123 
1897 - 98 .] Thermodynamics of Volta-contact Electricity . 
Hence 
dm(g, t) _t - T 
dt 
t) 
0 ) 
From (1) and (2) with the equation 4L and f rom 
v ' dg dt dt d£ 
(3) and (4) with the equation 41 = 4i we find 
d£ dt dt d£ 
and 
/(QV) 
( 5 ) 
( 6 ) 
Eliminating between (5) and (6) we find finally 
Ct£ 
dV 
dt 
or, 
JQ = 
dV 
d( log t) 
( 7 ) 
§ 10. The quantity of heat absorbed or produced in virtue of 
change of electric density on the surfaces II, CC must in all proba- 
bility in every case be too small to be detected by direct observation. 
But the difference between the quantities produced at the two sur- 
faces of the two different metals, per unit of electric quantity added 
to one surface or taken from the other, which is Q — H if H be the 
Peltier effect at the junction J, can by aid of equation (7) be 
readily and surely determined for any two metals for which the 
Peltier effect is known. In fact, it is easy to arrange apparatus 
for measuring V through a considerable range of temperature, say 
from 0° to 100° C., by the now well-known compensation method 
introduced independently by Pellat and myself, and thus to measure 
and so find Q by equation (7). I am at present commencing 
experiments for this purpose with air, and with carbonic acid gas, 
between the two opposed metal surfaces, so that we may judge 
what precautions, if any, will be necessary to eliminate disturb- 
ances due to different condensations of gas on the metals at 
different temperatures. Very interesting and important experi- 
ments by Pellat and by Erskine Murray have shown large tern- 
