(3) The tube was well excited, and the four bulbs (other than the negative one), upon stopping the current, glowed strongly. The current was then sent through reversed, so as to throw the negative glow for a few seconds into the positive bulb. The after-glow in the positive bulb was at once extinguished. On once more reversing the current, it was only restored after a certain amount of continuance of the positive stream. [Illustration: Plate XVIII.] [Sidenote: Effect of change of current on the three central bulbs.] The time during which the negative glow was thrown into the positive bulb did not appear sufficient to have heated it. After rapidly changing the direction of the current several times and then stopping it, the three central bulbs alone had an after-glow, the two extreme ones having none, being both equally transparent. [Sidenote: Effect of heat on the bulbs. Effect of cooling by ether-spray.] (4) A moderate heat from a spirit-lamp was applied to the centre bulb (_a_) while the current was on; and also (_b_) when this was stopped, and the bulb glowed. In the first case the bulb was found to get more transparent; and in the second case the after-glow disappeared in a proportionately shorter time in the heated bulb than in the others. To test the result of cooling the bulbs, the negative-pole bulb and also the central one were each subjected to the action of ether-spray, and also of ether and water-spray mixed. This was done, (_a_) when the current was passing, and (_b_) when it was stopped and the glow only was in the bulb. The bulbs were cooled until a marked cold effect to the touch was produced. We did not notice any difference in the behaviour of the bulbs so treated as compared with the others, either when the current was passing or in the case of the after-glow. [Sidenote: Negative-pole bulb between the armatures of magnet. Effects on negative and positive glow.] (5) We placed the negative-pole bulb between the conical points of the armatures, and excited the magnet. The negative glow contracted itself into a condensed violet-tinted crescent, in accord with the magnetic curves. The positive glow of the same bulb lost its beaded (stratified) character, and was condensed into a bright stream of light, which latter protruded from the small inner tube and formed a spreading spiral set of cloud-rings within the bulb (see Plate XVIII. fig. 2). The action of the magnet seemed to be exercised in subduing the stratification, condensing the glow into a bright stream of light, and forcing the latter to “tail over” at each extremity of the tubular joints into the bulbs—this effect extending even so far as the second bulb. When the positive bulb was placed between the poles of the magnet, the glow was simply condensed into a bright stratified stream, which flew to either side of the bulb. [Sidenote: Effect of magnet on glow in bulb No. 4.]