698 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
thermometers was most nearly steady during a few minutes of flow of the current, 
first in one direction and then in the other. 
Current entering Readings, in arbitrary scale-divisions, of 
by end next 
Thermometer A. 
Thermometer B. 
1. (A 
43 
1b 
42 
441 
II. (A 
41 
41 
Ib 
41 
341 
III. { A 
31i 
261 
Ib 
291 
16i 
IV. (A 
m 
22 
Ib 
201 
Hence the conclusion (see below, §§ 102 and 103), that the Vitreous Electricity 
carries heat with it in brass, which I anticipated three years ago from the mechanical 
theory *, is now established by a direct experimental demonstration. 
PART II. ON THERMO-ELECTRIC INVERSIONS. 
78. Cumming’s discovery of thermo-electric inversions having afforded the special 
foundation of that part of the theory by which I ascertained the general fact of 
electric convection in metals, and every observation of a thermo-electric inversion 
being a perfect test as to the relative positions of the two metals between which it is 
observed in the Table of Convections (see below, § 103), I was induced to make experi- 
ments with a view to finding new instances of inversion, and to determine in each case, 
with some degree of precision, the temperature at which the two metals are thermo- 
electrically neutral to one another. 
79. In the experiments on thermo-electric inversion described by Gumming, and 
by Becquerel, the only other experimenter, so far as I am aware, who has published 
observations on the subject, one junction between the two metals is generally kept 
cool, while the other is raised until the current indicated by the galvanometer, 
instead of going on increasing, begins to diminish, comes to a stop, and then sets in 
the reverse direction-j'. 
80. In this way Gumming found that “ if gold, silver, copper, brass, or zinc wires 
be heated in connexion with iron, the deviation [indicating the current], which is at 
first positive, becomes negative at a red heat]:.” Many other experimenters have 
professed themselves unable to verify these extraordinary results, and have attempted 
* See “ Dynamical Theory of Heat,” § 132. 
t Cumming’s ‘Electro-Dynamics,’ section 104, p. 193. Cambridge, 1827. 
I Cambridge Philosophical Transactions, 1823, addition to p. 61. 
