1913-14.] Resistance of Iron in Crossed Magnetic Fields. 265 
gives certain results for the change of resistance of mercury in a transverse 
magnetic field. Treating his numbers in the same way, I find that A 
has the value 6 x 10~ 13 or 5 x 10 -13 for mercury filaments of diameter 
0-7 mm. or 0'5 mm. respectively. The diameter of the mercury filament 
used in the experiment just described was almost exactly 1 mm. The 
discrepancies are considerable ; and j Jo is difficult to believe that the effect 
in mercury should depend on the diameter of the filament within the 
limits indicated.] 
Now in field 3750 the corresponding changes of resistance per 10,000 in 
iron and steel are respectively — 6 ‘9 and —9*2, that is, twenty or thirty 
times the numerical value for mercury. 
In the earlier experiments with nickel the highest transverse field 
reached was only 815 ; but it was obvious that in much higher fields the 
change of resistance would not exceed the value — 95 x 10 -4 , that is, about 
ten times the value for steel. Changes numerically equal to those given 
above for iron and steel were obtained for nickel in fields of only twenty 
and thirty Gauss respectively. 
It is well to bear in mind that, as proved in the earlier paper, the 
numerical value of the change due to a given transverse field is a function 
of the width of the strip of the magnetic metal, for the simple reason that 
on that width also depends the value of the magnetization. 
I now pass on to the consideration of the main object of the research, 
namely the influence of a steadily maintained magnetic field upon the 
changes of resistance due to a cyclically applied field at right angles to the 
former. 
With regard to the numbers given in the Table three pages back, it 
should be noted that the last figure in the measured changes of resistance 
is of no value, being well within the limits of experimental error. 
The smaller number of data for the iron strip was due to the overheat- 
ing of the magnetizing coil round the strip and the consequent breaking 
down of the insulation between the contiguous turns of the coil. But the 
nature of the results is obviously the same in both metals, and may be 
expressed qualitatively in the following words : — 
1. Under the influence of longitudinal magnetization the electric resist- 
ance of iron and steel is increased ; but this increase is notably diminished 
when the longitudinal magnetizing force is superposed cyclically upon a 
steadily sustained transverse magnetization. In the highest transverse 
fields used the change of resistance due to the superposed longitudinal field 
was in most cases very small, being a small fraction of the value when the 
longitudinal field acted alone. 
