604 Mr. Williams on the Influence of Stress and 



tlie apparatus at my disposal enabled fields of 800 units to be 

 obtained, it is thought that an account of these experiments 

 will not be without interest. 



Appa7'atus. 



The magnetizing field was obtained by means of a solenoid 

 1 metre long, having a resistance of about 3 ohms and giving 

 afield of 40 c.g.s. units per ampere. The highest field used 

 was 800 units. 



The coil was provided with a water-jacket through which 

 flowed a current of cold water from the mains. Inside this 

 jacket a glass tube was placed, and the space between the two 

 was filled with cotton. The coil was fixed vertically in a 

 wooden framework. 



The ends of each of the wires experimented upon were 

 soldered to two thick copper rods, one of which was fixed at 

 the top of the glass tube so that the wire hung vertically 

 inside it. The wires were shorter than the coil by about 

 20 cms., and being placed centrally in the coil were in an 

 approximately uniform field. All the wires on which expe- 

 riments were made were less than J mm. in diameter, so 

 that the demagnetizing factor may in all cases be neglected. 



The wires were stretched by placing weights in a pan 

 tittached to the low^er copper rod. The change of resistance 

 was measured in the same way as described in the previous 

 paper. The curves for the temporary and for the residual 

 effect were found separately, and the ordinates added together 

 to construct a curve for the total effect. 



Experiment I. — The Influence of Stress on the Magnetic 

 Change of Resistance in Nickel. 



The nickel wire used was '35 mm. in diameter, and was 

 annealed by heating in the flame of a Bunsen burner. 



The wire was suspended as described above and change-of- 

 resistance curves were taken for a number of different loads 

 np to the breaking-load. The results obtained are shown by 

 the curves in PL XXVIL fig. 1. It will be seen that the 

 efi'ect of loading is to diminish the change of resistance at low 

 fields and to increase it at high fields. The residual change 

 of resistance is diminished by loading at all the fields tried. 



It is well known that the magnetic properties of nickel 

 and iron are altered by stretching even after the load has 

 been removed. Some measurements were made to investigate 

 this effect for the magnetic change of resistance, and the 

 results oblained are shown in fig. 2. The cnrve A in this 

 figure is the original curve taken before loading the wire, the 

 curve B w^as taken after the load of 10'9, kilos had been 



