PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 711 
permanently strained by traction, and probably all non-brittle (or plastic) transpa- 
rent solids when permanently strained otherwise than by uniform condensation or 
dilatation in all directions, possess double refraction as a property of the molecular 
alteration which they acquire under the stress and retain after the stress is removed. 
Again, magnetization, as Joule discovered, causes an elongation of iron in one direc- 
tion (that of the magnetization) and a contraction in all directions perpendicular to it, 
with no sensible change of volume. Faraday discovered the wonderful dipolar optical 
property of transparent bodies in a magnetic field (the first and only case known of 
any dipolar qualities, other than those of magnetic and electric reactive forces, called 
into existence by induction): Maggi discovered that magnetized iron conducts heat 
with a greater facility across than along the lines of magnetization*. 
106 . In applying the dynamical theory of heat to thermo-electric currents in con- 
ducting crystals, I was led to consider the probable effects of mechanical strain, and 
of magnetization on the thermo-electric properties of non-crystalline metals, and in 
consequence entered on the investigation, of which the results, so far as I have yet 
advanced in it, are now laid before the Royal Society. 
107. To find the effect of longitudinal tension on the thermo-electric quality of a 
metal, I first took eight thin copper wires each capable of bearing about 10 lbs., and, 
attaching their upper ends to a horizontal wooden arm at distances of about ^ of 
an inch from one another, allowed them to hang down, each kept stretched by a 
weight of about |^lb. They were connected with one another in order, and the first 
and last with the electrodes of a galvanometer, 
by nine wires soldered to them, as shown in 
the diagram ; the junctions between the suc- 
cessive wires being alternately in the upper 
and lower of two horizontal lines 4 inches 
apart. Every alternate wire was then stretched 
with a weight of about 3 lbs., and a slip of hot 
plate glass was applied, sometimes to the upper 
and sometimes to the lower row of junctions. 
A deflection of the galvanometer needle was observed in one direction or the other, 
according as the glass heater was applied to one set of junctions or the other. The 
deflection was also reversed when the weights were changed to the alternate set of 
wires, and the heater kept applied to the same set of junctions. In every case the 
deflection was such as to indicate a current from stretched to unstretched through 
hot junctions. The uniform and consistent nature of the indications was such as 
could leave no doubt as to the result ; and I concluded that copper wire stretched 
* Doubts have been thrown on this result, I believe, by other experimenters, who have not succeeded in 
verifying it by their own observation, but its close correspondence with a result I have recently discovered by 
experiments on the electric conductivity of magnetized iron, have diminished the impression such doubts pro- 
duced on my own mind ; and I look with much interest to a repetition of Maggi’s experiment. 
