S78 Mr. W. Petrie's Results of Experiments 



Experiments on the Shape of Magnets. 

 Eight steel bars were made of nearly the same hardness 

 and quality, being six inches long, half an inch broad, and of 

 various thicknesses, from 0*5 inch to 0*02 ; they were all 

 magnetized by the single touch, being repeated until they 

 would become no stronger ; the power was measured by the 

 force exerted at the extremity to turn into the magnetic meri- 

 dian, when placed at right angles to it ; this force, which we 

 will call fn, was ascertained by calculation from the formula 



"W being the magnet's weight. 

 jj2L N its number of single vibrations per 



f n = W ' 824-00(n minute - 



L the total length of the magnet in 



r inches. 



The fn of all these magnets was the same, within the errors 

 of experiment. Similar experiments were made with steel 

 bars of 18, 6 and 2 inches in length, all being 0*5 inch broad; 

 the fn of each was in the exact ratio of their length, as was 

 partially ascertained by Captain Kater* and Coulomb. 



Similar experiments were made with bars of 3, 0*5 and 

 0*25 inches in breadth, all being six inches long ; the ratio of 

 their fn was nearest that of the square root of the breadth 

 (the smallest being too strong and the largest too weak ; 

 probably the ratio of the square roots would have been cor- 

 rect had each been rubbed with a magnet of corresponding 

 breadth). 



Similar experiments were made with bars of 0*5, 0'25 and 

 0*15 inch square, all being six inches long; the ratio of their 

 fn was between that of the square and cube root of the 

 breadth of their sides, being nearest the square root. 



The relativeyw of magnets of the accompanying forms, is 

 as follows : their length was four inches, and breadth 0*8. 



4' 



7 



12 



^They were magnetized in the 

 direction of the shade lines. 



Effect of approximating Magnets. 



Two bar magnets, six inches long, 0*5 inch broad,and 0*05 

 inch thick, were placed parallel with similar poles adjacent 

 flatwise, one over the other at various distances, and their 

 magnetic power ascertained by vibrations, according to the 



[* Capt. Kater's paper on this subject will be found in Phil. Mag., vol. 

 lix. p. 359.— Edit.] 



