496 WEBER ON THE MEASUREMENT 



in a semicircular form in the centre, to allow the tube in which the 

 bifilar wires are suspended to pass through them. In fig. 8, vv 

 exhibits the glass plate at the side ; v' to is the wooden plate, 

 with the flat parallel glass S at the other side ; v v' is one of the 

 upper glass plates. The letters kk are loops, through which 

 the conducting wires ffg and g' g' in fig. 6 descend; these wires 

 are fixed in these loops to avoid their lying loosely throughout 

 their entire length ; they terminate in pegs, or small cylinders. 



Fig. 9 also exhibits a vertical section, but at right angles to 

 that of fig. 8 ; jn is the multiplier, and R a section of the reel 

 vibrating within it. At the side of the case we perceive four 

 metallic knobs, marked uu' zz'. These are perforated crucially, 

 and the perforation most distant from the case is furnished 

 with a screw^ ; on the inner side of the case it is fixed to it by 

 another screw. Two of these knobs, u and u', are in metallic 

 contact with the commencement and termination of the multi- 

 plier, so that a current from the knob u can run through the 

 multiplier into the knob «', and vice versa. The other two knobs, 

 z and z', are perfectly isolated ; but all four of the knobs are 

 very useful for reversing the current, and for effecting various 

 combinations. In this figure also we see the index vibrating 

 above the scale-plate, as also in fig. 3, where the case is sup- 

 posed to be removed. 



Let us now trace the course of a galvanic current, which 

 enters the instrument at the knob u ; it passes from u through the 

 multiplier m and towards ti' ; if the conducting v.ireg' g' with its 

 metallic cylindrical extremity be now inserted into this knob, 

 the current ascends in g'g', and (fig. 5) towards the nut c' c^ 

 above the roller a', then down w ithin the tube to o'; thence (fig. 2) 

 from through the centre jy of the transverse beam to r' m' k d, 

 through the vibrating reel to ni rp o, and (fig. 5) to o, ascending 

 above the roller a in the nut c c, to the second conducting 

 Avire g g, and (fig. 9) through g g down into the knob z, whence 

 it runs into the other of the two exciting surfaces. 



By means of the upper rotating part of the serpentine 

 pedestal, the instrument may be arranged in any part of a hall 

 or room as required. All the figures are drawn one-fourth of 

 the linear magnitude of the electro-dynamometer, excepting fig. 5, 

 which is one-half the real magnitude. 



The wire on the vibrating reel is 200 metres in length, that of 

 the multiplier 300; the first forms about 1200 coils, the latter 

 about 900. The length of the bifilar wires, (which are very fine, 



