REPORT ON THERMO-ELECTRICITY. 307 
poles on the northern, and two poles similarly posited on the 
southern side; the poles of different names being opposed to 
each other on the contrary sides of the equator. 
I cannot refuse myself the pleasure of quoting another result 
from an apparatus still more resembling the connexion of the 
earth and its atmosphere. When heat was applied to a point in 
the equator of a copper shell surrounding a sphere of bismuth, 
it resulted that the deviation of the end of the needle of the 
same name as the latitude, was always towards the west when 
the place of heat was above the horizon, and towards the east 
when it was on the meridian below . It is unnecessary for me 
to say to this audience, how perfectly this and the preceding 
results accord with observation. 
I should be happy, if it were in my power, to bring before 
you a series of researches of equal interest with this paper by 
Mr. Christie; but, as I am not aware of any of recent date 
that are worthy of your notice, it only remains for me to point 
out to you what I conceive to be the present state of our know¬ 
ledge of thermo-electricity. 
Thermo-electricity may be developed in a homogeneous 
metallic mass, or in two distinct masses of the same metal un¬ 
equally heated at their point of contact. In the former case 
the direction of the electricity in the same metal is influenced 
by the figure; in the latter, as when a metallic disk is touched 
by a heated wire, the action is anomalous, varying with the 
metal. 
A metallic bar affords as many circuits as there are portions 
into which it can be divided. Therefore dividing it indefinitely, 
it may be conceived that each particle has its distinct current, 
the positive and negative electricities being separated by heat. 
The currents between the extremities of a metallic bar are not 
affected by varying the directions of the intermediate currents. 
Charcoal, plumbago, and some, possibly many, of the metallic 
sulphurets, are capable of thermo-electric excitation. This spe¬ 
cies of electricity is therefore not confined to the simple metals. 
The metals compose a thermo-electric series of which bismuth 
and antimony are the extremes, and are the most efficacious 
within certain limits. 
A battery may be formed of pairs of such elements, the action 
of each depending in some degree on the surfaces in contact; 
and the effect is greater when they are arranged in sequence, 
as in the voltaic series, than when connected as in Hare’s 
calorimeter; but in neither case increases in the same ratio as 
the number of elements. 
The ratio between the temperatures and the corresponding 
u 2 
