54 Professor CumMING on the Developement 
of the same dimensions, and under the same circumstances, gave 
deviations of 65° positive, and 4° negative respectively. These are 
selected as being some of the most remarkable results, and as 
serving, at the same time, to form a comparison between the 
effects of electro-magnetism as excited by heat, and by the usual 
process. The results of similar experiments on -some metals and 
metallic alloys are given in Table IT. 
The metals which were found to be most powerful im their 
action, and at the same time most readily prepared for experi- 
ment, were bismuth and antimony. These were sufliciently ener- 
getic to have their effects estimated by a compass needle 4 inches 
in length, the terrestrial magnetism of which was not neutralized, 
and therefore the deviations caused by different increments of tem- 
perature may be employed as data for the corresponding electro- 
magnetic action. With this compass, rods of zine and copper, 
each = inch diameter, and ; apart, excited by equal parts of mu- 
riatic acid and water, gave a deviation of 27°. A circular rod of 
bismuth 43 inches long and + inch diameter, gave a deviation of 
21° at the melting point of bismuth, 12° at the temperature of 180°, 
and 5° at 100°, the cooler extremity being in water at 60°. A si- 
milar bar of antimony: gave 19° with the utmost heat of a spirit 
lamp, and a bar of platina which weighed 565 grains, and was 
7 inches long by 4 broad, gave 19° at a red heat. Hence it appears, 
that, at the same temperature, bismuth is more powerful than an- 
timony, and antimony more powerful than platina: the other metals, 
with the exception perhaps of iron, were inferior to platina. 
Were it possible to acquire the same accumulation of power 
in this species of electro-magnetism as is obtained by the use of 
large plates in the galvanic, or by numbers in the voltaic appa- 
ratus, it would be an interesting object of research to compare 
the electricities thus differently excited. For this purpose bars 
were cast, differing both in length, breadth, and thickness, but 
