292 Dr. Tvndall on the Laws of Mtignetism, 



from the latter law occurs. From this to absolute OQiataot, tJbjO 

 passage from oue law to the other is gradual. •., -.i;,ff ■>•.{) hun 

 36. To demonstrate this gradual change, it Mas necessaiy to 

 apply a lilm much thinner than the paper formerly used, and to 

 pi-ocm'e this was a matter of some difficulty. Gold leaf was oja 

 many accounts unsuitable. The following expedient occurred to 

 me on observing the extreme tenuity of a gun-cotton balloon. 

 It is well known that a solution of gun-cotton in sulphuric ?ether, 

 when spread thin and suffered to di-y, forms a tough unyielding 

 film. A layer of such a solution was laid, like a wash of water 

 colom', with a camel's-hair pencil upon the smooth Hat pole. 

 The layer must ha\"e been exceedingly thin, for when dry it ex- 

 hibited the colom's of thin plates with great brilliancy. The 

 ball was brought down upon this, weights from 300 to 900 

 grammes were laid upon the scale-pan, and the corresponding 

 magnetic powers determined. This done, another wash of the 

 solution was laid on, and the same process repeated. The result 

 may be thus stated : — Through the interval between the surface 

 of the pole and the sixth layer no law is recognisable. At the 

 lower limit of this space the attraction of the ball is directlij propor- 

 tional to the mafjnetizing stream ; at the upper limit, and beyond it, 

 ths attraction is proportional to the sguare of the maf/netizinr/ stream. 

 wf. i37. I will for the present limit myself to an observation or 

 two on this smgidar result, and defer the fuller discussion of the 

 subject to a future paper. In the case of contact the fundamental 

 law of magnetism is, to all appearance, contradicted — that law 

 which affirms that the attraction is expressed by the product of the 

 magnetism of the magnet into that of the sphere. But this law 

 supposes that opposite magnetic flinds act upon each other. In 

 speaking of the attraction of a mass of soft iron by a magnet, it 

 is usual to say that a north ])ole excites a south pole, and a south 

 pole a north pole, attraction being the consequence. A cylinder 

 of soft iron 1 inch thick and 6 inches long was laid upon the 

 end of the excited magnet. According to the notion generally 

 entertained, two O])posite poles embraced each other at the place 

 of contact. Tliis being the case, we might infer, that on raising 

 a magnetic needle from the centre of the magnet u})wards, that 

 end of the needle which was attracted below the place of contact 

 ought, upon passing the latter, to be immediately repelled ; but 

 there is no such action exhibited. From the centre of the mag- 

 net to the top of the soft iron cylinder the same end of the needle 

 was attracted, the combination of magnet and cylinder behaving 

 in all respects as if they were one continuous mass*. To obtain 



* Since making the above experiment, I have learned that the same has 

 heen observed by Prof. PoggendorfF and M. Van Rees. See Poggeudorif 's 

 Annalen. vol. Ixxiv. pp. 213-230. 



I 



