284 MAGNETISM 



of it of the same shape and size as the glass, if the restoring force 

 in air was 0, 



KB fa QC <f> 



100 



whence /c. AT, on the air-water scale was X . 



( ). / > 



Full details of the work are not given, but apparently a bull) 

 was used containing in succession various liquids and ga-ex. and the 

 same bulb was used when exhausted so that the values of * relative 

 to a vacuum could be deduced. Taking a vacuum as having zero 

 susceptibility and K* K W = 100, Faradav found for K* the value 

 3-5, and therefore for water K W = - 96'5. The following Table 

 gives some of his measurement^ on this scale. Becquerel's \.-dues 

 for some of the same substances are given, adapted to Faraday's 

 scale. Becquerel took /c /or = 10 and *-, 0. To compare 

 them his numbers have been multiplied by 10. and have been added 

 to 3*4. The numbers for air and water must ofcour- 



Oxygen 17'5 '.Ml 



Air :$! 



Vacuum 



Carbon dioxide o 



Hydrogen 01 



Zinc -7H; -21fi 



Alcohol 



Water <>ll li - %<> 



Carbon disulphide - 99'6 



Sulphur -lls - in- 



Bismuth - 1 -21 



The closeness of the \ahus in many cases is e\ideme o! tin- 

 excellence of the work of the two experiment! ; 



Faraday made experiments at different distances and found 

 different results. The value for bismuth, for instance, rose <]uite 

 considerably as the distance increased. This has not been con- 

 firmed by later work, and there was no doubt some undetected 

 error important at the greater distances, but apparently not so 

 important at the half-inch distance, since the result ^ there obtained 

 are in very fair accordance with the best measurements. 



We shall now describe some of the later experiments, selecting 

 typical methods. 



Rowland's experiment. Rowland and Jacques * used 

 the method of a vibrating needle, of which the simple theory i^ 

 given on p. 260. According to that theory the time of vibration 

 of a needle-shaped body in the field is 



* Rowland's Phytical Paper*, pp. 75 and 1 -I. 



