980 REPORT—1885. 
fall far short of 1-5 volt, being in certain cases barely appreciably greater than 0. It 
would thus seem that whilst the heat evolved during the condensation of 1 gramme 
of hydrogen and of 8 grammes of oxygen to the physical state of the gases occluded 
or attracted in the form of these films must jointly amount to something close to 
34,100 gramme degrees at least, the energy requisite to break up water not into 
ordinary free oxygen and hydrogen, but into the nascent forms of these bodies, 
unmodified by subsequent coalescence or spontaneous modification must be equal 
to something not far short of three times that amount; whilst the total heat 
evolved during the subsequent modification of these elements (passage from atomic 
to molecular condition ?) must represent at least 2°5 volts. 
A large number of observations have been made (mostly published) on the 
relationships between the positive E.M.F.s generated in various kinds of voltaic 
cell when at their maximum, and the amounts of heat corresponding with the nett 
chemical changes ensuing between the electrolytes and electrodes. The general 
result of these investigations has been to show that in but few instances are the two 
values approximately coincident (say within a departure of +01 volt), and even 
in these cases very material fluctuations in the actual E.M.F. of an electromotor 
may be brought about by alterations in solution, strength, and plate surface nature 
that exert but little influence on the nett heat development. Oue conclusion 
deducible from this and other allied facts is that the seat of primary action in such 
cases lies at the junctions of the surfaces of the electrodes and fluids, and that the 
modus operandi of a voltaic cell is in certain respects closely akin to that of a 
thermo-couple or Peltier couple, consisting of two dissimilar forms of solid matter 
(metals, &c.) where conversion of sensible heat into current takes place, or vice 
versa. In fact the actual maximum E.M.F-.s set up in such cells as those examined 
(mostly after Daniell’s construction, e.g., Zn | ZnSO, | CuSO, | Cu) are found to: 
be conveniently represented by assigning to each given metal immersed in a given 
solution of one of its salts a numerical value, or thermo-voltaic constant (analogous 
to the thermo-electric values assignable to metals, &c., when used in thermo- 
electromotors), and adding the algebraic difference between the values for any two 
given pairs of metal and salt (which difference may be a positive or negative 
quantity) to the value in volts corresponding with the heat evolution due to the 
nett chemical change. According as this difference is plus or minus, the E.M.F. 
actually set up exceeds or falls short of the amount due to the nett chemical 
change. Relatively to zinc, some metals have thus in general a negative and 
others a positive thermo-voltaic constant assignable, no matter what the class of 
salt employed. The cells where the algebraic difference between the thermo- 
voltaic constants is positive in sign are characterised by the peculiarity that in the 
production of a current by them sensible beat must become converted into current 
energy, causing cooling of the cell; for with a large external resistance more work 
is done outside the cell than corresponds with the heat development due to the 
nett chemical change. With cells where this difference is negative in sign, the 
reverse holds so long as the numerical value of the difference is not greater than 
that corresponding with the difference in heat of formation between the two fluids 
surrounding the plates; the cell in this case being warmed by the passage of the 
current, and less external work being done than corresponds with the nett heat 
development due to chemical change. When, however, the numerical value of the 
negative algebraic difference between the thermo-voltaic constants exceeds that 
corresponding with the difference in heat of formation, the remarkable result ensues 
that the current circulates in the direction opposite to that predicable from the 
relative heats of formation of the fluids in the cell; so that not only is there 
absorption of heat in the cell itself, due to the character of the chemical changes 
occurring as the current passes, but, further, any work done outside the cell must 
be at the expense of the sensible heat of the cell itself. 
