on the Thcorij of Magnetic Reaction. 4-7 i 



ceases to allow of any fuither induction from this source. 

 Now apply a very hard steel lifter to both poles, keeping up 

 the same galvanic current, and the magnet will be found to 

 have gained a considerable increase of )iermanent magnetism. 

 After ihe magnet thus operated upon ceases to indicate any 

 further increase of its magnetism, apply a lifter composed of 

 very soft iron of precisely the same dimensions as the former 

 one, the galvanic action being continued, and the magnet will 

 be found to have acquired a still higher charge of permanent 

 magnetism. Now as the induction from the galvanic current 

 was the same in each stage of this experiment, and the effect 

 produced varied according as the condition of the keeper was 

 more or less favourable for induction, the increase of )ierma- 

 nent magnetism would appear to depend upon the reaction of 

 the keeper, which, by the converse of Dr. Ritchie's reasoning, 

 must then possess a higher state of magnetism than the magnet 

 to which it is applied. 



The experiment proposed by Dr. Ritchie would certainly 

 set this matter at rest provided it answered; but in case it does 

 not, it will not in the least invalidate these facts concerning 

 magnetic reaction : less conclusive evidence gained from ex- 

 periment and reasoning is not to be totally abandoned because 

 perfectly decisive experiments are unattainable. I believe the 

 experiment proposed by Dr. Ritchie will not succeed, in con- 

 sequence of the keeper, under the circumstances he has named 

 not possessing an inducing power sufficient to overcome the re- 

 sistance whicli steel furnishes to magnetic induction. It is well 

 known that magnetism having once been induced in a piece of 

 steel and partially destroyed by heat or the action of similar poles 

 of another magnet, maybe in some measure restored by there- 

 action of the keeper. This is probably owing to the particles 

 of iron having acquired a tendency to resume their former 

 state when the exciting cause is applied. This would be an 

 instance of induction just as much as if the steel had never 

 been magnetized, for the addition of magnetism is just as 

 necessary to restore that which has been lost as to impart it 

 in the first instance. Dr. Ritchie says that if the explanation 

 which I have given be admitted, it will completely overthrow 

 the Newtonian law of the perfect equality of action and reac- 

 tion. I had always considered this law as applicable only to 

 mechanical forces, and not extending in the least to these 

 physical phaL^nomeua, the acting cause of which is altogether 

 unknown. Suppose a number of })ieces of steel properly tem- 

 pered, and for convenience made into the form of theconnnon 

 horse-shoe magnet, anil one of these magnetized to saturation; 

 now by this one let all the others be magnetized, ard after- 



