on Magnetic Testing. 5 



cannot be got rid of, or even affected, by any magnetic 

 operation or process of operations whatever, unless the sign 

 of the field is changed. Thus if on applying a field of intensity 

 -\ H the corresponding intensity of magnetization is I l9 

 cyclic variation of the field strength between the limits +H 

 and H — h, where h may have any positive value ranging 

 from to H, does not result in the value of I corresponding 

 to the field +H being diminished. If, however, h becomes 

 greater than H, in other words if the field is reversed, then 

 the value of I corresponding to +H is reduced. The effect 

 of applying even an extremely small negative field is to bring 

 about a perceptible diminution in the value of I, that is to 

 reduce, more or less, the " Sensitive State." 



A further series of experiments showed that the " Sensitive 

 .State " induced by equal augmentations or diminutions of the 

 temperature are of widely different amounts depending on 

 the position of the temperature ranges on the temperature 

 scale. Even a small increment of only 25° C. in the neigh- 

 bourhood of 180° C. produced, in the case of the steel tested, 

 a percentage ' : Sensitive State"'' of 10 for a field of 10 c.G.s. 

 uuits. 



Such facts have a very important bearing on magnetic 

 testing. For example, suppose that it is required to test the 

 magnetic properties of a specimen at 200° C. We shall 

 consider the specimen to be initially neutral and at room 

 temperature. It may now be heated to 200° C, and tested 

 at that temperature. Or again it might be heated to say 

 250° C, then cooled to 200° C, and tested in like manner 

 at that latter temperature. From what has been said it 

 will be evident that if the specimen has not been rendered 

 neutral at 200° 0. in each of the two cases previous to exami- 

 nation, the results of the two tests will not in general be in 

 agreement. 



Or suppose that it is desired to test the magnetic properties 

 of a specimen at temperatures lying between room tempe- 

 rature and the critical temperature of the material. The 

 specimen might be heated up above the critical temperature 

 and allowed to cool very slowly, tests being carried out 

 at different stages of the process. This method has been 

 adopted by many experimenters. If the specimen were com- 

 posed of steel, it would be heated up to the neighbourhood 

 of 900° C. and then cooled. A test carried out at 800° C, 

 without the specimen having been rendered neutral, would 

 give a certain magnetization curve. The specimen might 

 now be submitted to the process of reversals, allowed to cool 

 to 700° C, and then tested in like manner. This procedure. 



