532 



PROFESSOR KNOTT ON SOME RELATIONS BETWEEN 



Current - 18, wire tapped 

 Current - 24, „ „ 



- 45° twist applied ; deflection rises to 650 







+ 10 



20 



30 



45 



30 



20 



10 







-10 



-20 



-30 



-45 



-30 



-20 



-10 







then falls and 



settles at 



Deflection. 

 580 

 589 

 535 

 620 

 650 

 687 

 711 

 750 

 695 

 630 

 567 

 545 

 610 

 590 

 570 

 540 

 590 

 655 

 655 

 660 



From this condition of apparent negative polarity the wire was reduced by means of 

 a series of alternating and diminishing currents, being vigorously tapped at every stage. 

 Finally, with 



Current - 5, the deflection was 

 Current off, „ „ ,, 



600 

 596 



In this condition the wire, although set longitudinally in the earth's horizontal field, 

 had practically no action on the magnetometer. Then followed the twistings. 



0° twist 

 - 5 ,, applied 





 + 5 





 + 5 





 + 5 







Deflection. 

 597 

 578 

 600 

 625 

 606 

 585 

 604 

 625 

 606 



Thus a very slight twist is sufficient to disturb the equilibrium of apparently 

 balanced magnetisms in a wire. 



At first, the inductive effect of the earth's field was balanced by the superposed 

 magnetic effect of a current flowing in a coil surrounding the wire. This negative 

 current must, as is well known, be a little greater than the positive current, which, acting 

 alone, would have produced a field equal to that of the earth. My belief is that the 

 principal magnetic effect of the negative current is confined for the most part to the 

 surface layers. Here the molecular groupings may be supposed to be reversed as regards 

 polarity. These, by their very reversal, will have a sustaining effect upon the contiguous 



