265 



attract the needle till it is touched. Then it is that 

 one pole begins to attract one end and repel the 

 other. But even the repelling pole, will attract upon 

 contact, or at a very small distance. But how odd 

 are the following experiments. I cut a piece, says 

 Dr. Knight, of a loadstone into an oblong square. 

 In this I placed the magnetic virtue in such a man- 

 ner, that the two opposite ends were both south 

 poles, and the middle quite round was a north pole. 

 I made the two opposite ends of another stone, north 

 poles, the opposite side south poles. An irregular 

 stone had two broad, flat surfaces opposite to eacli 

 other. I made half of each of these surfaces a north, * 

 pole, and the other half a south pole. So that the 

 north pole of one surface was opposite to the south 

 pole of the other. I took a stone that had a grain 

 very apparent, running the lengthways of it. At 

 one end of it I placed a north pole, surrounded by 

 a south ; at the other a south surrounded by a 

 north pole : so that the edges of each surface had a 

 different pole from that which occupied the middle. 



Many varieties of this kind might easily be devised, 

 But these examples are sufficient to shew, how ma- 

 nageable the magnetic virtue is, with respect to its 

 direction : and how defective all the hypotheses are, 

 which are brought to account for the phenomena 

 of the loadstone. 



Mr. Howard sailed to Barbadoes in company with 

 another ship, commanded by one Groston. Suddenly 

 a terrible clap of thunder broke Groston's fore- 

 mast, and did some damage to his rigging. When 

 the noise was past, he was surprised to see Mr. 

 Groston's ship steering directly homeward. He tacked 

 and stood after him, and found that Mr. Groston did 

 indeed steer by the right point of his compass, and 

 that the card was turned round, the north and south 

 point having changed places. If he set it right with 

 his finger, as soon as if: was at liberty, it returned 

 to its former posture. And on examination^ he found 



VOL. n. N 



