THE SMITHSONIAN INSTITUTION. 369 



rest upon metcal springs, one of whicli presses against tlie front end 

 of the Lrass cylinder 7i, the otlier against the brass ring of the 

 wooden cylii)(ler, (the springs are not shown in the figure.) The 

 winch 711, which can be removed, serves ibr turning the cylinder 

 about its axis. Placing it on the cylinder A, and turning to the right, 

 the wire is unwound from the wooden cylinder and wound upon the 

 brass one ; on the other hand, placing it upon g, and turning to the 

 left, the reverse tak'^^ place. Since the coils are insulated on the 

 wooden cylinder, and kept apart by the screw-thread, the current 

 traverses the wire throughout its whole length on this cylinder ; but 

 on the brass cylinder, where the coils are not insulated, the current 

 passes at once from the point where the wire touches the cylinder to 

 the s})ring at k. The resisting part of the length of the wire is there- 

 lore the variable porJon which may happen to be on the wooc'en cylin- 

 der. 



There are forty screw-threads of the wooden cylinder to an inch. 

 The wire is of brass, and 0.01 of an inch in diameter. 



For counting the number of coils uuAvound, a scale is placed be- 

 tween the two cylinders, and the fraction of a turn is estimated by an 

 index fastened on the axis of one of the cylinders, and which points., 

 to the divisions of a graduated circle. 



§ 2*7. Differential measurer of resistance. — For determining the* 

 resistance of metallic wires, Wheatstone has given a very simple pro- 

 cess. The rheostat is inserted in the conducting arc of a constant 

 element with the galvanometer and the wire whose resistance is to be 

 determined, and the whole resistance is so regulated that the needle 

 can come to rest at any desired point a of the graduated circle. Now, 

 removing the wire from the circuit, the needle will indicate a greater 

 deflection, and to bring it back to the point a, a definite number of 

 turns of the' rheostat must be added to the existing resistance. We 

 find in this manner how great the resistance of the wire in question 

 is, expressed in turns of the rheostat. 



By this method neaily equally accurate results are obtained, whether 

 a multiplier, the much less sensitive tangent compass, or any other 

 galvanometer, be used. The reason is as follows : To produce in a 

 tangent compass a deflection of, say 45°, the entire resistance of the 

 closing conductor must not be very great. Suppose R is the entire- 

 resistance of the whole battery, and an increase or decrease r of this* 

 resistance produces such a change in the strength of the current that 

 the deflection of the needle is varied by 1°. 



Now, by using a multiplier, which is about 150 times more sensi- 

 tive than the tangent compass, the entire resistance of the battery 

 must be about 150 R to cause a deflection of the needle of 45°. 



To produce a like change in the strength of the current as that above- 

 mentioned, the resistance must now be increased or decreased by 150 r. 

 But, since the multiplier is 150 times more sensitive than the tangent 

 compass, the 150th part of this change of resistance, or r, will suffice 

 to advance or bring back the jjosition of the needle by 1° ; thus the 

 same change of resistance r produces in both instruments nearly equal 

 changes of deflection. 

 24 



