Observations on the Spark Discharge. 



71 



form had a mossy outline, with but little branching and few and faint 

 spark veins. Then a negative cloud-edged pattern, with finally a 

 second mossy development of the + form at the centre. At 10" 

 pressure the — pattern was a very regular halo, with uniform texture. 

 No spark veins, and central mossy + form. The + pattern was a 

 mossy edging, with a second mossy pattern at the centre and an 

 intermediate undisturbed region, in which no definite marking could 

 be detected. At 6" pressure the halo of the — form is more extended, 

 very faint, and shows the moss pattern at its centre. No spark veins. 

 The + form consists of a few coarse, straggling lines, extending 

 towards the centre, a region within of unmarked powder, and a few 

 more thick straggling lines wandering from the centre. Ail these 

 thick lines show a central core of unmoved powder, each, in fact, con- 

 sisting of two parallel lines in which the powder has been removed, 

 leaving an undisturbed central axis. No spark veins. 



In air at 3" pressure both forms have become very indistinct. The 

 suggestion of a halo in the negative : a little pitting here and there, 

 not deep enough to expose the glass, in the positive. 



It appears from these experiments that the nature of the gaseous 

 dielectric exerts a considerable influence on the nature of the path 

 taken by the outspreading current. It would seem to be also a ques- 

 tion as to the degree of conductivity possessed by the gas. The 

 ring-like symmetry of the negative pattern, as well as the generally 

 more symmetrical outspread of the positive, and the absence of spark 

 veins in both, seem to indicate a uniformity of spread of the current 

 in the better conducting media, as hydrogen, or air, at reduced pres- 

 sure. This suggests, in fact, that these dust-figures owe their forms 

 chiefly to the manner in which the current spreads in the surrounding 

 gas. It has already been seen that the nature of the powder in general 

 exerts little qualitative influence. The nature of the plate carrying 

 the powder has yet to be dealt with. 



(6.) Formed on the surface of a sheet of vulcanite, the figures 

 exhibited no distinguishing feature from those on glass. Thus it 

 appears that a difference of specific inductive capacity (vulcanite is 

 two-thirds that of glass, according to Gordon) does not appear to 

 affect what influence the non-conducting plate may exert on the form 

 of the figure. It will now appear, however, that the isotropic quality 

 or otherwise of the plate is an important factor in determining the 

 path taken by the current. 



A plate of selenite cleaved on the clinodiagonal, measuring about 

 6x7 cm., was polished smooth on opposite faces, having been re- 

 duced to a thickness of about 4 mm. Owing to perforations, due to 

 loose crystallisation, the crystal had to be laid down on glass with 

 melted paraffin. Dust-figures taken on this crystalline surface showed 

 a very marked variation from those effected on glass. This is espe- 



