ELECTRICAL WAVES AND SOME OE ITS APPLICATIONS, 
19 
If we assume the value of B to be about 14,000, we see that the permeability for 
the extreme surface layer would be about 4, The value of the magnetizing force 
diminishes from the surface inwards, so that the mean permeability of the iron to the 
discharge should be greater than the value at the surface. These considerations 
show that the permeability of iron to these discharges is by no means constant, but 
depends on the diameter of the wire and the intensity of the discharge. 
The resistance of iron wires was found to vary with the length of the spark. 
Short air-breaks gave higher values of the resistance than long ones. When the length 
of the spark was so adjusted that the maximum current was af)proximately constant 
for different periods, the resistance was found to vary as the square root of the 
frequency, as we should expect from theory. 
Several specimens of pianoforte steel wire were examined to see whether the larger 
waste of energy due to hysteresis in steel materially affected the value of the 
resistance, but the increase of resistance was not so great as for soft iron wires of the 
same diameter, although the loss due to hysteresis in steel is much greater than in 
soft iron for slow cycles. 
Absorption of Energy by Metal Cylinders. 
This subject has been treated mathematically and experimentally by J. J. Thomson 
(‘ .Recent Researches,’ p. 321-326). It is there shown by observing the electrodeless 
discharge that a cylinder of iron placed in a solenoid absorbs considerably more energy 
than a copper one of the same dimensions. The method adopted here admitted of 
quantitative as well as qualitative results. 
An ordinary Leyden jar was discharged through a solenoid of about thirty turns 
and 14 centims. long. The metal cylinder was then placed in the solenoid and the 
damping of the oscillations observed. The cylinder was then removed and a non- 
inductive resistance added until the damping was the same as when the metal cylinder 
was in the solenoid. The absorption of energy in the cylinder was then equal to the 
absorption of energy in the added resistance, whose value was known. In the above 
we have taken no account of the change of inductance of the circuit due to the metal 
cylinder being placed in the solenoid. The change is small, and could be made 
negligible by making the inductance of the solenoid small compared with the rest of 
the circuit. 
(1.) A test-tube was filled with finely laminated iron wire, *008 inch in diameter. 
The test-tube was filled with paraffin oil, to insure insulation from edd}^ 
currents. 
The absorption of energy in this case corresponded to an added resistance 
of 10*25 ohms to the circuit. 
(2.) A test-tube filled with steel filings and insulated as in (1). Increase of resist¬ 
ance, 9 ohms, 
D 2 
