RESEARCHES IN CONTACT ELECTRICITY. 273 
only to the metals ; for, as pointed out by Professor CLERK MAXwWELL,* VoLta’s 
electromotive force of contact is in general greater than that indicated by the 
Peltier effect, and sometimes of opposite sign—a discrepancy to be accounted 
for by the fact that in direct contact experiments there is always a film of con- 
densed air or other gas between the metals when they are in so-called contact, 
and that possibly the chief effect ‘must be sought for, not at the junction of 
the two metals, but at one or both of the surfaces which separate the metals 
from the air or other medium which forms the third element of the circuit.” 
After a few preliminary experiments I concluded that direct contact of the 
surfaces under investigation was a sufficiently accurate and constant method 
for indicating any appreciable change which might occur. It was found neces- 
sary, however, to keep the surfaces continually polished in a particular manner, 
since they gradually altered their surface condition when exposed to the action 
of the air—a fact formerly established by HANKEL.+ Previous to any discus- 
sion of the results obtained, it is advisable first to give a description of the 
apparatus and method of experiment. 
Of the two metallic surfaces which were the subject of experiment the 
lower was the upper surface of a flat cylindrical flask-shaped vessel, which 
rested on an insulated stand in electric connection with one pair of opposite 
quadrants of a THomson Quadrant Electrometer. The temperature of this 
surface was determined by the temperature of the water contained in the flask. 
Three such flasks were used—one of iron, one of zinc, and the third with the 
one flat surface copper and the other tin. The diameters of the plane faces, 
the thicknesses of the flasks, and their volume capacities, are as follows :— 
Flask Diameter in | Thickness in | Volume in Cubic 
ee Millimetres. | Millimetres. | Millimetres. 
ron ater, f : 128 uy 198,000 
Zin 0 ‘ ‘ 131 16 168,000 
Copper and Tin, . 129 16 196,000 
The upper plate of the condenser was a disk of approximately the same 
area as the lower, on which it pressed during contact by its own weight. It 
could be readily adjusted by screws to a practically accurate parallelism with 
the lower plate, and had only one degree of freedom—an up-and-down motion 
directed by a pin and guiding slot. It depended from the brass top of a 
cylindrical glass case which surrounded the insulated stand and flask on all 
sides, if we except the small aperture through which the internal arrangement 
was put in connection with the electrometer. Great care was necessary in dry 
* Electricity and Magnetism, vol. i. § 249. 
+ Pogg. Ann., vol. exxvi. p. 286 (1865), 
