^60 Dr Traill on Thermo-Magnetism 
depend on the position of the initial point of the disturbance of 
the equilibrium of temperature, but on the direction in which it 
arrives at the needle. 10. A bar of bismuth, under all circum- 
stances, gives opposite deviations to one of antimony. An ap- 
paratus constructed with a bar of the former metal, was of con- 
siderable power ; but the low melting point of bismuth renders 
it less suited to such experiments than antimony. 11. Various 
metals affect the needle as antimony, others as bismuth : Among 
the former are, silver and copper, zinc and copper, iron 
and copper : Among the latter are, platinum and copper, 
platinum and silver, lead and copper, brass and copper, 
Chinese tutenag and copper, & c. 12. When a piece of metal 
has one of its surfaces applied flatly to another equal metallic 
plate, a thermo-magnetic combination is formed ; less powerful 
indeed than when united as already described. Even a speck 
of solder on a piece of copper appears to produce this effect. 
13. Dr Traill was unable to form any thermo-magnetic combi- 
nation of one metal alone , when the pieces united were of equal 
purity ; but he found that two different copper wires, of 
unequal purity, gave slight indications of thermo-magnetism * 
14. Nor did he succeed in forming such combinations with one 
metal and water, nor with a metal and pieces of stony matter, 
though he seems to think, that a nearly neutralized needle 
might enable us to detect thermo-magnetism in large masses 
even of different kinds of stone. 
The very remarkable effect of helical connecting pieces in 
electro-magnetic combinations, induced Dr Traill to try their 
thermo-magnetical influence ; and he found a striking analogy 
between their effects in both cases ; the relative position of the 
zinc plates to the copper, in the former class of phenomena, 
producing similar effects on the needle as heating either extre- 
mity of the bar of antimony in the latter. A horizontal right 
helical connecting piece produces, with a bar of antimony, in- 
version of poles, when the N. end of the apparatus is heated ; 
* The author finds, that, though he was at first unable to produce thermo- 
magnetic effects by heating a single metal , yet the experiment of Bequeret taught 
him, that they could be shewn by heating the joined wires of a nice galvanoscope. 
This does not alter any of the hypothetical deductions of the memoir, but even 
render* the thermo-magnetism of the earth more easily conceived. 
