736 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
to be drawn. In the second place, the galvanometer circuit may be broken and com- 
pleted, and reversed, as often as is desired, by its own commutator, without affecting 
to the slightest sensible degree, the strength of the current through the tested and 
testing branches ; while in the former mode of experimenting the indicating needle 
was always under the action of the divided current, unless the current in one or the 
other of the branches was broken, which introduced irregularities lasting for a con- 
siderable time, by the consequent changes of temperature through the conductors. 
This was an immense convenience in every experiment, and allowed small deflections, 
amounting to the tenth of a degree, to be tested with ease by using the commutator 
of the galvanometer, and getting oscillations. But it was of especial advantage in 
the experiments on the effects of transverse magnetization, since the galvanometer 
circuit had only to be kept broken for a few seconds during the making, breaking, 
or reversing of the magnetizing current, to get entirely rid of all disturbances of the 
needle due to induced currents ; and in all experiments in which the Ruhmkorff 
magnet was used, since by breaking the galvanometer circuit and using a little steel 
magnet in the hand, the galvanometer needle could be let down in a few seconds 
into its position as affected by the direct action of the large magnet, before proceed- 
ing to test the current due to the change of resistance under investigation. In the 
third place, it is possessed of almost unlimited capacity for increase of sensibility. In 
some of the experiments on the influence of tension on electric conductivity, I have 
tested with the greatest ease effects amounting to only of the whole resistance 
of the wire under examination, and I see no difficulty in testing effects amounting 
to only the tenth part of that, or even hundreds of times smaller effects, by using 
more powerful currents, and applying artificial means to keep the wires cool. 
PART V. ON THE EFFECTS OF MAGNETIZATION ON THE ELECTRIC CONDUCTIVITY OF 
METALS. 
154. The remarkable effects which I found produced in the thermo-electric quality 
of a metal by magnetization and by mechanical strain, appeared to render it highly 
probable that the same agencies would also influence their electric conductivities. 
To demonstrate this if I could, and to discover the nature of the anticipated effects, 
I commenced an experimental investigation of the subject, and, after various nuga- 
tory operations, arrived at a variety of positive results by the following processes, 
155. Exp. 1. On the longitudinal electric conductivity of longitudinally magnetized 
iron wire. — A length of seventy-two yards of silk-covered copper wire was rolled in 
six strands, or altogether in about 860 turns on a core made up of two concentric brass 
tubes, connected at their ends by a ring of sheet brass, and arranged to have water sent 
through the space between them by suitable entrance and exit pipes soldered to aper- 
tures in the outer one ; the external diameter of the brass tube was about inch, and 
the internal diameter of the inner one about ^ inch ; the metal of both outer and inner 
