116 M. F. Kohlrausch on the Determination of the Absolute 



the latter be determined by measuring the moment of inertia and 

 the time of oscillation, the two deflections will give the product iT 



i 

 and the ratio ^ in absolute measure, from which i and T can be 



determined separately. Just in this way, we know, the product 

 and quotient of the magnetism of a bar into the earth's magne- 

 tism are ascertained in Gauss's method. Since the deflections of 

 the tangent-compass and of the bifilar galvanometer are simulta- 

 neously observed, and the other measurements are independent 

 of the earth's magnetism, the result of these determinations gives 

 the horizontal intensity for the place and time of observation 

 without any further reductions. 



On a further comparison of the magnetic and voltaic me- 

 thods a few other points appear in favour of the latter. First 

 of all, the measurements of distance (here consisting of two mea- 

 surements of the length of wires) only require to be made once, 

 inasmuch as the coiling of the wires ensures the unalterability of 

 the dimensions. Then, again, the difficult comparison of the 

 magnetism of the needle in the meridian (in the oscillations) 

 with the magnetism of the needle which is at right angles to the 

 meridian (in the deflections) is quite dispensed with. In the 

 third place, the observation of the deflections at two different dis- 

 tances is got rid of, inasmuch as the deflection of the needle is 

 produced, not by a magnet in which the distribution of the mag- 

 netism is unknown, but by a current the distribution of which is 

 most accurately known, and which has been brought into the 

 simplest and most favourable form — that of a circle. Since 

 there is only this single determination, the needle which is to 

 Jbe deflected may easily be so short that an approximate estima- 

 tion of the position of the poles will be sufficientfor the correction 

 required. Finally, the magnitude of the forces and deflections 

 aue quite at our disposal, inasmuch as the current may have any 

 desired strength. 



Of the instruments, I shall have especially to devote to the 

 bifilar galvanometer a short description elucidated by some figures. 

 The suspended solenoid is a circular ring 67 centims. in 

 diameter, formed of eighty-four coils of a covered copper wire 

 3*2 millims. in diameter. The ends a of this wire (fig. 1) are 

 screwed and soldered to brass pins, which move with friction in 

 a vertical direction in a horizontal axis, and are fastened above 

 by means of immoveable nuts. The axis is placed in the rectan- 

 gular hooks of a suspension which is represented in fig. 1 half 

 its natural size. All the parts represented in the figure as shaded 

 are made of vulcanite; the others consist of gilt brass. 



The suspension consists of three disks on the same rod, s, 

 and turning against each other. In the middle one the rod is 



