68 



Mr. J. Joly. 



discharge may "be chiefly confined to this surrounding dielectric y 

 spreading over the surface of the layer in ramifying lines. 



This ramifying discharge is depicted in some measure in Lichten- 

 berg's figures and in Antolik's. But these lose some of their value in 

 the fact that they represent an electrified condition of the surface of 

 a dielectric obtaining subsequent to the period of discharge. The 

 path of discharge revealed in the figures described in this paper is, on 

 the other hand, laid down by the outspreading current in the act of 

 discharge. There is some resemblance between all three patterns of 

 figures, more especially with Antolik's, whose figures are obtained 

 with a similar disposition of apparatus. 



The following are the arrangements used in obtaining the dust- 

 figures : — 



A plate of glass is coated evenly with a thin layer of lycopodium 

 powder. This is best done by placing the glass on the bottom of a 

 deep box and shaking the powder from a linen bag, surrounded with 

 a couple of folds of gauze, through a hole in the lid of the box. The 

 plate is next transferred to the surface of a smooth sheet of metal. 

 Wires from the + and — poles of a Ruhmkorff coil are then brought 

 down just to meet the surface of the glass, touching it at points 6 or 

 8 cm. removed from each other and symmetrical about the centre of 

 the plate. It is immaterial if the underlying conductor be connected 

 to earth or not. By drawing back the hammer of the coil and again 

 bringing it sharply forward, a single make and break is effected. At 

 the moment of " make," hardly any disturbance of the dust, save for 

 a couple of millimetres around the poles, is noticeable, but at the 

 moment of " break " the dust is suddenly agitated and thrown into 

 the pattern shown in Plate 1, a flash of burning lycopodium some- 

 times accompanying. 



TUis figure depicts the case where the poles have been brought into 

 such proximity as to permit, in addition to the ramifying discharge, 

 the direct passage of a spark. If the poles on the glass be so far 

 removed that no spark passes from pole to pole, the figures appear 

 each separate and complete, but in general branching towards one 

 another. If one pole of the coil be put in connexion with the metal 

 plate beneath the glass and the other be applied centrally to the glass, 

 a figure corresponding to the nature of the pole so applied to the 

 glass is produced. These figures tend to become more symmetrical 

 in development and rounded in outline according as care is taken to 

 centre the pole on the plate and the conductor beneath the plate, and 

 also when the plate used is large. Omitting the underlying conductor 

 diminishes the extent of the figures. 



(2.) Of these figures, the + pattern is moss-like and irregular on 

 the edge, the — pattern smooth and cloudy in outline. It is seen 

 that v.'ithin each of the figures the pattern corresponding to the 



