278 : DR CARGILL G. KNOTT ON 
the electrometer deflection was, as far as the method admitted of judg- 
ing, the same. Probably after a longer lapse of time than that here speci- 
fied, a change might become manifest—such a change as HANKEL long ago 
established for iron and other metals at the ordinary temperature of the 
air. In order to compare this time-variation of surface condition with the 
temperature-variation established above, I made a series of observations, 
at sufficiently distant intervals of time, of the deflections produced by con- 
tact and separation of two iron surfaces, one of which was kept constant 
by polishing, while the other was permitted to vary, by being simply left to 
itself. Both were initially polished to be the same electrically—a state of 
affairs evidenced by the absence of any effect on the electrometer when the 
two plates were separated after contact. Readings were first taken at intervals 
of five minutes, then at intervals of ten minutes, fifteen minutes, and finally 
at half hour intervals. Each number in the following table is the mean of 
five readings taken in rapid succession within the lapse of one minute. 
EXPERIMENT X. (May 20). 
(Fig. 1, 0). 
Time (in minutes), Deflection (iron against iron). 
0 0 
5 —11 
10 —14 
15 —15 
20 —16 
30 —18 
45 —19°4 
75 —20 
The curve corresponding to these numbers is given in the diagram (fig. 1, 0). 
In its main characteristics it is very similar to an ordinary curve of cooling, 
and is markedly dissimilar to the curve which represents the temperature- 
variation of surface condition, Curves @ and ¢ on the same diagram indicate 
the corresponding variations for copper and aluminium respectively. The 
copper was electrified by contact with iron, both surfaces being allowed to 
vary; and the real time-variation of the copper was obtained by properly 
introducing the ‘nown time-variation of the iron. The aluminium was elec- 
trified by contact with polished zine, to which it was originally positive, but in 
the course of half an hour became as strongly negative. The contacts were 
instantaneous, and except immediately before the taking of a reading the 
surfaces were kept far apart. The tabulated values for these metals are given 
below, the chemical symbol for each metal being employed to represent the 
corresponding surface, and the suffix p signifying that the surface to which it 
is suffixed was kept polished and therefore constant. 
’ 
