296 



Mr. J. Brown. 



[Nov. 18, 



5. The principal piece of apparatus used was a bi-metallic 

 quadrant electrometer represented at fig. 1, about one-fourth actual 

 size. Its design was an elaboration of the temporary apparatus 

 described in a previous paper ('Phil. Mag.,' voL 2, 1878, p. 142). A 

 is a metal case supported on levelling screws ; B a thick vulcanite 

 tube provided at its upper end with a brass cap containing a collar of 

 leather in which slides the brass rod C. D is a brass ball with 

 clamping screw to retain the rod at any desired height. E is a 

 platinum torsion wire, O'OOl inch diameter, carrying the stouter wire 

 F, on which are fixed a light concave mirror G, the " needle " H, and 

 a glass weight J dipping in water or other liquid contained in a glass 

 vessel, resting on a vulcanite plate on the bottom of the instrument. 

 R and L are the quadrants of the metals under examination screwed 

 to a vulcanite piece M which is adjusted by hole, slot, and plane on 

 the points of three long fine-threaded levelling screws, 0, P, Q, 

 passing up through stuffing-boxes in the bottom of the instrument. 

 The quadrant L is connected to an insulated wire N passing oat air- 

 tight through the back of the case. R is connected to the body of 

 the instrument and to " earth " by a slip of foil under the vulcanite 

 support touching the screw 0. A stopcock T in the top of the 

 vulcanite tube and another in the bottom of the instrument of which 

 the opening is shown at S, provided means for varying the nature of 

 the gas surrounding the metals. The opening in front of the instru- 

 ment was provided with a flange, faced up to a plane, on which a pane 

 of plate glass was jointed by grease or other suitable means. The 

 needle was electrified either positively or negatively by connecting 

 its suspending wire through a Pohl's reversing commutator to either 

 pole of a battery of 100 small Daniell cells : the opposite pole being 

 at the same time connected to the body of the electrometer and to 

 "earth." 



6. To make an experiment with this apparatus the quadrants were 

 first carefully adjusted with their upper surfaces in the same plane, 

 then put in position, roughly levelled, the connexion to wire N made, 

 and the pair dusted with a camel-hair brush. The needle which had 

 been raised by the sliding-rod C was now lowered to within about 

 0*05 inch of the surface of the quadrants, and adjusted with its 

 centre line over the slit between them and its suspension wire in the 

 centre of the system. 



If now on connecting the quadrants metallically, and electrifying 

 the needle positively and then negatively, it was found that the 

 deflection (as observed by lamp and scale) was greater on one side of 

 zero than on the other, due to one quadrant being higher than the 

 other, the necessary adjustment was made by means of the screws O 

 and P. A second concave mirror fixed on the ball D was now 

 clamped, so that its reflected beam also fell on the zero of the scale, 



