489 
Probably the form of the wave of difference of potential 
between the terminals of the induction-coil has much to do 
with the ratio of increase. 
With a certain length of spark and a certain capacity of 
condenser, a luminous brush is perceived at the positive 
terminal, and a glowing point at the negative terminal ; 
as the capacity is increased or diminished, more or less of 
a luminous brush appears also at the negative terminal. 
This probably means that more or less of positive electrons 
are discharged from the negative pole, which will complicate 
the phenomena under observation. 
Hxperiments were tried to show the effect of altering the 
capacity of the condenser :— 
an Inducteon-Coil by the Aid of Auailiary Wires. 
Fraction of whole | Spark-length. Ratio | 
capacity used. , Unassisted. Negative assisted. ae | 
9 12 14 mm 1:16 
9 2mm mm. 
7 WO fea 
9 12 mm. 15 mm. E25 
3 r 29 
9 15 mm, 22 mm. 1:46 
if at, Re. 
9 15 mm. 24 mm. 1°60 
9 5 B es 
9 82 mm. 107 mm. io 
8 a = 
9 91 mm. 116 mm. 1°27 
i 88 mm. 116 mm. 1:32 
| 
Thus the effect of the variation of the condenser capacity 
varies considerably with the length of the spark. 
It was found that with a spark-length of 60 mm. or more, 
a capacity of % of the full gave the best results. 
Experiments were made to try and discover the causes of 
these phenomena ; and first with regard to the action of the 
positive assister. In these experiments it must be premised 
that the error in the measurement of the spark-length may 
be as much as 1 or 2 per cent., owing to slight irregularity 
in the vibrator. Also the effects varied somewhat on different 
days, apparently owing to changes in the temperature and 
humidity. ) 
To determine whether the assisting effect was due to some 
form of radiation or ionization, the assister was so placed that 
