Erskinb. — On Magnetic Screening. 179 



If a metallic screen was placed inside the coil, so as to 

 -surround the needle, the demagnetization produced by the 

 discharge was less. 



The screening depends on the thickness and on the con- 

 ductivity of the screen, and on the frequency of the discharge. 



The condenser, which consisted of ordinary 40oz. leyden- 

 jars, was charged by a Voss influence machine, and connected 

 in series with several coils wound on glass tube of 21mm. 

 diameter. In the circuit was a spark-gap, of length 3-7mm., 

 and the diameter of the knobs was 2-8cm. ; hence the potential 

 at discharge was about 43-5 electrostatic units, or 13,000 volts. 

 (J. J. Thomson, " Eeceut Eesearches," p. 77.) It was found 

 that the effect of the discharge varied less with this length of 

 spark than with a shorter spark. 



The needles used were of glass-hard pianoforte-steel wire. 

 They were magnetized to saturation by placing them in a coil 

 (of 191 turns, and of length 9-9cm.) through which passed a 

 current of from six to ten amperes, produced by a Grove's 

 battery or by an accumulator. The needles, for convenience 

 in handling, were sealed in fine glass tubes. 



The needle, after being magnetized, was placed at a dis- 

 tance of 102-5cm. from the needle of a magnetometer. The 

 magnetometer readings were taken by the ordinary lamp-and- 

 ■scale method. The needle was then placed in one of the coils 

 in the leyden-jar circuit and a discharge was passed. The 

 needle was again tested by the magnetometer and the new 

 reading noted. Great care had to be taken in replacing the 

 needle, and in order to avoid error the magnetometer was kept 

 in a fixed position throughout the experiments, while the 

 needle was placed in a groove cut in a bar attached to a 

 stand, which was itself screwed down to the table on which 

 the instrument stood. 



The screens were metal cyhnders, which could be placed 

 inside the coils. They were formed by winding thin sheets of 

 metals on glass tubes. 



Great care had to be taken to make the contact at the 

 junction good ; if there is merely touching contact the screen- 

 ing is much reduced. 



The screening produced by different metals was com- 

 pared as follows: Tubes were wound with different num- 

 bers of layers of tinfoil. The reductions in the deflection 

 produced on the magnetized needle when surrounded by these 

 and placed in a certain coil (of 2-04 turns per centimetre) 

 were observed. Similar experiments were made with cyhn- 

 ders of other metals, and the numbers of layers of tinfoil pro- 

 ducing the same effect w^ere calculated by interpolation. 



The diameter of the tubes on wnich the metals were 

 wound was 14-4mm., and the length of metal 16cm. The 



