding variation of the other will be known. By making the distance between the poles vary, and comparing the resulting direction of the needle with the observations of M. de Humboldt, he found that the nearer the poles are supposed to approach one to another, the more did the computed

and observed results agree; and when the poles were assumed to coincide, or nearly so, the difference between theory and observation was the least possible. It is evident, therefore, that the earth does not act as if it were a permanently magnetic body, the distinguishing characteristic of which is, to have two poles at a distance from one another. Mr. Barlow has investigated this subject with much skill and success. He first proved that the magnetic power of an iron sphere resides in its surface; he then inquired what the superficial action of an iron sphere in a state of transient magnetic induction, on a magnetised needle, would be, if insulated from the influence of terrestrial magnetism. The results obtained, corroborated by the profound analysis of M. Poisson, on the hypothesis of the two poles being indefinitely near the centre of the sphere, are identical with those obtained by M. Biot for the earth from M. de Humboldt's observations. Whence it follows, that the laws of terrestrial magnetism deduced from the formulæ of M. Biot, are inconsistent with those which belong to a permanent magnet, but that they are perfectly accordant with those belonging to a body in a state of transient magnetic induction. It appears, therefore, that the earth is to be considered as only transiently magnetic by induction, and not a real magnet. Mr. Barlow has rendered this extremely probable by forming a wooden globe, with grooves admitting of a copper wire being coiled round it parallel to the equator from pole to pole. When a current of electricity was sent through the wire, a magnetic needle suspended above the globe, and neutralized from the influence of the earth's magnetism, exhibited all the phenomena of the dipping and variation needles, according to its positions with regard to the wooden globe. As there can be no doubt that the same phenomena would be exhibited by currents of thermo, instead of voltaic, electricity, if the grooves of the wooden globe were filled by rings constituted of two metals, it seems highly probable that the heat of the sun may be the great agent in developing electric currents in or near the surface of the earth, by its action upon the substances of which the

globe is composed, and, by the changes in its intensity, may occasion the diurnal variation of the compass and the other vicissitudes in terrestrial magnetism evinced by the disturbance in the direction of the magnetic lines, in the same manner as it influences the parallelism of the isothermal lines. That such currents do exist in metalliferous veins appears from the experiments of Mr. Robert Fox in the Cornish copper-mines. However, it is probable that the secular and periodic disturbances in the magnetic force are occasioned by a variety of combining circumstances. Among others, M. Biot mentions the vicinity of mountain chains to the place of observation, and still more the action of extensive volcanic fires, which change the chemical state of the terrestrial surface, they themselves varying from age to age, some becoming extinct, while others burst into activity.

It is moreover probable that terrestrial magnetism may be owing, to a certain extent, to the earth's rotation. Mr. Faraday has proved that all the phenomena of revolving plates may be produced by the inductive action of the earth's magnetism alone. If a copper plate be connected with a galvanometer by two copper wires, one from the centre and another from the circumference, in order to collect and convey the electricity, it is found that, when the plate revolves in a plane passing through the line of the dip, the galvanometer is not affected; but as soon as the plate is inclined to that plane, electricity begins to be developed by its rotation; it becomes more powerful as the inclination increases, and arrives at a maximum when the plate revolves at right angles to the line of the dip. When the revolution is in the same direction with that of the hands of a watch, the current of electricity flows from its centre to the circumference; and when the rotation is in the opposite direction, the current sets the contrary way. The greatest deviation of the galvanometer, amounted to 50° or 60°, when the direction of the rotation was accommodated to the oscillations of the needle. Thus a copper plate, revolving in a plane at right angles to the line of the dip, forms a new electrical machine, differing from the common plate

glass machine, by the material of which it is composed being the most perfect non-conductor; besides, insulation, which is essential in the glass machine, is fatal in the copper one. The quantity of electricity evolved by the metal does not appear to be inferior to that developed by the glass, though very different in intensity.

From the experiments of Mr. Faraday, and also from theory, it is possible that the rotation of the earth may produce electric currents in its own mass. In that case, they would flow superficially in the meridians, and if collectors could be applied at the equator and poles, as in the revolving plate, negative electricity would be collected at the equator, and positive at the poles; but without something equivalent to conductors to complete the circuit, these currents could not exist.

Since the motion, not only of metals but even of fluids, when under the influence of powerful magnets, evolves electricity, it is probable that the gulf stream may exert a sensible influence upon the forms of the lines of magnetic variation, in consequence of electric currents moving across it, by the electro-magnetic induction of the earth. Even a ship passing over the surface of the water, in northern or southern latitudes, ought to have electric currents running directly across the line of her motion. Mr. Faraday observes, that such is the facility with which electricity is evolved by the earth's magnetism, that scarcely any piece of metal can be moved in contact with others without a developement of it, and that consequently, among the arrangements of steam engines and metallic machinery, curious electro-magnetic combinations probably exist, which have never yet been noticed.

What magnetic properties the sun and planets may have, it is impossible to conjecture, although their rotation might lead us to infer that they are similar to the earth in this respect. According to the observations of M. M. Biot and GayLussac, during their aërostatic expedition, the magnetic action is not confined to the surface of the earth, but extends into space. A decrease in its intensity was perceptible, and as it