370 



TENTH ANNUAL REPORT OF 



If the multiplier is required to indicate very minute changes in the 

 closing conductor, care must be taken that the corresponding diifer- 

 ence of current shall act in the multiplier, without a very considera- 

 ble resistance being inserted in the conductor. Wheatstone has ac- 

 complished this by means of the contrivance represented in Fig. 19^ 

 which he calls a differential measure?' of resistance. 



Fig. 19. 



On a board about 14 inches by 4 wide, the small brass knobs a, h, 

 c, and d are fastened, forming a paralellogram, and between a and d 

 are placed e and/, and g, h between d and b. These knobs, which 

 are furnished with binding-screws, are connected by wires, as seen in 

 the figure. 



One of the wires of the pole of the electro-motor is screwed in a, 

 the other in b ; tlie ends of the wires of the multiplier are fastened in 

 c and c?, so that the knobs c and d are in conducting connexion through 

 the multiplier m; between e and / a piece of wire is inserted, and 

 another between g and h. The currents here diverge in various 

 branches ; but we have to consider only those which pass through the 

 cnultiplier. 



A current passes from a to c, from c through m to cZ, from d past g 

 and h to b, as indicated by the unbroken line in Fig. 20 ; another 



Fig. 20. 



current, "which traverses the multiplier in the opposite direction, goes 

 from a, through e and/, to d ; from d, through m, to c, and finally 

 from c toi), as shown by the dotted line in Fig. 106. If the resist- 

 ances in the two conducting wires a, c, d, b, and a, d, c, b, are perfectly 

 equal, so are also the two currents passing through the multiplier 

 equal ; consequently the needle will remain at rest at the zero point. 



Now, by mafcidg the wire, inserted between e and/, only a little 

 longer or shorter., the two currents going in opposite directions 

 through tke multipHier will be no longer equal, and the difference of 



