[ 363 ] 



XXXVII. 



ON THE CONSTITUTION OF ELECTEIC SPAEKS. 

 By W. N. HAETLEY, F.E.S., Eoyal College of Science, Dublin. 



[Read March 20, 1889.] 



In the course of my researches on the ultra-violet spark spectra 

 of the elements, a number of isolated observations have from time 

 to time been made, and it is these which I propose to lay before 

 the Society. 



An induction coil and a condenser, consisting of a pane of 

 glass coated on two sides with tin-foil, in place of a Leyden jar, 

 is used for the purpose of obtaining brilliant sparks at the highest 

 possible temperature. 



Introduction. — Numerous memoirs on the subject of the electric 

 spark have been published, and the following may be quoted as 

 dealing with matters treated of in this communication. Masson's 

 "Etudes de Photometrie Electrique" {Ann. de Chimie et de 

 Physique, 3, xlv., pp. 385-454, 1855) describes the use of the 

 Ruhmkorff coil, and its application to the production of nietallic 

 spectra. These are considered to be caused by the incandescence 

 of vapour derived from particles of metal torn 'off the electrodes. 

 As early as 1853 this author employed a condenser with the coil, 

 for the purpose of increasing the brilliancy, of the rays. The action 

 of the spark upon gases was studied. One of the most important 

 conclusions arrived at was the co-existence of two opposite electric 

 currents in the same circuit. 



Perrot minutely studied the decomposition of water by the 

 electric spark, and found that two actions occurred side by side : 

 the spark decomposed the vapour of water by reason of its intense 

 heat, and thu.s yielded the bulk of the mixed gases obtained ; but 

 there was also a slight voltaic action. He further caused oxygen 

 and nitrogen to combine by means of the spark, and two well- 

 known lecture experiments illustrating these points are adaptations 

 of his demonstration — " Becherches sur Taction chimique, de 

 etincelle d'induction de l'appareil Ruhmkorff" {Ann. de Chim. et 



