Observations on the Spark Discharge. 



73 



•covers a cloud on the negative. The branching spark veins, too, cor- 

 respond or overlie closely. Sometimes, however, the coincidence of 

 these latter is not perfect, for, if a plate be dusted and breathed on 

 upon both sides, the spark veins, then showing out more clearly, are 

 seen to diverge a little in some cases. 



That this coincidence of development at each side of the plate is 

 inductive and not ascribable to luminous action, possibly initially 

 developed on one side and then determining by its influence dis- 

 charge along certain paths on the other, is shown by repeating the 

 experiment on red photographic glass, when the coincidence is as 

 striking as before. 



It is hard to make out the exact nature of the coincidence. It 

 does not appear to be that of photographic positive and negative 

 throughout. However, the clear margin around the clouds on the — 

 pattern is invariably backed by the marginal frill of the + pattern. 

 It is interesting and curious to observe in the dark this inductive 

 transmission of the current across a sheet of clean glass, arranged 

 as in the above experiment. The close but not perfect coincidence of 

 the spark veins is then easily noticeable. 



(9.) If a bundle of plates making up about a centimetre in thick- 

 ness be arranged as in the last experiment (i.e., the extreme surfaces 

 powdered and touched centrally by the poles), the figures no longer 

 show the coincidence observed in the experiment with the thin di- 

 electric, but are smaller and more of the Lichtenberg type. That is, 

 the + tends to be more straggling and tufted, the — more rounded 

 and lobed. The inductive action is here feebler, and matters are 

 more as in Lichtenberg's experiment. 



(10.) On the other hand, diminishing the thickness of the dielectric 

 gives rise to figures of great minuteness and detail. With very thin 

 dielectrics it is difficult any longer to distinguish the + pattern from 

 the — . If such thin dielectrics be laid down on a metallic surface as 

 in the first-described experiment (1), the extreme delicacy of form is 

 still obtained. In this way the figures of Plate 2 were formed upon 

 a sheet of the thin glass used for microscopic cover glasses. Still 

 greater delicacy may be obtained by using thin sheets of mica, but 

 ultimately the piercing of the plate by the spark sets a limit to the 

 experiments. In these experiments the conditions are those for a very 

 strong inductive action. The double nature of each figure is apparent 

 in these as well as in the former figures, and the similarity, or perhaps 

 identity, of the forms of opposite sign is very well seen by comparing 

 the inner with the outer of these patterns. 



(11.) Figures taken on opposite sides of a thin insulator at 

 diminished air pressure present some peculiarities. A thin plate of 

 micro-cover glass was arranged to stand vertically beneath the re- 

 ceiver of an air-pump, the poles touching it centrally at each side. 



