ASTATIC NEEDLES. 219 



space comprised within the two poles and to be able at need to 

 place the two poles in the prolongation of the needle. 



841. ASTATIC NEEDLES. A second method, due to Nobili, 

 consists in using a system of astatic needles (299). The action 

 of the earth on the system may be reduced at pleasure. By placing 

 one of the needles within the coil and leaving the other outside, the 

 action of the coil is slightly increased, the action on the external 

 needle, which is moreover very slight, being always in the same 

 direction as the principal action. 



The apparatus is more symmetrical when two superposed coils 

 are used, with a needle in x the middle of each, and passing the 

 current in contrary directions in the two coils. As a system almost 

 astatic tends to set at right angles to the meridian, this property, as 

 well as the time of oscillation may be used to investigate the degree 

 of compensation of the system. 



842. BIFILAR SUSPENSION. A third method, pointed out by 

 Gauss,* consists in supporting the bar by a bifilar suspension, so 

 that its position of equilibrium is still in the meridian, but opposite 

 that which it takes under the action of the earth. The resultant 

 directing couple is equal to the excess of the bifilar couple over the 

 terrestrial couple. This difference may be made very small ; but the 

 experimental arrangement is less convenient than the preceding, and 

 is only suitable for very heavy magnets. 



843. DAMPING. With instruments which are not damped the 

 observations are always long and tedious. Although it is un- 

 necessary to wait until the magnet is quite at rest, yet we must 

 wait until the amplitudes are small enough to deduce the position 

 of equilibrium of three successive elongations (695). When the 

 galvanometers have a very slight damping of their own, the needle 

 may be stopped by means of a small magnet held in the hand, which 

 is so manipulated as to act in opposition to the oscillations. It is 

 convenient in that case to use a magnet jointed like a compass, the 

 action of which is inappreciable when the two branches touch. It 

 is often advisable to adjust near the galvanometer a coil through 

 which the current of an auxiliary battery can be sent in either 

 direction by means of a key in the hands of the observer. Yet for 

 ordinary galvanometers it is better that the instrument should have 

 an appreciable damping of its own. 



The natural damping of the oscillations is due to the resistance 

 of the air and to the induction currents developed by the motion 



* GAUSS. (Euvres, Vol. v., p. 367. Resultate des Magn. Verdns, Vol. I. 1837. 



