372 Bcientific Proceedings, Royal Duhtin Society. 



bands, and hydrogen lines, but no trace of any lines due to lead or any other 

 impurities, this proving that the tubes were quite clean and the water pure. 



Standard solutions containing one gramme of the metal per litre were 

 made of the chlorides of manganese, lead, copper, and lithium respectively. 

 The solutions of the substances were then introduced into the tubes, the 

 method being precisely the same for all. The stopper containing the electrode 

 was detached and O'l. c.c. of the standard solution introduced into the lower 

 bulb of the quartz tube from a clean burette. The other end of the quartz 

 tube with the attached stop-cock was joined to a Buusen pump, the tube 

 heated and a current of dried air drawn through, the solution in tlie tube 

 being kept as near as possible to the lower limb of the capillary portion, and 

 heated gently until all the water was driven off. The tube was then cooled 

 in a current of dried air. When cold, tlie tubes were disconnected from the 

 Bunsen pump, the second electrode attached, and the tube connected to the 

 Geryk pump, exhausted, and the current passed as with the distilled water, 

 having the capillary portion and the lower limb of the quartz heated to 

 volatilize the chlorides ; five minutes' exposure was given for each photograph, 

 and the spectrum of cadmium was superposed as before. 



The spectrum given by this small quantity (O'OOOl gramme, metal) 

 consisted of many lines, manganese especially giving a great number (Plate 

 XXIV., No. I). In all the photographs there were lines due to hydrogen 

 and water-vapour, and some showed bands due to nitrogen which had not 

 been expelled from the tube owing to the smallness of the quantity of metallic 

 vapour present. In the case of the new gases — helium, argon, neon, krypton, 

 and xenon — Collie and Ramsay have shown that the presence of hydrogen 

 greatly influences the spectrum obtained, as under certain conditions of 

 pressure the spectrum of tlie rare gas is entirely invisible and only that of 

 hydrogen shows ; but so far as I have observed their presence in no way 

 afi'ects the spectrum of metals ; and it is very difficult to eliminate the last 

 traces of hydrogen and water- vapour without the introduction of nitrogen. 



In order to avoid the presence of nitrogen and to take another photograph 

 with each tube, the lower limb was opened and a small glass rod, which had 

 been moistened with strong hydrochloric acid, was inserted, allowed to 

 touch the side of the tube, and then withdrawn. This was repeated with each 

 tube, which was then closed, and photographed as before. The small quantity 

 of hydrochloric acid vapourised when it was heated, and drove out any 

 nitrogen present ; it also combined with and helped to volatilize any of the 

 metal which might have been reduced during the experiment. Of course 

 most of the chloride had probably disappeared out of the tube ; but apparently 

 not all, as there still remained many lines in the photographs, due to the metal, 



