424 Dr. Faraday on the Physical Character of 



wire, about 15 or 20 inches in length, and 005 of an inch in 

 diameter. It can hardly be effected by bars ; and when the wire 

 is afterwards examined by filings (32*34.), it is found to have 

 irregular, and consecutive poles, which vary as the magnetization 

 is repeated with the same wire, as if they broke out suddenly by 

 a rupture of something like unstable equilibrium ; the effects 

 apparently being chiefly referable to the cause now assigned. 

 Again, when a magnet is made out of a thin, hard, steel plate, 

 whose length is ten or twelve times its width, it is well known 

 how the hues of force issue from it in greatest abundance at the 

 extreme angles, and then at the edges ; and how a spot ou the 

 face gives exit to a much smaller number of lines than a like 

 spot on the edge, at the same distance from the magnetic equator. 

 Iron filings show such results readily, and so also do the vibra- 

 tions of a magnetic needle, and likewise the revolutions of a wire 

 ring (3212.). Now this state of the plate-magnet is precisely 

 that which would be expected from the hypotheses of the neces- 

 sary and dependent state of the magnet on the medium sur- 

 rounding it. 



3291. The mutual dependence of a magnet and the external 

 medium, assumed in the view now put forth, bears upon, and 

 may probably explain, numerous observations of the apparently 

 superficial character of the magnetism of iron and magnets in 

 different cases. If a hard steel bar be magnetized by touch of 

 other magnets, both the vicinity of the superficial parts of the 

 bar to the exciting magnet in the first instance, and afterwards 

 to the surrounding sustaining medium, will tend to cause the 

 magnetism to be superficial in the bar. If a small magnet or 

 a horseshoe bar be surrounded by a thick shell of iron as its ex- 

 ternal medium, the inner surface of the iron, or that nearest to the 

 magnet, with its neighbouring parts, will convey on more power 

 than the parts further away. If a thick iron core be placed in a 

 helix carrying a feeble or moderate electric current, it is the part 

 of the core nearest to the helix which becomes most highly 

 charged. Probably many other like results may appear, or be 

 hereafter devised, and may greatly help to assist the discussion of 

 the question of physical lines of force now under consideration. 



3292. When, in place of considering the medium external to a 

 magnet as homogeneous or equal in magnetic power, we make it 

 variable in different parts, then the effects in it appear to me still 

 to be in perfect accordance with the notion of physical lines of 

 magnetic force, which, being present externally, are definite in di- 

 rection and amount. The series of substances at our command 

 which affect the surrounding space in this respect, do not present 

 a great choice of successive steps ; but having iron, nickel and 

 cobalt, very high as paramagnetic bodies, we then possess hard 



