6 Prof. W. G. Adaras on the Forms of 



the curves in the neighbourhood of the axis to be similar to those which 

 .could have been traced on a sheet of infinite extent. 



The third equip otential curve from the negative pole C is the rect- 

 angular hyperbola, and its vertex O divides the distance AC ; so that 



AO is to AB as 1 to si 2. 



AO is equal to 53*75 millimetres. 



Case 6. Plate 2. fig. 6 represents the case of a circular disk where the 

 current enters at the edge and leaves at the centre of the disk. 



Around the centre the curves are very nearly ellipses of small eccen- 

 tricity, the focus being at the centre of the disk. If a is the radius of 



. . 2r 



the disk and r the distance from the centre, the eccentricity is 



a 



The curve cutting the axis at the point L at a distance of 16 millims., 

 i. e. (3 — 2s/ 2) a from the centre, has two branches cutting one another 

 at right angles at the point Q, and each cutting the edge of the disk at 

 an angle of 45°, the radius of the disk being 3*75 inches. 



"When the fixed galvanometer-electrode is at L, it is difficult to find a 

 succession of points forming a continuous equipotential curve ; the 

 tracing-electrode may at one time be placed on the boundary of the 

 shaded portion of the figure, and at another may be placed on the axis 

 near the point Q, without causing any current through the galvano- 

 meter. 



On placing the fixed galvanometer-electrode at Q, the tracing-elec- 

 trode marks out two straight lines in the neighbourhood of that point of 

 the same potential as the point Q, and each cutting the edge of the disk 

 at an angle of 45° at that point. 



The uncertainty in tracing this equipotential curve is explained, as 

 will be presently shown (p. 25), by the fact that each of the galvanometer- 

 electrodes was rather more than 1 millimetre in diameter. 



The equipotential curves which lie further from the centre cut the 

 edge of the disk at right angles. 



To determine experimentally the Lines of Flow and the Equipotential 



Surfaces in Space of three dimensions. 



If two platinum wires sealed in glass tubes, with only a short piece of 

 wire projecting from the sealed end, be immersed in a liquid, the other 

 ends being connected with the poles of a battery, we shall have a close 

 approximation to the case of currents flowing from one point to another 

 within a liquid ; and by means of two other platinum wires similarly 

 arranged but attached to a galvanometer, we may trace out the forms of 

 equipotential surfaces within the liquid. 



If dilute sulphuric acid be employed there will be polarization on the 

 electrodes ; but by reversing the current alternately, and making contact 



