[ 361 ] 
X. Effects of Stress and Magnetisation on the Thermoelectric Quality of Iron. 
By Professor J. A. Ewing, B.Sc., F.R.S.E., University College, Dundee. 
Communicated by Sir William Thomson, F.R.S. 
Received February 24,—Read March 11, 1886. 
[Plates 21-23.] 
§ 1. In May, 1881, the writer submitted to the Royal Society a paper entitled “Effects 
of Stress on the Thermoelectric Quality of Metals, Part I.,” an abstract of which was 
published in the ‘Proceedings,’ vol. 32 (1881), p. 399. The paper was described as 
incomplete, and its further publication was postponed until additional results should 
be submitted. The experiments were continued in 1882-3, in the physical laboratory 
of the University of Tokio, in conjunction with others on the magnetisation of iron. 
These latter, which were communicated to the Ptoyal Society in January, 1885, and 
are now being published in the ‘ Philosophical Transactions,’* have a very intimate 
relation to the subject of this paper, and the writer has for this reason deferred the 
publication of the thermoelectric experiments until the appearance of the experiments 
in magnetism. The present paper embodies the results of the paper referred to above 
as Part I. (sufficiently fully to make separate publication of that part unnecessary), 
along with those of subsequent work, in which the writer has to acknowledge the 
very valuable help of one of his Japanese students, Mr. S. Sakai. 
§ 2. In the earlier experiments the thermoelectric effects of stress on an iron wire 
were studied by exposing a piece of the wire to longitudinal pull, while one of the 
junctions between the stressed and the unstressed part was kept at a temperature of 
about 100° C., the other junction being at the atmospheric temperature, or about 
15° C. to 20° C. The wire was held at one end of the stressed portion by being 
twisted round a fixed hook. Various plans of heating this junction were tried; the 
plan finally selected, and followed throughout all the experiments, was to immerse the 
hook in a bath of hot oil, whose temperature was kept very nearly uniform bv a 
regulated flame under it. The other end of the stressed portion of the wire was con¬ 
nected to a cord which passed over a fixed pulley, and from which a light water-tank 
hung. Pull was applied to the wire by running water into the tank, and could be 
relaxed by allowing water to escape through a stopcock at the foot of the tank. 
Water was run in at a sensibly uniform rate, and the pull on the wire was given by 
* “ Experimental Researches in Magnetism,” Phil. Trans., vol. 176 (1885), p. 523. 
MDCCCLXXX VI. 3 A 
