6 Art. 9.— T. Terada : 



into shape as shown in Fig. 2, T. This was made to hang on the 

 looped end of the suspension wire ivw, for which fine Wollaston 

 wire of 5-10 /^ diameter was used. Tlie upper arc of the alumi- 

 nium hook is made to ride on the central V-shaped recess cc of 

 the stirrup. JM is a liglit plane mirror of 5 mm. diameter, attached 

 to the hook. The whole system was hung in the metallic case of 

 Mascart's magnetograph made by Carpentier, Paris, from which 

 the usual attachment in the interior was removed. The regulating 

 screw attached at the top of the case for adjusting the Itreadth of 

 the supension was utilized as such. For damping the natural 

 vibration of the suspension, an electromagnetic damper ]) made 

 of a copper block was introduced. The damper could Ije rotated 

 on a cylindrical brass block P, which in its turn could be rotated 

 about the screw S, fixed to the bed plate of the case, passing 

 through the groove cut along the projecting arm F at the foot of 

 L. The damping was very effective, making tlie vibration com- 

 pletely aperiodic. 



To set the suspension in working condition, we proceeded as 

 follows. At first, the stirrup only was hung on the wire, making 

 the breadth of the suspension sufficiently large. After hanging 

 the magnetic needle carefully in its natural direction, twist the 

 torsion head slowly till the needle is turned about 180°, and the 

 luminous image of the slit formed by the reflexion of the mirror 

 M appears within its proper range. Then, after bringing the 

 damper in position gradually narrow tlie breadth of the bifilar 

 suspension, at the same time adjusting the torsion head so as to 

 keep the luminous spot always wdthin the range assigned to it. 

 Proceeding very carefully in this way, the sensibility can l)e made 

 extremely great, till at last the suspension attains its unstable 

 position. It must l)e remarked that near this extreme position 

 any minute chan'ge in the elastic property of the wire, the breadth 

 of the bifilar suspension or the weight of the suspended system, 

 may sensibly affect the deflection. Nor is the system independent 

 of the slight inclination and the vertical acceleration of the 

 instrument. In one instance, an inclination of about 4' produced 

 a dis]3lacement of about '1 cm. on the photographic record. Since 



