56 Professor CumMinG on the Developement 
that of the single plate of bismuth*. A battery of 6 plates of 
bismuth, (Fig. 11.) gave at 140° a deviation of only 83°; but two 
plates alone of the same battery, gave a deviation of between 9° 
and 10°. This diminution of power arose, as appears by Table V. 
from some of the plates being inferior to the others; as is the 
case in the voltaic battery, when the plates are of different di- 
mensions or differently excited, the weaker having a tendency to 
reduce the others to their own standard. The greatest deviation 
I have as yet produced was by a double bar 7 inches long, com- 
posed of two bars, the one antimony, the other bismuth, soldered 
together at the middle. This, at the melting point of bismuth 
gave a deviation of 36° with the same compass which shewed 
a deviation of 28° with the zinc and copper rods excited by dilute 
muriatic acid. As these rods were capable, though slightly, of 
magnetizing a needle inclosed in a spiral wire, and of exciting 
the limbs of a frog, I endeavoured, but without success, to effect 
the same by the double bar of antimony and bismuth. Whether 
this failure were owing to a want of sufficient action, or to some 
peculiarity in the electro-magnetism excited by heat must be de- 
termined by the application of a more powerful apparatus; but 
it does not seem improbable that it may be owing to some pecu- 
liarity in this mode of excitationt. There is something perhaps 
analogous in common electricity; a large battery may be dis- 
* The bars of bismuth and antimony, having their extremities connected by copper 
wires, were fixed in a wooden trough, in such a manner that one half of their length 
projected below it. The trough was then filled with hot water or sand, and placed upon 
another vessel containing cold water. The mean of four thermometers in the upper part 
was assumed as the temperature of the heated extremities, and those experiments alone 
retained, in which the thermometers did not differ materially from each other. As the melting 
point (475°) of bismuth, the more fusible metal, is at nearly the temperature (500°) of boiling 
linseed oil, it is evident that a considerable increase of power may be obtained, if the trough 
be filled with ice, and placed.upon a vessel of oil heated nearly to ebullition, 
+ Possibly from its low intensity, 
