The aura was somewhat larger in extent upon one electrode than on the other. In the first case, it sprang from a considerable number of minute illuminated points; on the other electrode, these illuminated points were fewer in number, and the flame was more purple in tint. Reversing the current these effects were reversed. The aura was uniformly thin and disk-like, and the curved edge remarkably true in shape (see Plate XVII. fig. 6). The lateral direction of the aura was changed when the current was reversed. [Sidenote: Aura not proportionate to length of spark.] 3. The aura was found not proportionate to the length of the spark. When the electrodes were approached, so as to very much shorten the spark, the aura still sprang out to a distance and extent quite out of proportion to the length of the spark. Even when the electrodes were approached so close that the spark was very short indeed, still, under the magnetic influence, a very considerable aura made its appearance. [Sidenote: Effect of working coil-break upon the aura.] 4. Upon working the coil-break, it was found that in proportion as the contact screw was drawn apart from the break, so the aura gradually diminished in extent, until at last, by continuing to increase the distance between the screw and the break, a point was reached when thin bright sparks, without any aura, passed. Upon the screw being worked up closer, thicker sparks passed, and the aura again made its appearance. As the aura diminished in size it gradually changed in tint from yellowish rose-pink to purple. [Sidenote: Spark taken in glass bulb.] 5. The spark was taken in a glass bulb, the tube in which it was blown being open at both ends, with the same effect as in the open air. 6. A plate of glass was laid on the poles of the magnet, and the spark was passed _along_ the poles (in the same direction as the heavy glass was laid in the Faraday experiment). No aura was formed. The points were then moved round, so as to carry the spark at right angles to the poles, and the aura was formed as before. [Sidenote: Aura could be blown away from the spark.] 7. The aura, it was found, could be blown away at right angles to the spark. When strongly urged, it assumed the shape of a flickering tongued curtain of flame, flying away in the contrary direction to that from which the current of air proceeded, and again returning to its original shape as the impulse was removed. The spark proper was not influenced (see Plate XVII. fig. 8). [Sidenote: Effect of withdrawing spark from central position between the poles.] 8. As the spark was withdrawn from its central position between the poles of the magnet, the convex edge of the aura became gradually less perfect, and assumed a ragged and broken-up appearance, the inequality at times amounting almost to jets or flickering sprays of light. The spark was also slightly curved away from the electrodes (see Plate XVII. fig. 7).