CHANGE OF DIME^•SIOi\S BY MAGXETIZATIOX. 79 



the front view of these rings. R and R' were two rods in con- 

 tact with the ends of the tube. The ends of these rods were ben.t 

 upwards and so filed down, that they could easily slide between 

 two parallel wires of the coil, which were specially fixed at a dis- 

 tance of 1 mm. from each other. The rod ?• served to communicate 

 the motion to the prism P. The other parts of the apparatus 

 remained unchanged. The apparatus was put into a magnetizing 

 coil, 30 cm. long and wound in 12 layers with copper wire of 

 2 mm. diameter. The field at the centre of the coil due to a 

 current of one ampere was 37.97 C.G.S. units. The current 

 through the outer coil produced the change of length by longi- 

 tudinal magnetization and that through the inner coil gave rise 

 to the change of length by circular magnetization. 



To study the efi'ect of temperature on the change of length, 

 the circular magnetizing coil was wound, not by a single wire, 

 but by double Avires ; thus connecting the four ends of these 



wires to a reversing key as shown in 

 the figure, the circular field can be 

 made or annulled by turning the key 

 one way or the other. The total 

 number of turns of the circular mag- 

 netizins; coil was 44 for the nickel 

 tube, 40 for the wolfram steel tube and 36 for the soft iron tube. 

 The magnetizing currents were measured by Thomson graded 

 galvanometers which were compared with a decianipere balance 

 before each experiment. 



3. The samples used in the present experiment had the 

 following dimensions : 



