20 REPORT — 1855. 



the ends of the iron conductor are connected with the electrodes of an astatic 

 needle galvanometer of very moderate sensibility, suddenly moving the coil from one 

 side to the other of the flame of the spirit-larap. 



The author next explained a series of experiments (not so easily described without 

 the apparatus which was exhibited to the Section, or drawings of it), by which it was 

 ascertained that magnetized iron, with electric currents crossing the lines of magnet- 

 ization at right angles, differs from unmagnetized iron, thermo-electrically, in the 

 same direction as bismuth, that is, in the opposite way to that previously found for 

 iron magnetized along the line of current ; and it was verified that an iron conductor, 

 obliquely magnetized, and placed in a circuit of conducting matter, has a current 

 excited through it when its two polar sides are maintained at different temperatures. 

 The author also described and exhibited an experimental arrangement nade, but 

 not yet sufficiently tried, to test whether or not magnetized iron possessed a certain 

 thermo-electric rotatory property which his theory of thermo-electricity in crystal- 

 line conductors had led him to believe might possibly exist in every substance 

 possessing, either intrinsically or inductively, such a dipolar directional property as 

 that of magnetism. 



Regarding the thermo-electric properties of magnetized steel, the only experiments 

 yet made, being on longitudinal magnetization, showed most decidedly the same 

 kind of effect subsisting with the permanent magnetization, after the magnetizing 

 agency is withdrawn, as had been found in iron while actually sustained in a state 

 of magnetization by the electro-magnetic force. 



The effects of magnetism on the conductivity of iron both for heat and electricitj', 

 in different directions with reference to the direction of magnetization, had been 

 tested by different experimenters with no confirmed indications in the conduction of 

 heat, and with only negative results regarding electric conductivity. The author of 

 the present communication, feeling convinced that only tests of sufficient power are 

 required to demonstrate real effects of magnetization on all physical properties of 

 iron, tried to ascertain the particular nature of the conjectured effect in the case of 

 electric conductivity ; and at last, after many unsuccessful attempts, succeeded in 

 establishing, that an iron conductor, sustained in a magnetic condition by a longi- 

 tudinal magnetizing force, and brittle steel wires retaining longitudinal magnetism, 

 resist the passage of electricity more, or, which is the same, possess less electric 

 conductivity, than the same conductors when unmagnetic. It remains to be seen 

 whether either iron or steel has, when magnetized, the electro-crystalline property of 

 possessing different electric conductivities in different directions ; and whether 

 either has the possible rotatory property as regards conduction, which the intrin- 

 sically dipolar type of magnetization suggests. 



It is important to observe, that both the thermo-electric quality, and the effect on 

 electrical conductivity induced in iron or steel, and sustained by the magnetizing 

 force, are retained with the permanent magnetism in steel after the magnetizing force 

 is removed, as Joule found to be the case with the alteration of dimensions, which 

 he discovered as an effect of magnetism ; while on the other hand, as the author 

 showed in a previous communication to the Section, the thermo-electric quality he 

 had discovered as an effect of mechanical strain, becomes reversed when the con- 

 straining force has been removed, if any permanent strain has been produced. 



On the Thermo- Electric Position of Aluminium. 

 By Professor W. Thomson, M.A., F.R.S. 



The author, through the kindness of Baron Liebig, having been enabled to make 

 experiments on a bar of aluminium with a view to investigating its thermo-electric 

 properties, found that it gave currents when its ends were at different temperatures, 

 and an inch or two of its length was included in the circuit of a galvanometer by 

 means of wires of copper, of lead, of tin, or of platinum, bent round it. These 

 currents were in such directions as to show that the Aluminium lies, in the 

 thermo-electric series, on the side towards bismuth, of Tin, Lead, Copper, and a 

 certain platinum wire (Pg) ; and, on the side towards Antimony, of another 

 platinum wire (P3). They were in the same direction as regards the higher 

 and lower temperatures of the two junctions of the aluminium with the other 



