12 Trans. Acad. Sci. of St. Louis. 
charge lines such as are shown in Plate IV. It is with 
difficulty that the negative discharge is forced out of a 
straight wire and over the film, although the wire is sur- 
rounded by a luminous glow extending outward a centi- 
meter or more from the wire. There are rudimentary 
lines shown in this black belt. The negative particles can 
be much more easily forced from one positive ion to the 
next within the body of the copper conductor, than into a 
photographic film in contact with the wire. This seems 
to be true notwithstanding the fact that the negative par- 
ticles are easily drawn from the film surface into the wire 
through several centimeters of distance when the copper 
wire is in an ionized condition in the positive circuit. 
This inward flow must cross the discontinuity at the con- 
tact of film and wire. It is evidently not the resistance 
offered by this discontinuity, which accounts for the dif- 
ficulty of forcing the electricity out of the wire on the 
compression side of the machine. The discharge lines 
shown in Plate IV begin at the wire and increase in 
length during the discharge. This may easily be shown 
by drawing pencil lines on the film. Some of these dis- 
charge lines will terminate on these pencil marks, while 
others not thus arrested will continue to lengthen to a 
greater distance from the wire. Such a result, with a 
different form of discharge wire, is shown in Fig. B, 
Plate X. It might be supposed from this result in Plate 
IV that positive discharges can be forced outward over 
the film from the positive wire, while the negative dis- 
charge cannot easily be forced to leave the wire, under 
like conditions. It might be supposed that the electricity 
in the positive wire is therefore under greater compres- 
sion. 
If these discharge lines from the positive wire are due 
to a positive discharge outward from the wire, then the 
placing of the grounded wire below the center of the plate 
should permit a negative discharge to flow upwards, and 
to distribute itself over the under side of the glass plate. 
This should produce a condenser effect similar to that 
