ELEMENTARY MAGNETISM 381 



acted on, being along the line p q, which is no longer 

 parallel to the magnet. Verify this deduction by actual 

 experiment. 



In this way we might go round the entire magnet; 

 and, considering its two poles as two centres from which 

 the force emanates, we could, in accordance with ordinary 

 mechanical principles, assign a definite direction to the 

 magnetic needle at every particular place. And substi- 

 tuting, as before, a bit of iron wire for the magnetic 

 needle, the positions of both will be the same. 



Now, I think, without further preface, you will be able 



Fio. 12. 



to comprehend for yourselves, and explain to others, one 

 of the most interesting effects in the whole domain of 

 magnetism. Iron filings you know are particles of iron, 

 irregular in shape, being longer in some directions than 

 in others. For the present experiment, moreover, instead 

 of the iron filings, very small scraps of thin iron wire 

 might be employed. I place a sheet of paper over the 

 magnet; it is all the better if the paper be stretched on a 

 wooden frame, as this enables us to keep it quite level. 

 I scatter the filings, or the scraps of wire, from a sieve 

 upon the paper, and tap the latter gently, so as to liberate 

 the particles for a moment from its friction. The magnet 

 acts on the filings through the paper, and see how it ar- 



