448 Prof. F. Kohlrausch on the Increase and Decrease 



It has always appeared to me most probable that this rela- 

 tion must exist, but not in the way assumed by Lamont, but 

 simply as an equality. Lamont's experiments do not seem to 

 be sufficient to justify the view originally taken by him. 

 Later, Lamont appears* to return to the former view without 

 expressly saying so. He refers only briefly to some experi- 

 ments " previously published " by him, but does not mention 

 their result again. On the contrary, he calculates as if both 

 coefficients were of equal magnitude for very small forces, 

 and speaks of this as the generally accepted view. 



This latter view is indeed, a priori, the most probable, for it 

 corresponds to the law of permanency. 



The analogy, employed by Lamont, that a distorted elastic 

 body is more easily caused to return to its natural form than 

 to recede further from it, cannot be considered as of much 

 force. This fact of molecular mechanics is no doubt true for 

 persistent alteration of form. The limit of elasticity of a bent 

 body iSj we know, greater outwardly than inwardly ; and so, 

 of course, a magnet will in general behave towards forces 

 which are sufficiently great to cause a permanent alteration in 

 its magnetism. But we are not now concerned with that. The 

 change in magnetic moment depending upon change in posi- 

 tion relative to the earth is to be compared to elastic changes 

 in form, and for these there is no doubt that the modulus of 

 elasticity is of equal magnitude in both directions. 



I have abundantly convinced myself of the correctness of 

 both of these assertions. A magnet which had been exposed 

 some 8000 times to a weakening force of the magnitude of 

 the earth's horizontal magnetism (0*2 centimetre-gramme- 

 unit) showed still the same magnetism as before to within ^oVo- 

 And, as far as the modulus of elasticity is concerned, I found 

 this the same in both directions in a soft-steel rod which 

 had been just before much bent. I have endeavoured, with 

 the important assistance derived from the observations of 

 M. Kreichgauer, to decide this question in the following 

 manner. 



First I employed Weber's method of currents induced in a 

 rotated spiral by the varying magnetism, employing only rota- 

 tions of 90°; viz. first, from the east and west position of the 

 magnet into the normal position (north pole towards the north), 

 and, secondly, into the reversed position (north pole towards 

 the south). The deflections of the galvanometer-needle in the 

 two cases were found to be perceptibly equal. 



A difference of less than 1 per cent, in favour of the coeffi- 

 cient of increase is easily accounted for by a small error in the 

 adjustment of the spiral. 



* Handb. d. Magnetism, p. 871 (1867). 



