160 ^ Professsoi^s Liveing and Dewar, On the [March 25, 



March 25, 1878. 

 Professor Liveing, President, in the chair. 



A communication was made to tlie Society by 



Professors Liveing and Dewar, " On the reversal of the lines of 

 Metallic Vapours." 



The apparatus employed by them is represented in the ac- 

 companying diagram. D is an iron tube half an inch wide and 

 2 feet long, the lower end closed, and coated with borax or fire- 

 clay. This was placed vertically in a small assaying furnace, E 

 (shown in section) 9 inches square and about 12 deep with a good 

 draft, and fed with Welsh coal so that a welding heat was attain- 

 able in it. A gentle current of hydrogen from the Kipp's ap- 

 paratus A, purified and dried by passing through the tubes B and 

 C, was led into D by a narrow brass tube reaching to the hot 

 part of D. i'^ is a small mirror by which the light from the 

 interior of the hot tube was reflected on to the slit of the spectro- 

 scope G. In the earlier experiments the mirror was omitted, and 

 the spectroscope held above the tube. Fragments of the metals 

 examined were dropped into the tube D, and the consequent 

 absorption of the light issuing from the hot bottom of the tube 

 observed with the spectroscope. The most characteristic lines of 

 thallium, indium and magnesium were thus seen directly reversed. 

 The red line of lithium was only seen reversed when potassium, 

 sodium and lithium chloride were all. introduced into the tube 

 together. 



In the case of sodium besides the D lines they observed a 

 dark line in the green with a wave length about 5510, and in the 

 case of potassium a dark line with a wave length about 5730. 

 These lines do not correspond with any known emission lines, 

 though they are in each case near to, but more refrangible than, 

 well-known emission lines of those elements. The channelled 

 spectra described by Roscoe and Schuster {Proc. R. S. v. 22) were 

 also seen sometimes, but the above-mentioned lines were seen at 

 times when the channelled spectra were not visible. Besides ob- 

 serving potassium and sodium in the iron tubes, they examined the 

 absorption produced by those metals volatilized in glass tubes about 

 S inches long, of the form H in the accompanying diagram, looking 

 through the length of the tube at a lime light. The ends of these 

 tubes were drawn out so as to present approximately plane faces 

 at the ends, one end being drawn out into a narrow tube by which 



