70 



Mr. J. Joly. 



very numerous. All the other powders behaved like lycopodium, but 

 showed less perfectly developed figures. 



(4.) When a plate has been exposed to the action of the spark in 

 producing these dust-figures, the powder which before was loose on the 

 surface of the plate will be fonud to have become fixed, or to a con- 

 siderable extent adherent, to the plafce where the figure has been 

 formed, and continues to remain so for many weeks ; so far as I have 

 observed, indeed, indefinitely. 



If a plate bearing a dnst-fignre be laid by and subsequently (a 

 couple of weeks later) be dusted clean and then breathed upon, breath- 

 figures, differing in the finer detail from those formed on the powder, 

 appear. They more nearly resemble the curious photographic figures 

 obtained recently by Mr. Brown, by passing coil sparks over photo- 

 graphic dry plates ('Philosophical Magazine,' December, 1888). 

 Brisk rubbing, or washing with soap and water, destroys the " magic " 

 qualities of these breath-figure plates. I have not obtained these 

 breath-figures at all so distinctly developed on plates which had very 

 recently been sparked over. It would appear that a certain lapse of 

 time is necessary to confer this quality on the plate. Breath-figures 

 of somewhat similar character have been noticed before. 



(5.) Formed in an atmosphere of coal gas, the patterns show a 

 marked variation. 



In the negative a very regular, halo-like ring surrounds the pole, 

 through which the characteristic cloudy fronds of the negative pattern 

 break out, as it were, in places, extending further on the plate. The 

 -f- pattern appears in the centre of the halo. The development of 

 spark veins is less conspicuous. 



In the positive there is less variation from the normal pattern in 

 air. There are fewer spark veins, and hence less branching. The 

 characters are more those of an irregular outline, with deep mossy 

 edging. 



Formed in hydrogen, the negative is reduced to a faint, circular 

 halo, with a very faint positive pattern within. No spark veins 

 observable. The positive pattern shows only a few thin, sharp- 

 branching spark veins, fragmentary and radiate to the pole, with an 

 indefinite aggregation of the powder around them. 



Formed in carton dioxide, there is no notable change from the figures 

 formed in air. The distinction between the two patterns, the + and 

 — , is perhaps better developed, there being some increased likeness 

 with the Lichtenberg- figures. 



When the figures are developed under the receiver of an air-pump 

 at diminished air pressure it is found that at a pressure of 15". of 

 mercury the negative form appeared very much as in coal gas, i.e., 

 with a regular halo, having straight radiate marking, few and faint 

 spark veins and a mossy positive pattern near its centre. The positive 



