As a source of electricity, use is made of a bichromate of potash battery of 6 elements, capable of giving 10 volts and 15 amperes.  The current from this battery is converted into a current of high tension by means of a strong induction coil capable of giving sparks more than eight inches in length.  The discharge shown in Fig. 4 was obtained by means of a Holtz machine.  Each experiment lasted less than a second.

[Illustration:  FIG. 3.]

Figs. 2 and 3 represent the efflux that occurred under; the following conditions:  The disk, P, was of metal, and was connected with the negative pole of the induction coil; and upon it was laid the photographic plate with the sensitized film downward, and consequently touching the disk.  This is what produced the opaque circle in the center.  Then the photographic plate was entirely covered with a thin ebonite plate, above which there was a second one supported by small wedges, so as to allow air to circulate between them.  Finally, upon this second ebonite plate there was placed another photographic plate, with its sensitized film upward and directly in contact with an upper metallic disk, and connected with the positive pole of the coil by the conductor, L. An inspection of Figs. 2 and 3 shows that the, efflux does not possess the same form at the two poles.  We remark at the positive pole a quite wide opaque circle surrounded by a sort of aureola composed of an infinite number of very delicate rays, while at the negative pole the aureola seems not to have been able to spread.  We see, moreover, the same phenomenon in examining Fig. 4 (which represents the efflux obtained by means of a Holtz machine), but this time in a horizontal direction.  The photographic plate was here placed upon the non-conducting disk, P. As the sensitized film was upward, it was put in contact with the balls at the extremity of the conductors, H and N.

[Illustration:  FIG. 4.]

It will be seen here again that the efflux spreads out widely at the positive pole, while it is contracted at the other.  The conducting balls were spaced 0.04 inch apart.  A spark leaped from one to the other at the moment the current was being interrupted.

In Fig. 5 we are enabled to study with more ease a spark obtained with nearly the same arrangement.  The balls, H and N, did not here rest directly upon the sensitized film, but upon two small sheets of tin cemented to the extremities of the plate at 0.06 inch apart.  In addition, the source employed was not the Holtz machine, but the pile with induction coil.  Two nearly parallel sparks were obtained.  It will be seen that these are very complex.  Each of them seems to be formed of four lines of different sizes, entangled with one another and presenting different sinuosities.  Aside from this, the plate is traversed for a space of 0.04 of an inch by curved lines running from one pole to the other, and exhibiting numerous sinuosities.

[Illustration:  FIG. 5.]