450 Prof. Oliver Lodge 
reckoned from Tait’s experimental data of thermoelectric 
powers: the term “thermoelectric power” being defined, 
for a circuit of two metals, A and B, thus:— 
Ki e 
It is now easy to follow out the changes in magnitude and 
sign of the four forces I,, H,, ©,, and ©, for a circuit of any 
two metals as their junctions are varied in temperature ; and 
the diagrams (Plate III.), illustrative of twelve typical cases 
observable with an iron-copper circuit, are not without 
interest. 
The Peltier force acts from copper to iron for all tempera- 
tures below the neutral point 275° C.; at this point it vanishes, 
and above this point it is reversed : in other words, copper is 
the metal of higher thermoelectric power below 275°, but iron 
is, above. The Thomson force happens to act from hot to 
cold in iron, and from cold to hot in copper, always ; and it 
increases uniformly with the difference of the squares of the 
absolute end-temperatures. 
The temperature of each junction is indicated on the figures 
in Centigrade degrees. The arrows indicate the direction of 
the electromotive forces at the different places, and numbers 
attached to the arrows show their magnitude. The central 
arrow represents the resultant H.M.F., or current, in magni- 
tude and direction. The unit in which all the H.M.F.s are 
specified is the ten-thousandth of a volt. 
It will be observed that in cases 4 and 7 the current flows 
against the force at the hot junction; and that in cases 6 and 
9 it flows with the force at the cold junction, being helped on 
thereby. The common statement that the current flows from 
a metal of higher to a metal of lower thermoelectric value 
across the hot junction is thus by no means necessarily true. 
Its truth depends upon circumstances. It is not even true to 
say that the self-generated current necessarily cools the hot 
junction and warms the cold one. It may warm both ; it may 
cool both. It may even, in some peculiar cases, cool the cold 
junction and warm the hot: see below. 
These diagrams, though they express nothing but what is — 
well known, the numbers being calculated from Tait’s table 
of thermoelectric values as given in any textbook, yet are 
interesting from the point of view of the second law of thermo- 
dynamics, being some of them apparent exceptions to, or con- 
tradictions of, some statements of that law. 
It will be observed that the current usually flows with the 
force at the hot junction, and therefore cools it; but it may 
P;—P;= 
