qn & 
Cage akan SPV 
reactant alga edi 
nena 
RESEARCHES IN CONTACT ELECTRICITY. 275 
ture of the lower surface only was made to vary, so that the surfaces were 
generally at different temperatures. By the former method it was found that 
the difference of potential between polished iron and polished copper fell off 
by at least 45th of its original value for a rise in temperature of 1° C. Many 
series of experiments were made with these two metals, and each day’s results 
gave the same general indication ; although, as might have been expected from 
the nature of the inquiry, it is hardly possible to deduce from them any definite 
quantitative law. 
The general results of eleven series of experiments are given in the follow- 
ing table. The first column represents the lowest temperature for which 
readings were taken ; the second gives the electrometer deflection for that 
temperature due to the electrification by contact of the lower surface ; the 
third indicates the like deflection for the higher temperature; the fourth 
registers that higher temperature ; and the fifth notes the percentage average 
decrease of the deflection for unit increase of temperature. 
T Lower _ Deflection. Deflection. eres Eezcanlage 
emperature. | Temperature. | Decrease. 
Tee 70 50 30°C. 1:32 
14 60 35 40 1:60 
13 78 58 45 83 
10 ie 56 45 78 
18 110 80 50 85 
20 93 76 41 87 
23 110 85 50 84 
16 110 60 48 1:42 
16 110 85 35 1:22 
20 112 91 38 1:04 
16 . 112 90 36 98 
The first four experiments give smaller readings than the last seven—a 
discrepancy easily accounted for by the change of circumstances occasioned by 
removal to another room, and a refitting of the surfaces. Yet, that in such 
altered circumstances the average percentage temperature-variation of the 
deflection should be so consistent throughout, argues strongly in favour of the 
reality of such a variation. A like series of experiments was commenced with 
a zine surface substituted for the copper or under surface ; but, though there 
were indications of a somewhat similar variation, these were too vague to 
admit of any definite deductions being made. The same was true of the 
aluminium-zine pair. In this mode of experimenting, however, it was impos- 
sible to determine how much of the resultant variation of a given pair was due 
to the action of a particular component, or how far this variation depended 
directly upon the change of temperature, or indirectly through consequent 
material alteration of the surfaces—through oxidation, for example. 
