OF ARTS AND SCIENCES. 203 



Perfect demagnetization is attained with varying difficulty. Ordi- 

 narily, it is rapidly and easily accomplished. Sometimes a longer 

 time is required, or a more intense action. In numerous experi- 

 ments, we derived the alternating current from a Wilde machine 

 which gave about 6,000 reversals per minute. This rate is probably 

 greater than is required or desirable, except in extreme cases. About 

 3,000 to 4,000 reversals would be a better general rate, although of 

 course the operator should arrange his apparatus so that he could get 

 more reversals if necessary. Failure will result if the speed is too 

 great, as might be expected if the view here taken is correct. Usually, 

 it will be more convenient to employ, instead of an alternating ma- 

 chine, a battei-y with a reverser arranged for varying speeds. 



One application of this method is to the demagnetization of watches. 

 Watches strongly magnetized are completely demagnetized by one to 

 three minutes' exposure in the coil. Frequently unsuspected traces of 

 magnetism cause annoying irregularity of action of a watch. This 

 method enables us entirely to remove this difficulty. 



Some very curious and interesting results were obtained by experi- 

 mentiujj with magnetite. 



A specimen of very pure magnetite, from North Carolina, showing 

 marked magnetic properties, was completely demagnetized by a some- 

 what long exposure to the action of an alternating current applied as 

 described above. Before demagnetization the piece had shown con- 

 sequent points, although in general it possessed polarity. After 

 treatment it attracted either end of a very light suspended needle 

 indifferently, and when any part of the mineral was presented, just 

 as a piece of soft iron would do. The demagnetized specimen was 

 then placed across the poles of an electro-magnet excited by a strong 

 current from a Gramme machine. It became strongly magnetic, with 

 distinct poles, and without the consequent points it at first had. 

 After this it was treated like an ordinary bar magnet, and magnetized 

 or demagnetized at will. 



Another more impure piece was originally less strongly magnetic, 

 and was demagnetized with great difficulty. At first it displayed no 

 general polarity, having consequent points irregularly distributed. 

 After demagnetization it received induced magfnetism and became 

 polar, but it was a much feebler magnet than the previous specimen. 

 Demagnetization was afterward performed more easily than at first. 



A still more impure specimen was treated, but with the means at 

 hand it was not perfectly demagnetized, although so nearly was this 

 done that only traces of magnetism were noticeable. 



