740 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
circuit broken during the starting, the reversal, or the stoppage of the current through 
the coil of the electro-magnet. 
158. The deflection due to the effect of magnetic force on the substance of the iron 
was corrected in each case by sliding the moveable electrode towards the part of the 
testing scale remote from the end connected with the iron spiral which experienced 
that effect ; and it therefore indicated a diminution of conductivity in the iron. 
159. If the lines of magnetization had been exactly perpendicular to the lines of 
electric current through the iron, we should now conclude that transverse magneti- 
zation diminishes the conductivity of an iron conductor ; that is, that it produces the 
same kind of effect on the conductivity as longitudinal magnetization. But the lines 
of current formed spirals inclined at an angle of 84° to the lines of the magnetizing 
force ; and the mutual influence of the consecutive parts of the magnetized iron spiral 
would have an effect (not wholly compensated by the mutual influences between 
the successive spires because of the thickness of the twine between them,) contri- 
buting to longitudinal magnetization; and therefore the lines of magnetization must 
have been inclined, not at 90°, but at some angle less than 84°, to the direction of the 
lines of current. Hence all we can conclude is, that not only longitudinal magneti- 
zation but oblique magnetization up to some angle of obliquity less than 84° from the 
lines of current, diminishes the electric conductivity of iron. 
160. It remains to be determined by experiment what is the effect of magneti- 
zation right across the lines of current: if a diminution of conductivity, whether a 
greater or a less diminution than is caused by an equal longitudinal magnetization? 
or if it is an increase of conductivity, what is the angle of obliquity of the magneti- 
zation which gives neither increase nor diminution of conductivity ? 
161. Exp. 4. To discover the differential effect of magnetization on the conductivity 
of iron in different directions. — A square of 1^ inch each side was cut from thin 
sheet metal, and powerful electrodes were soldered to two corners, A and B. A 
« 
reference electrode (§ 149) of No. 18 copper wire was soldered to C, one of the 
other corners, and the two extremities of a yard of the same kind of wire, to be 
used as a multiplying branch, were soldered to points D, E, about irth of an inch 
from one another on each side of the remaining corner. A current being con- 
ducted through the square by the principal electrodes A and B, the reference 
electrode was used to connect C permanently with the commutator belonging to 
the testing galvanometer. Another wire used as a testing electrode, was applied to 
connect any point of the plate, or of the multiplying branch, with the other galva- 
nometer electrode. In the first place, it was found that a powerful current was 
raised in the galvanometer coil if the testing electrode was applied to any point of the 
multiplying branch ; and it was necessary therefore, as was anticipated, to adjust the 
distribution of resistance through the square by filing, so that there might be some 
point on the testing branch which would give no current when touched by the testing 
electrode. (See below, § 176, where a less troublesome way of managing this part 
of the arrangement, in an analogous experiment, is described.) For this purpose, in 
