PoLLOK — The Vacuum Tube Spectra of some Metallic Vapours. 207 



5. That witli the chlorides you get precisely the same lines as you do with 

 the vapour of the metal, with or without the addition of the lines of chlorine, 

 according to circumstances. 



6. That the chlorine lines, as a rule, do not show unless a Leyden jar is 

 introduced in the circuit, and even then they do not show with equal facility 

 with all compounds. 



7. That in all cases where a line is enhanced or otherwise by the intro- 

 duction of a condenser in the circuit with the vapour of the metal, a precisely 

 similar change takes place with the vapour of the chloride. 



8. Lines that show in the arc spectrum of an element, but not in the 

 spark spectrum, are not seen in the vacuum tube spectra of the element, 

 either with or without a condenser. 



9. That in many of the spectra, especially when much vapour is present, 

 and no Leyden jar is used, very beautiful bands develop, some of which are 

 entirely new to science, and these will require further study before any general 

 conclusion can be drawn regarding them. They are quite independent of 

 the lines of the spark spectra of the elements, and probably owe their origin 

 to the molecules of the compounds under observation. 



In the tables, as usual, tlie strongest lines are marked 10, the numbers 

 decreasing to 1 for lines that are just visible ; n indicates a nebulous line, 

 s a sharp line, c a continuous line, d a discontinuous line, r a line frequently 

 reversed, b a band measured at the centre and shading off on each side, b" a 

 band measured on the ultra-violet edge and degrading towards the red, i'' a 

 band measured on the red edge and degrading towards the ultra-violet ; 

 u>, ip, Y, (j> indicate the quantitative persistency of the lines in the spark spectra 

 of solutions of the clilorides, as used in the author's papers on Quantitative 

 Spectra. 



ID seen with '001 per cent, or more in solution. 



4i „ '010 „ „ „ 



X )' '100 „ „ „ 



</> „ 1-000 



Cadmium. 



Both the metal and its chloride give a very brilliant luminescence in the 

 tube, and the introduction of the condenser makes a remarkable change in 

 the spectrum of the vapour of the metal, but little difference with that of the 

 chloride, the spectrum of chlorine not being developed. 



The behaviour of the lines X2748'7 and A.2573-1 and the characteristip 



