TRANSVERSE EFFECT AND ON SOME RELATED ACTIONS IN BISMUTH. 227 



In diagram B, Plate P, with dotted circle to represent north pole above the paper, 

 the transverse effect is positive. 



To both ends of the plates strips of copper of the same breadth and thickness were 

 soldered ; to these latter, wires were soldered, which lead to the accumulators giving the 

 primary current. 



To two points in the middle of the sides of the bismuth plates w T ires were soldered ; 

 each wire was doubled on itself, the point of contact with the bismuth forming the bottom 

 of a V. Between the arms of the V mica was inserted to insure insulation. Both arms 

 were kept in the plane of the bismuth plate and perpendicular to its length. One arm 

 of each was joined to the galvanometer ; the other led to a mercury pool in the first 

 series of experiments, in the later ones it was unconnected. (Fig. C, Plate P.) 



Section II. — On the Effect on the Transverse Current of Inserting a Shunt 

 whose Resistance is of the same Ob der of Magnitude as that of the Plate. 



The transverse effect has up till now always been measured with a galvanometer whose 

 resistance was many times greater than that of the plate of metal experimented upon. 

 The question arises, How will this current be affected when a shunt is inserted between 

 the transverse electrodes whose resistance is of the same order of magnitude as that of 

 the plate ? If the plate when placed in a steady magnetic field behaves as a cell with 

 constant electromotive force would do, it will divide according to Ohm's law ; if, on the 

 other hand, it behaves as a cell whose electromotive force is not constant, the current 

 will not be obtainable from the equation — 



„, Electromotive Force 



Current = = — =— - . 



Resistance 



For example, Prof. Lommel, in his paper " Sichtbare Darstellung der aquipoten- 

 tialen Linien in durchstromten Platten. Erklarung des HaH'schen Phanomens," # has pro- 

 posed a formula, according to which the insertion of a shunt between the two transverse 

 electrodes would not affect the reading on a galvanometer whose resistance is great — com- 

 pared with the sum of the resistances of the bismuth plate and of this shunt. 



Each plate experimented upon was placed between the poles of the electro-magnet, 

 perpendicular to the lines of magnetic induction. 



Before the magnet was put on, a current from the accumulators at A was sent through 

 the plate. Two points, E and D, as near the middle as possible, were then found, so that 

 when wires joined them to the galvanometer G, no current passed. From E and D two 

 other wires lead to the mercury pools L and M respectively. Should it be found 

 impossible to find two points at the same potential, the current which goes through the 

 galvanometer circuit can be eliminated by joining E and N or D and N, as the case may 

 be, and inserting a suitable resistance. (Cp. fig. D, Plate P.) 



* Sitzungsbericht der Kong, bayerischen Akedamie der Wissenchaft, 1892, Bd. xxil. Heft iii. § 371. 



