STRAIN ON THE PHYSICAL PROPERTIES OF MATTER. 17 



Experiment XV. 



The bundle of soft-iron wires used in Experiment XIV. was subjected to gradually 

 increased amounts of magnetic stress according to the following plan : First the 

 magnetising circuit was closed and the deflection D! noted, and after the spot of light 

 had come to rest the circuit was opened and the deflection D 2 observed ; next the 

 battery was reversed, and the deflections D s , D 4 , produced by closing and opening the 

 circuit respectively, were read off as before. When the magnetic stress produces no 

 permanent set the four deflections are all of equal value ; but, if any such set exists, 

 D 3 D 4 will represent the change of magnetisation produced by the wrenching of the 

 molecules from their positions of permanent set on one side to their positions of perma- 

 nent set on the other. If the reversals be made a great many times, D 3 D + becomes 

 less and less, but never vanishes, and finally becomes a constant. Experimenting in the 

 manner mentioned above, the permanent set could be just detected when the magnetic 

 stress had reached the value of 0'01690, a value only a little less than half of that at 

 which it had been detected by the other method.* The method of reversals possesses 

 two great advantages over the previous method in attacking the question of a magnetic 

 elastic limit. In the first place the amount of permanent set detectable is about 

 doubled ; and, in the second place, we may go on repeating the reversals so that any 

 failure to detect a small amount of permanent set at the first reversal may be rectified 

 in the second or subsequent reversals. 



A very protracted examination of the values of D 3 D 4 for different values of 

 magnetic stress extending beyond H was now made,t and it was found that, provided 

 the reversals were for each stress continued long enough to make D 3 D t constant, the 

 permanent set as thus measured was exactly proportional to the square of the magnetic 

 stress. We can now plainly see how mutters stand with respect to a magnetic elastic 

 limit, namely, that no such limit is mathematically existent ; but from the rapid falling- 

 off of the permanent set, with decrease of magnetic stress, as indicated by the above- 

 mentioned law, and from the observations with hard and soft iron which have been 

 already made, we may, in all probability, safely assume that even with the very softest 

 iron* no permanent set amounting to one per cent, can be detected by the ballistic 

 method, provided the magnetic stress does not exceed |H.] 



* This might be expected from what has been said above. 



t It is not necessary to enter into the details of this examination, as they will be laid before tho Royal 

 Society in a subsequent paper. 



* The soft-iron wire used in these experiments was specially prepared for the author by Messrs. JOHNSON 

 and NKPHKW ; it can bo permanently elongated 25 per cent, before breaking. 



Before this note was written Lord RATLEIGH had by another method shown ('Phil. Mag.,' vol. -.'. 

 1887, pp. 225-241) that up to JH there is no sensible permanent set. Lord RAYI.KILII has also shown 

 in his paper that when permanent set is produced by higher magnetising stresses the amount of set is 

 proportional to the square of the stress. 



MDCXXILXXXVIII. A. I) 



