124 
Lord Kelvin on 
[r.s.e., sess. 
perature effects on the Yolta-electric-force between two plates, 
whether of the same or of different metal, when one of them is 
heated and the other kept at the ordinary atmospheric tempera- 
ture ; hut this does not supply what we now want from experi- 
ment, which is, the variation of the Yolta-electric-force between 
two metals at one and the same temperature, when this tempera- 
ture is varied. 
Addition, of date 26th March 1898. 
A number of preliminary experiments, carried on with the 
assistance of Mr W. Craig Henderson, have shown large but 
largely irregular temperature variations of the Yolta E. M. E. 
between copper and zinc, with all parts of the Yolta circuit at the 
same temperature. The substitution of carbonic acid gas for air, 
and re-introduction of air for carbonic acid, seemed to make but 
little difference on the results. The irregularities seemed to have 
been chiefly due to permanent or sub-permanent changes in the 
copper plate. At all events we found somewhat more nearly 
regular results with zinc and gold. The zinc plate used had 
never, so far as I know, been polished or much disturbed by touch- 
ing its surface since experimented on by Mr Erskine Murray * 
three years ago. The “ gold ” was a brass plate gilded for me 
about 1859. It was one of the two “standard gold” plates in 
Erskine Murray’s experiments, and, so far as I know, it has never 
been rubbed or polished since 1861. 
We found in a range from 16° C. to 50° C. an augmentation of 
Yolta-contact difference at rough average rate of about *002 of a volt, 
or ’2 x 10 6 C.G.S. units, per degree Centigrade. This is 800 times 
the thermo-electric difference of zinc and gold given as 250 C.G.S. 
units per degree Centigrade in Jenkins’ Electricity and Magnetism , 
p. 176, and Everett’s Physical Units , 1886, p. 173. And (§ 3 
above) gold, zinc are as Y, X in respect to orders in .the Yolta 
series and in the thermo-electric series of metals. Hence, according 
to the secure thermo-dynamic formula (7) above, and the old 
probable thermo-dynamic hypothesis for thermo-electricity (§ 4 
* A description of Murray’s experiments will be published probably in May 
in the Phil . Mag . , and in the Proc. Roy. Soc. Lond. 
