ON THE MAGNETIC EFFECT OF ELECTRIC CONVECTION 



137 



J = -380, 

 For this case from equation (1) 



1 



295000 



and from the electrification 



= -2926. 



= -00000339 , 



= -00000355 . 



281500- 



The error amounts to 3, 10 and 4 per cent respectively in the three 

 series. Had we taken Weber's value of v the agreement would have 

 been still nearer. Considering the difficulty of the experiment and 

 the many sources of error, we may consider the agreement very satis- 

 factory. The force measured is, we observe, about ^inr of the hori- 

 zontal force of the earth's magnetism. 



The difference of readings with -f- and - - motion is due to the 

 magnetism of rotation of the brass axis. This action is eliminated 

 from the result. 



It will be observed that this method gives a determination of v, the 

 ratio of the electromagnetic to the electrostatic system of units, and if 

 carried out on a large scale with perfect instruments might give good 

 results. The value v = 300,000,000- metres per second satisfies the 

 first and last series of the experiments the best. 



Berlin, February 15, 1876. 



