Prof. E. Wartmann's Fifth Memoir on Induction. 93 



total current between the two circuits may be increased by 

 shortening the arm for, and thus approaching the point z" to 

 the opening v. We have even inserted the end of the wire a 

 m- 025 into the channel AB, the diameter of which was larger, 

 so that there still remained a thin ring of mercury around 

 the wire. 



156. The position of the extremity of the wire /3 may be 

 changed at will, and all possible arcs be described around the 

 point v, by the terminal surface of the wire a, giving any value 

 to the ray of rotation for : the deviation of the rheometer re- 

 mains absolutely constant. This is due to the great con- 

 ducting property of the mercury ; for if a saturated solu- 

 tion of sulphate of copper is substituted for it, differences are 

 found in the position of the index. But these differences are 

 produced by the variations of resistance of the filaments of 

 liquid interposed between the aperture of the tube AB and 

 the extremities of the circuits. 



157. The same constancy in the rheometric indication con- 

 tinues when the depths of immersion of the wires a and /3 are 

 modified in a successive or simultaneous manner. 



158. The result of these experiments cannot be attributed 

 to a phenomenon of internal reflexions. Admitting a perfect 

 cylindricity of the vessel DD', all the electric radii would have 

 been brought, by lateral reflexions, toward the aperture v of 

 the lube AB which occupies its centre. Although the vessel 

 does not present exactly this geometric form, it differs so little 

 from it that we may admit that a very small number of reflected 

 rays reached the wire /3 immersed in z at m, 13 of the aperture. 



159. I also wished to demonstrate, that the relative distri- 

 bution of electricity between the two wires a and /3 is not due 

 to a total reflexion of the fluid on the surface of the mercury; 

 and with this view I constructed the apparatus represented in 

 fig. 5. A large copper wire ttt is curved so that its extremities 

 become parallel and not far apart : they are moreover encased 

 in glass tubes tin, which leave only the terminal surfaces of the 

 wire exposed. The latter is fixed between two pieces of 

 wood H, one of which is pressed against the other by means 

 of a clamp x. A slit y permits of establishing this forked 

 system on the glass plate bb, and extending it over the entire 

 length right and left of the cylinder d. The extremity of the 

 wire /3 is then introduced into the compartment D'. Expe- 

 rience proves that the deviation of the rheometer is in no de- 

 gree affected by the diversities of position and immersion of 

 the wires /3 and t, whether they are or are not accompanied 

 by analogous changes in the place and depth of immersion of 

 the other wire L the extremity of a. 



