100 Barrett, Brown & Hadfikld— 0;« the Electrical Conductivity and 



PART II. 



Magnetic Properties. 



In order to examine the magnetic properties of these alloys it was necessary — 

 if complete B and H curves were to be obtained in each case — to adopt a method 

 that would not consume too much time, and also one that would not involve the 

 making of another set of specimens, which in some cases would have been difficult. 



The specimens were accordingly made, as already stated, in the form of rods 

 as long and slender as could be conveniently obtained. In this form both the 

 electric and magnetic resistance could be readily determined. It would have 

 been better to have had the rods of smaller diameter, but this was found to be 

 impracticable in many of the alloys. As, however, tlie length of all the rods was 

 about 200 times their diameter, the demagnetising reaction due to the poles in 

 c 



Fig. 7. 



most cases was small, inasmuch as the permeabilit3' of the majority of the specimens 

 was considerably less than that of iron, varying in fact down to a practically non- 

 magnetic alloy.* It was this great range in the permeability of the specimens, 

 as well as in their number, whicli led us to adopt the following magnetometric 

 method of measurement. 



A diagrammatic view of the general arrangement of the apparatus is shown in 

 fig. 7. The magnetometer M was supported on a solid stone pillar ; after various 

 trials the most convenient magnetometer was found to be that used on Lord 



* Experiments were made to compare the B and H curves obtained by our method of magnetometric 

 measurement, with the results obtained by a ring formed of the same material, using the ballistic method of 

 measurement. A certain correction is necessary for our rods in specimens of high permeability (see foot- 

 note, p. 106, and Note A, p. 125). 



