Professor Leslie on Electrical Theories . 
air, a greater resistance must be overcome, and consequently 
the spark will be shorter, and the sound louder ; but if the air 
be rarefied, the effect will take place, though the balls are farther 
asunder ; and if, while they are in this situation, it be still more 
attenuated, the streams will play faster, and the light grow more 
vivid. A certain limit will, however, be attained ; for the quan- 
tity of air that serves to communicate the electrical virtue may 
be so much diminished, as to counteract any advantage from its 
rapid motion, which can never pass certain bounds ; the appear- 
ance will, therefore, become fainter and fainter, and, in a per- 
fect vacuum, the electrical light will totally vanish. If sparks be 
sent, about the distance of 18 inches, between two balls, in a 
glass cylinder, nearly exhausted, the streams of luminous air will 
seem to play in opposite directions, much attenuated, however, 
near the middle. 
The principles now investigated, will explain various other 
phenomena. Thus a point, projecting from a broad metallic 
plate, that is held at some distance from the prime conductor, 
will be entirely dark, if placed in the middle of the surface, but 
will become luminous, when it is removed near the margin. A 
sheet of tinfoil is bordered with light, but the rest of its sur- 
face, however rough or ragged, is totally obscure. Though a 
strip of half an inch, or even an inch, be cut out from the 
centre, the edges so formed will remain dark ; but a narrow 
slit made near the verge will be distinguished by its brilliancy. 
For, in these instances, the air is almost stagnant over the sur- 
face, and flows off, bending by the edges. 
But, whatever speculations we may form in regard to electri- 
cal light, and the mode in which the point and the knob pro- 
duce their different effects, we must admit that the electricity is 
never communicated, in any perceptible degree, to a remote and 
unconnected body, but by means of a current of air ; and this 
established principle will enable us to estimate the real effects of 
conductors or thunder-rods. 
When two portions of air, near the point of saturation, and 
of different temperatures, are mixed, a quantity of the dissolved 
vapour is precipitated, and resumes its aqueous state. By this 
conversion, the mass acquires electricity ; and the conse- 
quent repulsion exerted tends to disperse the minute globules 
of water, which will float in the atmosphere, or rather, will de~ 
