496 Mr. Bowlker on the Lengthening of the Spark from 
mica be interposed between this spark-gap and the assisted 
positive terminal the sparking distance is reduced to its 
original value. 
Thus in an experiment where X and Y are the positive 
terminals and N the negative terminal (see fig. 3), the 
Fig. 5. 
unassisted sparking distance between X and N was 44 mm., 
but with Y assisted this became 65 mm. ; with a mica screen 
interposed between Y and XN it fell again to 44 mm. ; 
showing that the lengthening of the spark-gap XN was due 
to the electrons proceeding from the assisted point Y. In 
the figure, M shows the position of the mica screen. 
The above experiments lead me to suggest the following 
explanation of the phenomenon under observation. 
The action of the negative assister in lengthening the 
spark may be due to the attraction of the negative electrons 
in the assister and discharged from the assister on the 
positive electrons. 
When the negative assister is impulsively charged across 
its spark-gap this causes an induced impulsive rush in the 
positive wire, causing the positive electrons to leave the 
terminal with greater velocity. The effect of the negative 
assister is probably added to by a discharge of negative 
electrons from its point when it is impulsively charged. 
This negative discharge from the point is in an effective 
position to act attractively upon the positive electrons, and 
endures long enough to produce its full effect. When the 
negative assister is surrounded by an insulator or when a 
