This section is from "Scientific American Supplement Volumes 275, 286, 288, 299, 303, 312, 315, 324, 344 and 358". Also available from Amazon: Scientific American Reference Book.
The experiments of the author have been principally directed to the alterations in shape and color produced in a flame when under the influence of positive or negative electricity. The flames were arranged so as to form one electrode of a frictional machine. When charged with positive electricity the flame became more blue, narrower, and pointed at the top, while little or nothing of the result was observed in negative flames.
A peculiar result is that the end of a negative flame returns to its own conductor, and that, according to the intensity of the electricity, and also depending on the width of the burner, this turning back of the flame is either intermittent or constant. Most noticeable are these results:
When the flame rises from a circular burner, or when burning round a metallic cylinder, in the latter case it returns to the metallic surface according to the intensity of electricity in an arc or angle, while the point of the flame divides into two branches, which separately perform more or less equal movements. If a body connected to the earth by a conducting wire is held opposite the flame at some distance, the flame will in all cases bend toward it; as the body is brought closer, the flame, if negative, will be repulsed, and, if positive, will be attracted, at least the upper luminous part of the flame, while the lower dark body of flame is also repulsed.
This phenomenon explains why a positive flame will burn through wire gauze, while a negative flame remains below the gauze. The positive flame becoming pointed explains the fact that this will drive a small fan wheel, while a negative flame will only just move it.
All these results are most prominently obtained with a pure gas flame, a stearine, wax, or tallow candle, very indifferently with a spirit flame, and least from a Bunsen flame rich in oxygen. They may not only be obtained with flames electrified direct, but also when placed under the influence of a long "Holtz" machine.
A flame placed between two small disks on the machine bends toward the negative pole, becomes widened, and, at a certain point of electric intensity, commences to vibrate and oscillate, exhibiting a peculiar stratification. Since these phenomena are also least observed in flames rich in oxygen, it appears to be a general law that carbon and hydrogen are more strongly attracted by the negative pole, while oxygen is more attracted by the positive pole, probably like in all polar differentially attractions, in consequence of a peculiar unipolar conductivity of the substances.
The return motion of the flame the author explains thus: The point of the flame loses more electricity by influence than it receives by conductivity. A paper strip fixed at one end to a large ball shows similar movements when its free end is pointed and made conductive. Why principally the negative flame returns may be explained in two ways--either the point of the flame loses much by radiation, or the base of the flame is a bad conductor. The former explanation would agree with the experiments made by Wiedemann and Ruhlmann, the latter with Erdman's theory of unipolar conductivity of flames. This theory is still further supported by the resistance on the negative electrodes noticed by Hittorf, which almost explains Erdman's experiments, because if negative electricity enters a flame with greater difficulty, then positive electricity must leave a flame with difficulty.--W. Holtz, in Wiedemanris Beiblätter to Poggendorfs Annalen.