904 Mr. C. W. Heaps on the Effect of 



Fig. 6. Nickel. 



IV. Transverse field, electromagnet. Same specimen as II. 

 Fig. 7. Cadmium. 



I. Longitudinal field, solenoid. Temp. 17°. 

 Wire, diara. 0*034 cm. Resistance 1*72 ohms. 



II. Transverse field, solenoid. Temp. 17°. 

 Strips. Resistance 1'31 ohms. 

 Fig. 8. Cadmium. 



III. Longitudinal field, electromagnet. Temp. 23°. 

 Wire, diam. 0*014 cm. Resistance L84 ohms. 

 IV. Transverse field, electromagnet. Same specimen as III. 

 Fig. 9. Hensler alloy. 



I. Longitudinal field, solenoid. Temp. 21 c, 7. Resistance 

 0-68 ohm. 

 Fig. 10. Heusler alloy. 



II. Transverse field, solenoid. Temp. 21 0, 2. Same spe- 



cimen as L, differently mounted. 

 Fig. 11. Heusler alloy. 



III. Longitudinal field, electromagnet. Temp. 23°. 



Resistance 0*68 ohm. 



IV. Transverse field, electromagnet. Temp. 22°-4. Same 



specimen as III. 



In addition to the results as plotted, the transverse effect 

 for the Heusler alloy was measured for different fields up to 

 30,000, and curve IV, was found to remain a straight line 

 up to that point. Also, as the alloy was mounted in the 

 form of rather large flat bars, it seemed possible that the 

 magnetic field might be affected by the form of these bars. 

 To test this a specimen was made up from iron in exactly 

 the same manner and a set of measurements made, but the 

 curve obtained did not differ appreciably from III. for iron. 



As curve II. for cadmium was obtained with the metal in 

 the form of flat strips and curve I. with wires, a new set of 

 observations of both transverse and longitudinal effects was 

 made for a given field in the solenoid, using. the metal in the 

 form of wire in each case. The means of a large number of 

 trials gave <ffi./R = 4"65 x 10~ 6 for a longitudinal field of 1400, 

 and for the transverse field of the same strength dR/R was 

 found to be 5*42 xl0~ 6 . This result is more trustworthy 

 than the results shown in curves III. and IV. in larger fields, 

 because in the solenoid the field is uniform over all parts of 

 the specimen. 



It is very evident by comparing curves I. and III. for iron 

 and I. and II. for nickel that a certain error is introduced 

 wherever the electromagnet is used, because of the non-uniform 

 field. The pole-pieces were so far apart that the bending of 

 the lines of force introduced a small transverse component ; 



