MoEKOW — On the Influence of Self-induction, ^c 609 



convenient for comparing tbe effects produced by tl>e different electrodes, 

 and for observing by inspection which lines were due to the gaseous 

 element. 



In order to obtain the spectra of the non-metallic elements which are 

 liquid or solid at ordinary temperature, a small quartz flask was used as 

 shown in fig. 2, Plate XLYIII. A wide glass tube passed through a rubber 

 cork in the mouth of the flask, and through it the platinum connexions 

 entered insulated by glass capillary tubing, and these were passed through 

 the wide tube in the rubber cork and held in position by a plug of cotton 

 wool or asbestos. As before, a photograph was taken of the electrodes in 

 air. Then the substance whose spectrum was required was put into the 

 flask, and vapourized by heating it with a Bunsen flame, taking care that 

 nothing condensed on the side of the quartz bulb which might interfere 

 with the image of the spark on the slit of the spectrograph. When the 

 flask was filled with vapour, the spark was passed, and photographs taken 

 with and without the self-induction coil. Gold and carbon electrodes were 

 used in each case, and six photographs obtained on one plate as before. 



The lines shown by each element examined are tabulated at the end 

 of the paper, and the effect of introducing self-induction is indicated by the 

 change of intensity of the line. The strongest lines are marked ' 10/ and the 

 scale of intensities diminishes to ' 1.' If a line has been looked for and not 

 found, it is marked '0.' Nebulous lines are marked 'n.'; discontinuous 

 lines ' d.' ; sharp lines ' s.' The more important photographs are reproduced 

 in the plates. 



In all I examined the spectra given by sparking both gold and carbon 



electrodes, with and without self-induction, in nitrogen, oxygen, hydrogen, 



chlorine, bromine, iodine, sulphur, phosphorus, boron trichloride, silicon 



tetrachloride, sulphur dioxide, sulphuretted hydrogen, carbon dioxide, 



carbon monoxide, and hydrochloric acid. Au examination of the spectra 



showed that when electrodes are sparked in au atmosphere of any kind 



the principal lines of the line spectra of the elements present in the 



atmosphere are obtained, and as a general rule with greatest intensity 



when self-induction is not introduced. In the case of compounds, only the 



line spectra of the component elements are seen, and the band spectra 



of the compounds are not seen except in the case of cyanogen, as is 



already known. In the case of hydrogen, the only electropositive gas 



examined, the effect of self-induction is to intensify and sharpen the 



hydrogen lines, and to remove, or almost entirely remove, the lines of 



gold or carbon, whereas, in the case of electro-negative gases such as 



oxygen and nitrogen, the effect of self-induction is to remove or almost 



4x2 



