72 
MESSRS. W. DE LA RUE AND H. W. MULLER ON THE 
Our results give a higher potential as requisite to produce a spark than the 
numbers of Sir William Thomson ; for example, compare in Sir William Thomson’s 
results 1278 volts, 67,260, with 1236 volts,. 84,590, in our table; again, 5208 volts, 
39,310, Sir William Thomson, with our numbers, 5562 volts, 44,460. In the first 
case our numbers are to Sir William Thomson’s as 1'258 to 1, and in the second 
IT 31 to 1 ; that is, in the first case 26 and in the second 13 per cent, greater than 
his. But the accordance of some and the discordance of others of our results with 
those of Sir William Thomson is best illustrated by the curves on Plate 6 , repre- 
senting on a reduced scale (the abscissae to §ths, the ordinates to the curves 
as originally laid down. Curves I., II., III., IV., V., VI., and VII. show the plotting 
down of the actual observations made by us ; curve VIII. the mean curve of our 
results ; IX. shows Sir William Thomson’s ; the first parts of VIII. and IX. are again 
given in curves VIII.® and IX.®, in the upper part of the plate, on the original 
unreduced scale. Sir William Thomson’s first observation was made with a 
difference of potential of 690 volts, and his result agrees very closely with our own. 
This accordance holds good up to 1690 volts, his lengths of spark rising a little more 
rapidly for definite increments of electro-motive force than our own ; but between 
1692 volts, which gave a length of spark equal to O'Oll inch, and 1854 volts, which 
gave a spark of 0'016 inch, there is a remarkably sudden rise in Sir William 
Thomson’s results, which is quite at variance with our experiments ; from that point 
(1854 volts), however, the two curves are sensibly parallel up to the limit of his obser- 
vations. 
It must be remarked that, notwithstanding the accordances in the distances 
between the terminals, when the spark jumped in our oft-repeated experiments, it 
seldom occurred actually at the point of nearest approach, although it usuallv did so 
in close proximity with it. The distances observed for successive increments in the 
number of cells are, it will he seen, considerably less with spherical surfaces (not so 
much as a fourth in many cases) for high tensions than with a point and a disc, and 
moreover the increase of increment with spherical surfaces does not conform to the 
ratio of the square of the number of cells, as before referred to in a communication 
to the Society, and published in the Proceedings, vol. xxiv. p. 167, 1876. 
The difference in our results with spherical surfaces of three inches radius of 
curvature and those of Sir William Thomson with very slightly curved surfaces, 
induced us to make other experiments, with some so nearly flat that at their peripheries 
(0‘687 inch from the centre) the distance was about 0’004 inch when the centres 
touched. The following three series of observations were made, viz. : — 
