Nov. 28. 1S78] 



NATURE 



81 



a special bearing on that -nhich follows, we will mention 

 them in detail. 



Let two bar-magnets of steel be placed on the table 

 with the north-seeking pole of one towards the south- 

 seeking pole of the other, but not touching. Over these 

 lay a sheet of stiff writing-paper, or card, or a sheet of 

 window-glass. FiU a pepper-box with fine iron-filings, 



FiG. 6. 



the continuous curves from pole to pole. Suppose you 

 could obtain a piece of steel imbued with magnetism of 

 one Kmd of polarity only— a magnetic " particle," in fact, 

 ot the same kmd of magnetism as the north-seeking pole. 

 It you were to put that magnetic particle down on one of 

 these hnes near the north-seeking pole on the right, it 



Fig. 5. 



and sprinkle them evenly over the sheet ; then tap the 

 sheet gently, until the filings have arranged themselves. 

 Observe (Fig. i) that the lines of force nm across from 

 pole to pole. A line of force represents the direction in 

 which the forces act. Suppose that the pole on the right is 

 the north-seeking pole, and that on the left a south-seeking 

 pole. The forces act across the space between them in 



would, as we know, be repelled away, since similar poles 

 repel one another. And it would move away along the 

 line of force (for that line of force represents the direction 

 in which the force acts), and would pass right over and 

 be attracted to the south-seeking pole on the left. Simi- 

 larly a magnetic particle of south-seeking polarity, if we 

 could get one and place it down on a line of force, would 

 be driven along the line in the opposite direction. 



Fig. 7. 



Notice, too, that a great many of the lines of force that 

 run out of one pole run into the attracting pole opposite. 

 This you will find always to be the case when two poles 

 attract : their lines of force run into one another. 



As a second experiment lay down the two magnets, but 

 put their north-seeking poles towards one another, and 

 then lay on them a sheet of card or glass, and sprinkle 





I .!V. 



Fig. 8. 



filmgs over, as before. The curves you obtain, which 

 are like those of Fig. 2, are quite different from those 

 obtained before. This time the lines do not run across 

 from pole to pole. They start out, but instead of uniting 

 and blending together, and bending over to run into one 

 another, they turn away sharply where they encounter 

 one another, and, without ever joining, swerve aside itt 



