363 
of Edinburgh, Session 1879 - 80 . 
that of the water contained within it. In this way contact experi- 
ments with surfaces at different temperatures were made and results 
obtained. In the method which gave most reliable results, the upper 
surface was kept at the temperature of the air, while the temperature 
of the lower was allowed to vary as the water contained in the metal 
flask cooled through time. The contact, effected by lowering the 
upper plate upon the lower from a height of 5 inches, was instan- 
taneous, so that the temperature of the upper surface did not change 
during the operation ; while immediately before every such contact, 
both surfaces were carefully polished with emery paper and dusted, 
and their temperatures carefully observed. The best results were 
obtained when both the surfaces were of the same metal, as, for 
example, iron against iron. When that was the case there was, of 
course, no electrification by contact and separation when both plates 
were at the same temperature. When, however, the temperature of 
the lower surface was raised, a deflection on the electrometer scale 
was obtained, indicating a difference of electric potential at the sur- 
face of separation of these metal plates. Thus it was found that iron 
hot was negative to iron cold, copper hot negative to copper cold, 
zinc hot negative to zinc cold, and the same seemed to hold for tin. 
Not only so, however, but the difference of potential between, say, 
the two iron plates increased apparently with the difference of tem- 
perature between them, and increased uniformly. Curves were 
drawn out representing the variation of this potential difference with 
the temperature of the lower plate ; and the points entered clustered 
approximately round three straight lines representing the temperature- 
variations for iron, copper, and zinc respectively. The tangents of 
the angles of inclination of these lines to the temperature axis are 
given in the following table : — 
Metal. 
Tangent of 
Inclination. 
Copper, ...... '39 
Iron, -76 
Zinc, ’9 
Now it was proved by experiment in every case that this “nega- 
tive-growth ” of the metal surface when its temperature was raised 
was a 'permanent surface condition after the surface was cooled down 
to the same temperature as its fellow. Hence it follows that the 
main effect is not due to mere change of temperature, but to some 
2 x 
VOL. X. 
