PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 731 
instance, notwithstanding the imperfection of the arrangements. In this hope I was 
not disappointed. The glass needle (carried by the little suspended magnet, whieh 
was only about ^ an inch long), while moving steadily across its field, would receive 
an impulse forward and make two or three very rapidly diminishing oscillations, when 
the current was started through the magnetizing coil : when the current was sud- 
denly reversed, the needle would show little or no indication of any effect: when the 
current was broken, it would make a start backwards, and after two or three oscilla- 
tions would continue advancing as before, perhaps rather more rapidly. Traces of 
induced currents in the iron coil under the influence of the magnetizing helix were 
exhibited by scarcely perceptible differences in the bearing of the needle, according 
as the current was made in one direction or the other, and by slight impulses it 
received wlien the magnetizing current was suddenly reversed. After the current 
had been kept up for some hours through the iron wires, and when, partly by the 
heat developed by the magnetizing current during the periods of its flow, and partly 
by heat conducted from the iron wire within, the outside of the magnetizing coil had 
become very sensibly hot to the touch, the variation of the needle in the galvano- 
meter became much less rapid than at first ; and tolerably satisfactory indications, 
amounting to a fraction of a degree of permanent deflection, showed with perfect 
consistence an increase of resistance in the iron wire under magnetic force when 
the magnetic current was sustained in either direction, and a diminution of resist- 
ance in the same iron wire following immediately a cessation of the magnetizing 
current. 
147 . I followed the same method in a first attempt to find the effect of transverse 
magnetization on the electric conductivity of iron ; two spirals made on the plan 
described above (§ 136) being used as the resistance branches in the two channels 
conveying the divided current, and one of them placed between convex poles of a 
Ruhmkorff electro-magnet. The induced currents in making, reversing, and break- 
ing the magnetizing current were of course most conspicuously indicated by the 
galvanometer needle, but the needle came to rest after a few oscillations ; and then 
it did not exhibit any deviations of a sufiiciently marked character, when the direct 
effect of the electro-magnet (which by a very troublesome process of shifting the 
position of the magnet, was reduced as much as possible in preliminary arrange- 
ments,) was eliminated by reversals, to allow me to draw any decided conclusion as 
to the effect of the magnetic force on the conductivity of the iron spiral across which 
it acted. 
148. Before carrying into execution various obvious improvements in the experi- 
mental arrangements just described, or applying the system with the differential 
galvanometer to other investigations, I began to think of Maggi’s experiment on 
the relative thermal conductivities of a magnetized iron disc in directions across 
and along the lines of magnetization. As the electrical analogue, the method which 
* De la Rive, ‘ Electricity,’ vol. i. part 3. chap. iii. (p. 316, English edition, 1853). 
MDCCCLVI. 5 D 
