dependent on the Motion of the Conductor. 91 



Therefore the change produced in the current-intensity by 

 putting the conductor in motion is proportional to the velocity 

 \ and the cross section a of the conductor. When the resist- 

 ance R is so inconsiderable that it may be neglected in 

 comparison with r , the alteration of the current-intensity is 

 independent of the resistance. 



At the first glance it might be supposed that the velocity hi 

 which with accessible means can be imparted to the conductor 

 would be so slight, in comparison with the velocity h of the 

 aether, that the alteration in question (of the current-intensity) 

 could not be observed even with the most delicate galvano- 

 meter. The explanation, however, given in the above-men- 

 tioned memoir (p. 56), of Quincke's diaphragm-currents, inti- 

 mates that the observations must be practicable ; and on this 

 account I resolved to test the theoretical result by experiment. 



II. 



Special mechanical arrangements would be required if we 

 would give to solid conductors the velocity necessary for these 

 experiments. I therefore employed liquids, which can with 

 facility be set in motion of sufficiently great velocity. But 

 here two inconveniences appear, the injurious effects of which 

 must be carefully avoided : one of them is the polarization of 

 the plates by which the current is conducted into and out of 

 the liquid ; the other consists in the difficulty of making the 

 two plates galvanically so nearly equal to each other that a 

 feeble current shall not be produced by the inequality as soon 

 as they are placed in the liquid. Yec neither the polarization 

 of the plates nor their inequality in a galvanic point of view 

 would injuriously affect the experiments, if these two conditions 

 did not vary as soon as the liquid is put in motion. But it 

 appears that such variation really takes place ; and therefore 

 the experiments cannot be made by simply sending a galvanic 

 current through the liquid and measuring the intensity, first 

 while the liquid is at rest, and again after it has been set in 

 motion ; for, as both the polarization and the galvanic quality 

 of the plates are altered by the motion of the liquid, the ob- 

 served difference in the current-intensity may just as well be 

 attributed to this alteration as to a change in the resistance of 

 the liquid. 



After putting to the test of practice some proposed methods 

 of observation which need not be here discussed, and finding 

 them not advantageous, I finally adopted the following, which 

 gave good results. (See fig. 1.) 



a b is a cylindrical glass tube of uniform diameter and fur- 

 nished with three side-tubes c, d, and e ; the distances between 



H2 



