252 DESCRIPTION OF A MAGNET! METER, BEING A NEW 



4. A bar of soft iron, held in any position, except in the 

 plane of the magnetic equator, maybe rendered magnetical by 

 a blow with a hammer, or other hard substance ; in such cases, 

 the magnetism of position seems to be fixed in it, so as to give 

 it a permanent polarity. 



Exp. 1. — Bar No. 1. freed from magnetism, and held perpendicularly 

 in the hand, was struck a smart blow over the upper end with a small 

 hammer ; though it had pre^dously evinced no attraction for the 

 compass-needle 1^ inch distant, when the elevation of the move- 

 able plate was 21|^°, the upper end of it now attracted the north 

 pole of the needle upwards of a point, at the same angle of eleva- 

 tion. The lower end of the bar produced an equal repidsion. 



Exp. 2. — The same bar was now inverted, held as before, and a blow 

 again struck on the upper end. The poles of the bar were found 

 to be reversed ; that which was before north had become a south 

 pole. 



Exp. 3. Many more experiments were made on the effect of blows, 



from which it appeared, that when the iron was held in or near 

 the position of the magnetic axis, a blow on any part of the bar, 

 struck in any direction, or with any hard substance, such as a ham- 

 mer, a piece of copper, brass, or ivory, or even vdth a bit of wood, 

 invariably rendered the bar magnetical, the upper end becoming a 

 south, and the lower end a north pole. In these experiments, the 

 iron, when struck, was held nearly vertical in the hand, without 

 resting on any thing ; yet both ends seemed to acquire an equal 

 degree of magnetism. 



Remark. The fixing of the magnetism of position in iron or steel 



by hammering, may be employed with advantage in the experi- 

 ment of producing a magnet, with the use only of unmagnetised 

 bars of steel and two rods of iron. For in the commencement, the 

 bars of iron and steel may have peiinanent polarity given them by 

 hammering in a vertical position, by which the process wHl be con- 

 siderably shortened. 



5. An 



