92 Prof. E. Wartmann's Fifth Memoir on Induction. 



portion is isolated in a coating of wax. These two pieces have 

 their extremities amalgamated : they constitute a very simple 

 rheotrope, which has proved very useful to me in a great 

 number of researches*. 



J 53. The reservoir C, the lube AB, and the two compart- 

 ments D, D' are filled with purified mercury up to the general 

 level 0' 0" 0'". C is then connected with the cup E by the 

 metallic wire 8. The current then follows the mercury of the 

 channel AB. To conduct it back from D into F, two ways 

 are presented, the resistances of which are equal, or nearly so. 

 One is the wire «, which terminates at the aperture f of the 

 piece represented in figs. 2 and 3 ; an excellent Gourgon's 

 multiplicating rheometer G, the needles of which only make a 

 simple oscillation in twenty-six seconds, was interposed in its 

 development. The other way is furnished by the wire /3 of 

 determined dimensions, and which is immersed at z in the 

 mercury of the chamber D : it communicates with this liquid 

 only by its lower point, and is isolated for a sufficient length, 

 by means of a covering of glass. This arrangement is analo- 

 gous to those which I have described in employing the method 

 of derivations (74 to 79.). 



154. The principle on which this mode of experimenting is 

 founded is easily understood. Let us suppose that the wires 

 a and /3 exert an equal resistance, and that the second termi- 

 nating in z, at a distance Bz from the extremity of the chan- 

 nel AB, any suitable lengthening of the stem i causes the first 

 to terminate in z', the length Bz' being equal and symmetrical 

 to B#. Each of them will then afford a passage for a current 

 of the same intensity, and that intensity will be half that which 

 flowed off by the channel AB. Now the pair L* being of small 

 dimensions and feebly excited, whilst the wires a and /3 are 

 thick and good conductors, any one of the latter, taken singly, 

 would suffice for the discharge. If the electricity is propa- 

 gated by radiation, the portions of the total current transmitted 

 by each of the circuits will vary with the respective positions 

 of the points of contact of the wires with the mercury. The 

 extremity of /3 being constantly immersed in z, let us bring 

 that of a into z". Thus placed in the prolongation of the 

 axis of the tube, and in the course of the electric rays which 

 issue from it in a parallel bu?idle, this wire « will be traversed 

 by a stronger current, and the rheometric deviation will in- 

 crease. The quantity of the fluid which will run off by the 

 wire /3 will be proportionately diminished. 



155. This inequality in the distribution of the parts of the 



* Memoir es de la Societe de Physique et d'Histoire Naturelle de Geneve, 

 vol. ix. p. 1 19. Archives de VElectricite, vol. i. p. 74. 



