SCIENCE AND PRACTICE. 245 



and this can only occur when the line bisecting the angle a 

 stands at right angles to the line N s. 



When the magnetism of the two needles is of different 

 intensity, let n s have the magnetism f t and n s' an amount 

 equal to /' ; then, in order that the needles take a certain 

 direction, the forces must be balanced, or 



I __ 2/' 

 f ~ 2/ 



Now the limits of the possible values of _ . t are evidently 



I 



1 and oo on the same side of the line N s. The value of __ 



is 1 when the line bisecting the angle a is at right angles to 

 N s, and it will be infinite when n' s and N s coincide ; but 



when this takes place it follows that the proportion iZ. 



between the magnetism of the needles must also be infinite, 

 that 2 /, the magnetism of the needle n s, must be very small 

 infinitely so in comparison with 2/', the magnetism of 

 the needle n' s. Therefore, when the disproportion between 

 the magnetisms of the two needles of an astatic pair is very 

 great, the stronger magnet points magnetic north and south. 



Between these extremes, as the relation -Z. becomes finite, 



the needle pair places itself at various angles between 

 and 90. 



19. Magnetism of the Coils. Copper is not a magnetic 

 metal ; nevertheless, Dubois, TyndaLL, Melloni, and others, 

 have found large multipliers of insulated copper wire mag- 

 netic to a degree great enough to cause a permanent deflec- 

 tion of the astatic pair to 30 where no current passed through 

 the circuit. Suspended within such magnetic coils, the 

 needles usually show a disinclination to come to rest on the 

 zero line, but take up with equal facility a position on either 

 side of it. The magnetism of the copper coil has been 

 variously attributed to the mixture of magnetic metals in 

 the copper, to the iron which adheres to the wire as it leaves 



