ELECTRIC DISCHARGE WITH THE CHLORIDE OF SILVER BATTERY. 515 
Postscript. 
(Added July 10th, 1883.) 
Our battery now consists of 15,000 cells, all of tlie form in which the chloride of 
silver is fused into rods on a flattened silver wire.* * * § On the occasion of a lecture given 
by one of us at the Royal Institution on January 21, 1881, 14,400 cells, partly of the 
rod form, partly of the chloride in powder form, were used; the length of the spark with 
this number between paraboloidal points was 0 *7 inch (17*8 millims.), and between a 
point and disc 0*62 inch (15*7 millims.). It will be recollected that in Part I. of our 
researches,! we obtained with 11,000 cells between two points a spark 0*62 inch long, 
and between a point and disc one 0*47 inch long. It does not appear, therefore, 
that the law of the spark being as the square of the number of cells holds good 
beyond a certain number ; the falling off may be in part due, however, to the failure 
of insulation as the potential is increased. 
14400 2 x 0-62 inch 
11000 2 
= 1*060 inch (27 millims.) 
14400 2 x 0*47 inch 
11000 2 
= 0*806 inch (20 millims. 
) 
Since the removal of the battery from the Royal Institution we have not charged 
up the whole of it, as the experiments on which we have been engaged did not 
necessitate more than 11,000 cells. 
We have recently (May 10-July 4) repeated the experiments on the question of 
the polarisation of the terminals on a discharge taking place in gases, and have 
obtained the same results as those already described;! they confirm, therefore, our 
previously expressed opinion,§ that the experiments point conclusively to the deduc¬ 
tion that the small current, which is observed when the terminals of a vacuum 
tube (in which a discharge has taken place) are disconnected from the battery and 
instantaneously connected with a galvanometer, is due solely to a static charge 
and not to chemical 'polarisation. 
* Phil. Trans., Part I., Vol. 169, 1878, p. 109, separate copy, p. 55. 
f Ibid., p. 118, separate copy, p. 64. 
f Proc. Roy. Soc., Ho. 205, 1880, pp. 563-572. 
§ Ibid., p. 570. 
