608 Scientific Proceedings, Royal Duhlin Society. 



was carried across by the vapourized metal. Schenk' investigated the 

 appearance of the spark by a rapidly rotating mirror. The first discharge 

 showed a brilliant white line on the mirror followed by a few weaker ones. 

 The lines, which were seen to be curved, Schenk showed to be due to metallic 

 vapour and the first brilliant white flash was proved to be due to the air. 

 This white line did not show in the mirror when the self-induction coil was 

 in the circuit. 



The apparatus used in the present investigation consisted of a spectrograph, 

 designed by Sir Walter N. Hartley,^ an App's coil, condenser, and a 

 Hemsalech self-induction coil, all of which are described in my paper on 

 " The ultimate Hnea of the vacuum-tube spectra of manganese, lead, copper, 

 and lithium."^ In order to obtain the spectra of the gaseous non-metallio 

 elements, a small quartz tube was used (fig. 1, Plate XLYIII) fitted with 

 rubber corks through which passed glass tubes enclosing platinum connexions, 

 sealed into the glass with blue enamel in the usual manner, and to these 

 were attached the electrodes, the spark-gap being usually about a quarter of 

 an inch. The electrodes used were either of gold or carbon, the carbon 

 electrodes being cut from a piece of Acheson graphite. The electrodes 

 were first sparked in air with the condenser and Hemsalech self-induction 

 coil in circuit, and the spectrum so obtained photographed. Then the 

 gas whose spectrum was required was passed through the tube, having 

 been thoroughly purified and dried, and a photograph of the spectrum 

 taken with the condenser and Hemsalech coil in circuit. The current of gas 

 was passed continuously through the tube, so that the products which might 

 be formed by combination with the electrodes would be swept away. 

 Another photograph was then taken of the spectrum of the gas without 

 self-induction. Thus there were obtained three photographs with each pair 

 of electrodes, i.e. — 



(a) Gold electrodes in air with self-induction, 

 (J) „ „ in the gas with self-induction, 



(c) „ „ in the gas without self-induction. 



(«) Carbon electrodes in air with self-induction, 

 {h) „ „ in the gas with self-induction, 



(c) „ „ in the gas without self-induction ; 



so that on each photographic plate there were six photographs, three with 

 each pair of electrodes. This arrangement of the photographs was very 



' Astrophys. Journ., 1901, vol. xiv., p. 116. 



2 Sclent. Trans. Eoy. Dub. Soc, 1882, vol. i., pp. 231-238. See also Scient. Proc. Roy. Dub. 

 Soc, vol. iii., 1881, p. 93. 



3 Scient. Proc. Eoy. Dub. Soc, 1912, vol. xiii., pp. 269-287. 



