266 M. G. Quincke on Electrolysis^ and the 



and there is an increased degree of conductivity in the vicinity of 

 the platinum electrodes. This change in the conducting-power 

 of the water by the solution of the glass covering of the elec- 

 trodes would then explain the imperfect agreement of the ob- 

 served and calculated values of the condenser-charges in the 

 Table at the beginning of this section ; for there would thereby 

 be produced a change in the curves of the current : the electri- 

 city v/ould no longer flow in the direction of the normal of the 

 section of the column of water; and so the potential of free elec- 

 tricity could not be constant within the same section. 



§59. 



The preceding experiments have the disadvantage that they 

 require a condenser, the thick insulating shellac layers of v/hich 

 may readily become electrical v/ithout its being noticed, and 

 thereby occasion a tendency to one of the two electricities. The 

 electrometer which I used is not free from this objection. 



Hence, in order as much as possible to avoid this source of 

 error, I have recently made similar experiments with a Thom- 

 son's quadrant electrometer*, which may easily be brought up 

 to such a degree of sensitiveness that a condenser may be dis- 

 pensed with, even when a Grove's battery of only a few cells is 

 used. Its indications are also more perceptible, quicker,.and less 

 open to external accidental sources of error than those of Dell- 

 mann and Kohlrausch's electrometer. 



Thomson's electrometer consists essentially of a horizontal 

 thin aluminium plate which has a bifilar suspension by two co- 

 coon threads, and by means of a thin platinum wire is placed in 

 conducting communication with a constantly charged Leyden 

 jar, the outer coating of which is in conducting communication 

 with the earth, Fixed to and above the aluminium plate is a 

 small silvered concave mirror with a radius of about a metre, which 

 projects the image of a narrow petroleum flame upon a hori- 

 zontal scale. The position of the image determines, in the usual 

 manner, the deflection of the aluminium plate. 



The aluminium plate is suspended within a hollow space, which 

 has the form of a cylindrical box made of thin metal foil and se- 

 parating into four quadrants insulated from each other and 

 open inwards. Each two quadrants which belong to the same 

 diameter are in conducting communication with each other. 

 Both pairs of quadrants are symmetrically arranged in reference 

 to the aluminium plate; and this does not alter its position 

 when both pairs of quadrants are in conducting communication 

 or have the same electrical tension. 



But if one pair of quadrants is connected with one, and the 



* Brit. Assoc. Rep. 1867, p. 490, 



