fir Traill on Thermo-Magnetism. £59 
tions of the heated end, with regard to the magnetic meridian, 
was with facility ascertained, by laying it on a small wooden 
stand with a revolving top ; so that by a single application of a 
spirit lamp to one of the extremities of the apparatus, its 
thermo-magnetism in different positions could be easily deter- 
mined. 
The following are the general results of the experiments de- 
tailed in this part of the paper. 
1. Of all metals, antimony forms the most powerful combi- 
nation in thermo-magnetic experiments. £. The deviation of 
a finely poised needle, was in general greater when placed with- 
in , than without the rectangle. 3. In any one position of the 
heated extremity, the direction of the deviations is the same on 
any one of the inner surfaces of the rectangle ; and w T as uni- 
formly in the opposite direction on any of its outer surfaces. 
4. It is not necessary to place the support of the needle in con- 
tact with the metals, A suspended needle, or the interposition 
of a thick plate of glass, did not prevent the deviations caused 
by any of the surfaces. 5. The upper surface of the bar of 
metal produced deviations in the same direction, whatever may 
be the relative position of the connecting piece. 6. The direc- 
tion of the deviation is different when opposite extremities of 
the apparatus are heated. 7. In the same relative position of 
the two metals, and on their similar surfaces, the points on each 
side of the meridian which give the greatest deviations, are ex- 
actly opposite to those which give the least. Thus, if the 
heating of the W. end of the apparatus give a deviation equal 
to 180°, then the East end heated will not alter the direction of 
the needle within the rectangle, and vice versa . 8. The extent 
of the deviations in all cases depends not only on the nature of 
the metallic bar emp^ed, but also on its position with regard 
to the magnetic meridian. When the apparatus lies at right 
angles to the magnetic meridian, the deviations are the greatest. 
With antimony, in this position, the needle was completely in- 
verted ; but, for this effect the needle must be placed with- 
in the rectangle when the W. end is heated, and without it 
when the E. end is heated. 9. Experiments with the bars of 
antimony, bent at right angles, and heated at different points* 
shew that the direction of the magnetic deviation did not 
