228 MEASUREMENT OF CURRENTS. 



by a silk fibre, about i cm. in length, in a cavity of rectangular 

 section, just sufficient to allow of very slight motions on either side 

 of the position of equilibrium. Induction currents have only an 

 unimportant part in the damping, from the smallness of the needle, 

 its distance from the coil, and from the great resistance which the 

 wire usually presents. The coil is circular, and the section of the 

 channel is a rectangle, constructed according to the theoretical curve 

 of best winding (733). By means of a correcting magnet, the 

 directing force which acts on the needle can be varied at will. 



Sir W. Thomson has constructed galvanometers of this form, in 

 which the diameter of the several wires varies according to the law 

 indicated by theory (735). Each layer communicates with an 

 external binding screw, so that in the central part we may take 

 that length of the wire the resistance of which is equal to that of 

 the external circuit. We may thus work within the conditions of 

 maximum sensitiveness for all experiments. 



In the astatic galvanometer (Fig. 163) the two needles of the 

 system are placed separately at the centre of a circular coil. Each 

 of these coils consists of two, which fit against each other, only 

 having sufficient space between them to permit the passage of the 

 aluminum rod which connects the needles. This rod supports the 

 mirror of one of the coils, and in the centre of the other there is a 

 thin plate of aluminum or mica, for damping. The current is passed 

 in opposite directions in the two coils, so that the actions may add 

 themselves. A correcting magnet, which modifies unequally the 

 action of the field on the two needles, renders it possible to vary at 

 will the sensitiveness of the apparatus. If the system of needles is 

 very nearly astatic, the position of equilibrium depends on the 

 magnet only, and the zero is more stable. 



The resistance of the coils, unless for instruments intended for 

 experiments on radiant heat, is generally considerable, and amounts 

 to 8,000 or 10,000 ohms. 



These galvanometers have very rapid oscillations, which speedily 

 damp themselves. Their sensitiveness is very great, and they are 

 particularly suited for methods of reduction to zero, in which the 

 existence and the direction of a very feeble current is to be 

 determined. 



850. SIEMENS' GALVANOMETER. This galvanometer is especially 

 noticeable from the needle, which is a kind of horse-shoe magnet in 

 the shape of a cylindrical bell, grooved in a diametrical plane for 

 great part of its length. This turns about its axis in a hollow 

 cylindrical cavity in the centre of a solid sphere of copper. As the 



