272 DR CARGILL G. KNOTT ON 
extremities may, and frequently does, become a function of the intermediate 
structure, and is then no longer equal to the direct contact-difference between 
the extreme members.* With a circuit including such materials in its com- 
position, the resultant total electromotive force does not in general vanish, so 
that the existence of a current is possible and necessary; and the energy of this 
current is derived from the energy of chemical combination, which is the one 
aspect of the accompanying action, whose other aspect is the decomposition 
of the compound conductor. Except such chemical action were possible no 
current could be generated; so that, probably, the possibility of chemical 
action, and the non-vanishing character of the resultant electromotive force in 
the circuit, are necessarily co-existent phenomena. Such, at present, seems to 
be the most complete theory of the voltaic cell. 
Although no current exists in a circuit of simple conductors maintained at 
a uniform temperature because of the mutual balancing of the contact forces, 
it is possible to cause a current to flow by heating or cooling one of the 
junctions, and thereby destroying the equilibrium of the contact forces. The 
energy of the thermo-electric current so obtained is a partial transformation of 
the energy which was originally expended in unequalizing the temperature of 
the system. Apparently, then, the contact-difference of potential between two 
metals or other simple conductors depends upon the temperature—a conclusion 
verified in a very remarkable way by consideration of the Peltier effect, or 
reversible thermal effect, produced by the passage of a current across the 
junction of two different metals. By an application of the dynamical theory of 
heat, Sir W1LL1AmM Tuomsont proved that this evolution or absorption of heat at 
the junction, according as the current flowed in one ‘or other direction, indicated 
an electromotive contact force, acting against the current when heat was 
evolved, with the current when heat was absorbed. In other words, because of 
the difference of potential at the junction, the current has to do work when 
passing in one direction, and has work done upon it when passing in the other— 
giving rise respectively to an evolution and absorption of heat. From considera- 
tion of the principle of dissipation of energy, Professor Tarir{ has developed a 
formula for this electromotive contact force, expressing it as a parabolic 
function of the temperature ; and this theory has been indirectly verified by a 
long series of experiments upon the thermo-electric properties of metals, 
With a view of testing by direct contact experiments the variation with 
temperature of the contact-difference of potential between dissimilar metals, I 
undertook the experiments whose results form the subject of this thesis, It 
must be premised, however, that any positive result cannot be regarded as due 
* See the papers of KonLrauscn, Hanken, Crirron, and Ayrton and Perry, cited above, 
+ Transactions of the Royal Society of Edinburgh (1851). 
{ Ibid, (1870-71). 
