ff> r nVT 



27*6 Dr. Tj'udall on the Laws of Magnetisrki 



i;, 1 ,;■).- m:: i . :._ li' rtaiijS lI-J 'lo o'liv/ ;jii.t 



.l^Jj^'teramberofleave^if^'lJ^^ jM'"^ ~^' f^ 3^* S^i"*'"* 



.!:fl')i%Rd 'ill I ■iT'^^ "^Mit ,ii nuoirlj^ltt (nu/l iii'rj^af) !.)iIT 

 [vih%n-n mHl-)i?^i:!?f^!' fiuit'M;!'il^r?-rj;fft .f-ioJ*; yiov 0(> 



14 . . 26 Si : .13-.1,, If,; 115 ! '• 



15 27 10 13-2,., 3,{J,ir 



16 27 50, ,., ,,„-a3l2'io.'nof(.i. r.MMM,,^ 



17 28 35 ^"" 



18 29 00 '■ 



19 29 50 



20 30 15 



21 31 



22 -31 40 



23 32 25 



24 33 10 



We here observe tliat the condition '""^^ = const, is satisiied 



vd 



to a remarkable degree of exactitude. 



16. Considering the electro-magnet in its separate states^ 

 during the above experiments, to represent a number of distinct 

 magnets of different strengths, we see that if one magnet have 

 twice the strength of another, and if the latter exercise a certain 

 force at a certain distance, the former will exercise an equal force 

 at four times this distance ; if one magnet have three times the 

 strength of the other, the former will exercise an equal force at 

 nine times the distance ; and so on. 



§4. 



17. An apparently well-grounded doubt, however, may attach 

 itself to all these experiments. In damping the cm-rent, we 

 descended from a point at which the full power of the latter was 

 exerted to that at which the ball gave way. Now it is well 

 known that a soft iron core, when once excited, does not instantly 

 lose its magnetism on the cessation of the current, but continues 

 active for a short time afterwards. The question, " may not the 

 magnetism lag in the core dming the descent above alluded to," 

 natm*ally suggests itself here. But if the decrease of the stream 

 be not immediately accompanied by a proportionate fall of the 

 magnetic power, the strength of the magnet at the moment the 

 ball separated is not the true strength due to the current then 

 present, but a strength due to a greater current. It Avill be seen 

 immediately that experiment seems to justify this doubt. " ' 



18. To avoid everything of this kind, it was only neccsslal'y ti' 

 ascertain pretty nearly beforehand at what particvdar point in 



