Brown and O'Callaghan — 'I'he llujidity of Nickel Wire. 99 



which was fixed on the top of a rigid stand. After a great many trials and 

 preliminary experiments this arrangement was found to be practically fric- 

 tionless and torsionless. The small clutch also carried an aluminium arm, 

 consisting of a double sector of a circle of 4'5 cms. radius, which when 

 10 grammes weight were suspended from its circumference gave the initial 

 torque to the wire under test. 



The rigidity of the wire for a given torque varies inversely as the twist, 

 and the changes in this twist due to the magnetic fields are what are 

 recorded in the following tables and curves. 



In order to observe and measure the change in the twist of the wire, a 

 strong beam of light from a Nernst lamp was focussed on to the slit of a 

 collimator, and after reflection from the mirror attached to the wire the 

 image of the slit formed by the collimator lens was received in the eye-piece 

 of a microscope reading to O'OOl cni. The distance of the collimator slit 

 from the mirror was 56 cms., and the width of the slit was adjusted until 

 diffraction bands were seen in the microscope ; then a particular band was 

 chosen as the line of reference in measuring the change in the twist. The 

 distance from the mirror to the cross-hair in the microscope was 225 cms. so 



that a chancre in the twist of —— x 10"" radian, or 2'22 micro-radians, could 



4o 



be measured. The loads which were on the wire when being tested ranged 

 from 10° to 4 X 10^ grammes per square cm. ; and the final results here 

 recorded are in each case the mean of at least three different sets of measure- 

 ments. The temperature of the room during the experiments was fairly 

 constant, and did not vary more than half a degree above or below 15° C. 



Section 1. 

 Longitudinal Magnetic Fields. 



The solenoid used for this part of the work was 54 cms. long and 2'6 cms. 

 internal diameter, and consisted of 2058 turns of double cotton-covered 

 No. 18 copper wire in six layers, having a total resistance of 3'32 ohms. 

 The strength of the magnetic field per ampere at the middle of the coil was 

 47"89 c.g.s. units, which for this work was taken as 48 units. There was no 

 perceptible heating of the nickel wire under test when direct longitudinal 

 magnetic fields were round it, but there was a slight heating when alternating 

 magnetic fields were used. To eliminate, as far as possible, errors due to 

 this heating, the observations were taken rapidly, and repeated many times, 

 care being taken that the temperature of the room was kept constant, and 

 an interval of time allowed to elapse between each set of readings. 



The results for direct magnetic fields and for alternating fields of 



m2 



