Dr. Tyjidall uii the .Laws of Magnetism. 385 



,, ]\J[ult;iplyiug the weights in each column by^theu' correspond- 

 ing distai;ices, we obtain tlie following result :— ^ , ^, , ' , , 

 " ^ ■ ' - ° ' i; ri'j'io ^Bfl •)ln({ oil> 



'"DMtolde'. ' -Subrn. Xb. 1. Subin. No. 2. Siibm.iNWS. Sabm. No. 4. 



-Ml M,i, ■i-^.';;o.27- •■-•• ■■ 0-31 -K ■■■■ 0'35i<i/! ■.|.:,0-4 

 (Iffffug/ <-.:t ■-!■), (^^,gjr, lliv/ ii,',ip.'4(5..r. T,;^i|>, /Q-^e 'I'l! ''h/I Jr^,4^yi 



■fomaAf MtrnffrM^Wfi J.'nnbo-KpWtrf ;>/;,-[ ail-i^^. <fii([ 1/ ■"■*^ 

 •/fil -rSKl.; -Ill J 0-8|!>if- tfirJ/i'-OrBg-.-fiv i.i ifv.O-45< 'to >,■)>■- ^ 0-33 



:0gT0j|f f{l(7/ 'tl0^^)'iirfrurt.vi(lO>g4if(>lti;.(nii-ijyQ.4:^oIo liu ,bl0^g0^.1 



6 0-66 '"*'" '"m"^ ■'"•' ' b-^2'"^' ' 0-30 



, '"T . 0-6G •^""! ^©-'56' '•' ■'3'-'^'iOi40''- '■= .:"i.-0'3L' 



" ^^w''' 0-64 ^ ■"''"■ 'o^stf'""' ■■-■'• ^dW"^ ^'' "''■^' 



9 0-63 0-50 OricTnoT,,;. 



When we consider the disturbing action of the edges, and the 

 difficulty of preserving perfect parallelism between magnet and 

 submagnet, we shall not be surprised at the deviations which ex- 

 hibit themselves close to the poles. At a little distance the law 

 of action expressed above receives a good corroboration. 



The experimental portion of this investigation had been alreadj- 

 completed, when I learned that a series of experiments was made 

 in 18tl by M. Cramei", mechanician to the university in Kiel, 

 with a view to determine the influence of distance upon the lift- 

 ing power of the magnet. He experimented with steel magnets 

 of the horse-shoe form, placing, in each case, the unlike poles of 

 two magnets of the same size exactly opposite to each other. 

 The distances between the poles were measured, as in our case, 

 by leaves of paper. The result of M. Cramer's investigation, as 

 contained in vol. lii. p. 302 of Poggendorff s Annalen, is as 

 follows : — I , 



" Within the limits of these experiments the magnetic attrac- 

 tion by no means decreases in the inverse ratio of , the square of 

 tlic distance [selir naturlicn, "•)., , f^ , 



"When the various columns are ' compared^ '^liere is so little 

 agreement exhilnted, and even the members of the same series 

 decrease so iri'cgularly, that it apj)cars impossible to refer them 

 to any definite law. This la\vles.sness is to me as unexpected as 

 it is enigmatical. I cannot attribute it to the inaccuracy of the 

 method ])ur.sued, which in itself is simple, and in the carrying 

 out of wliich 1 expended all possible care and ))ains." 

 ' The irregularity of decrease mentioned by the author apjjcars 

 to arise from his manner of viewing the sixbject. lie begins his 

 I '■■■ > I,.. ,/ 



