559 
o f Edinburgh, Session 1879 - 80 . 
The truth or falsity of this assumption appears to he capable of 
being established with ease by means of the method of these 
experiments. 
From the end of the second table it appears that the ratio tends 
to change from being less than unity to being greater than unity, 
when the hissing discharges begin to appear. Suppose the negative 
spark preceded by hissing discharges, but the positive not. Then 
the occurrence of these hissing discharges is apt to diminish the 
deflection at the time when the negative spark passes, while their 
absence in the case of the positive spark allows the full deflection 
to be observed. Thus the ratio of the readings may change to be 
greater than unity. In the case of the positive spark, the electro- 
motive forces for the four distances lie well on the curve which 
from previous experiments we found to be true for the discharge 
between a ball and a plate, but in the case of the negative spark, 
only those for the first two distances. 
The discharge from the point is a more complex phenomenon 
than the discharge from the ball; its explanation probably requires 
many further experiments. 
These results accord with those published by Drs De La Rue 
and Muller, in Part I. of their “Research.” They state (Phil. Trans, 
vol. clxix. p. 76) that with high tensions — 5000 to 8000 chloride of 
silver cells — the spark between a point and a disc is longer when 
the point is positive, but with low tensions up to 3000 cells, it is 
generally longer when the point is negative. For the discharge 
they observed is single, like that which we obtained between the ball 
and plate, at the smaller distances, not intermittent like that which 
we obtained between the point and plate. 
I may mention that the observations of April 10, 1880, are 
supported by a previous series of observations, taken by means of 
a more roughly divided scale. 
The inductric and indue teous balls (to employ terms invented 
by Faraday), by which the measurement was effected, were at a 
distance from one another of 15 inches;, and the length of the 
wire connecting the inducteous ball with the electrometer was 
about 10 feet. 
In the record of observations the entries are given in the order 
in which they were observed. 
