A. W. Wright — Forms of the Ekctrwal Discharge in Air. 441 



ent inclinations as the thread of air ls moved in one direction 

 or another, either by currents in the surrounding air, or, as 

 apparently happens, by the attraction of the charged parts of 

 the machine. When no disturbing cause affects it, the plane 

 will tend to take a vertical position, from the diminished den- 

 sity of the air-thread which is heated by the discbarge. In 

 any case, as it moves from its first position, and the repeated 

 •iischavges pass through it, it occupies positions suceessi\ely 

 tanher from the original line, until from its increased length the 

 r^'sistance becomes too great, and the discharge occurs again 

 no;ir the initial position, and so on. Owing to the persLstance 

 of the impression in the eye several lines are seen at once, appa- 

 rently simultaneous, but that they are not so may be readily 

 made evident by suddenly moving the eye while looking at 

 them, or by viewing them in a mirror which is quickly turned 

 about, when the intervals between them are seen to be drawn 

 out and enlarged, whereas these would be unchanged if the dis- 

 charges occurred at the same instant, or were continuous. 



The effect becomes much more striking when the poles 

 are so shaped as to increase the vertical length of the space 

 between them For this purpose an experiment was made by 

 attaching to the inner extremity of each electrode a brass tube 

 about SIX inches in length and half an inch in diameter, the 

 upper end being hemispherical, and the tubes fixed in a verti- 

 cal position. When the electrodes were pushed in so that the 

 tubes were from one-quarter to one-half of an inch apart, and 

 the latter were brought as nearly as possible to parallelism, 

 sparks passed in quick succession 'between them, and now and 

 then appeared to rise rapidly from the lower portion of the 

 ^^^^^'al toward the top. When the condensers were employed, 

 ^M the tubes were brought nearer together, the same effect was 

 observed, but indistinctly and less frequently, as the discharges 

 succeeded one another too slowly to make it readily apparent 



.A phenomenon of the same kind is often presented in light- 

 ning flashes. It is no uncommon thing during a violent thun- 

 <ler-storm, to see two, three or even more discharges of light- 

 ning follow the same path in quick succession, the inter^'al 

 betw^n^ *i-„_. , . , i^ , . % . J —-sufficient 



r heated 

 i considerable 

 5 required for it to become cooledf by the enveloping air, 

 -- .„ therefore retains its superior conductivity long enough to 



t«rmine a repetition of the discharge through it. 



^tie lormation of the electrical shadow discussed i 



™er paper, as has been suggested by Mr. C. F. Yarl 



J""gioJiow the same path in quick succession, the 

 t/fiT^j^ them being but the fraction of a second, yet ! 

 h 1 ^^,^*^°ctly perceptible. In this case the mass of a 

 y the discharge is comparatively great, so that a com 



nine IS rennirprl f^^ U +„ T,„„^™^ «^^lr>rl >KTT iha Pn\'p]n 



uggested by Mr. C. F. Yarley,* 



' -J'a««fe, vol. iu, p. 238, ia report of Proc. Roy. Soc. for Jan. 12, 1871. 

 • Jour. Sci -Third Skries, Vol. I, No. 6.^Junk, 1871. 



