196 Mr. Rainey's Analysis ofDr, Ritchie's Reply. 



of its real power. A practical inference may be drawn from 

 these experiments, namehv that little advantage is gained by 

 the association of magnets, unless their hardness be such as 

 to prevent the dissipation of so large a portion of their mag- 

 netism in consequence of the contact of their similar poles. 

 Dr. Ritchie says that I have taken for granted that the action 

 of a galvanic current upon a steel magnet is the same, whether 

 its poles are connected by an armature or not. Certainly so far 

 as the identity refers to the nature of the eflect produced, 

 there can be no occasion to entertain any doubt. The usual 

 results of a galvanic current, such as polarity, &c., are precisely 

 the same in each case, varying only in degree. It is the ex- 

 planation of the difference in the quantity of effect of the gal- 

 vanic current upon the steel when the poles are connected, 

 and when separated, which forms the subject of discussion in 

 n)y paper, and which has been attributed to the reaction of 

 the keeper; therefore, so far from having adopted this as an 

 axiom, I have endeavoured to deduce it as a consequence. 



The last experiment which Dr. Ritchie adduces, p. 59, which 

 is to set the matter at rest, is a very singular one. " Take a bar 

 of steel, and bend another of the same size and length into 

 a square; magnetize the straight bar by drawing it lengthwise 

 over one of the poles of a magnet ; move the same pole the 

 same number of times round the square; break the bar into 

 four equal parts, and the square at the corners, and the bar 

 C D will be stronger than either [any is meant) portion of the 

 straight bar." It will be obvious upon a little reflection, that 

 in the process of magnetizing a closed circuit, of any form 

 whatever, one part of the steel composing the circuit will act 

 as a keeper to the other, and consequently that the bar thus 

 magnetized is under the same circumstances as a horseshoe 

 magnet would be when magnetized with a piece of iron in 

 contact with its poles. Hence we see that Dr. Ritchie's square 

 was under circumstances much more favourable for securing 

 its full complement of magnetism than the bar was; therefore it 

 is not to be deemed a matter of wonder that it was found to be 

 so much stronger. The most curious part of this experiment 

 is the laborious task of cutting each into four exactly equal 

 parts; why, if Dr. Ritchie had found that the bar which had 

 been bent was stronger than the straight one, surely he 

 might have inferred that the parts of which the bent bar was 

 made up would be stronger than those corresponding to 

 them in the straight bar, and that the most troublesome part 

 of the experiment might have been dispensed with. It is also 

 equally strange why Dr. Ritchie should have chosen an exact 

 ^(juare as the form necessary for the closed circuit. Does not 



