[ 339 ] 



XII. 



BANDED FLAME-SPECTRA OF METALS. By W. N. HAETLET, F.E.S., Honorary 

 Fellow of King's College, London ; Eoyal College of Science, Dublin ; and HUGH 

 EAMAGE, B.A., A.E.C.Sc.L, St. John's College, Cambridge. 



(Plates XXIX.-XXXIIL.) 



[Read Mat 22nd, 1901.] 



In ''Flame-Spectra at High Temperatures,"* it was proved by one of the authors 

 that metals easily volatilised exhibit banded spectra. 



Characteristic flame spectra of elements were described as follows t : — 



" Band Spectra — Antimony, bismuth, gold, tin, sulphur, selenium." 



"Band Spectra with Lines — Copper, iron, manganese, tellurium, lead, and 

 silver." 



In addition to these the " Spectra peculiar to Compounds," such as calcium 

 oxide, calcium fluoride, and magnesium oxide, were described as containing lines 

 and bands.J A summary of similar work previously published, giving a few 

 instances of bands in metallic spectra, was included in the paper and ajjpendix. 



It is stated on pp. 166— 7 that the band-spectra of silver and gold are really due 

 to the metals, "since no oxides of these metals can exist at the temperature of the 

 flame employed." Reasons are also given for attributing the bands in the flame- 

 spectra of manganese and of its compounds to the metal itself. The relation of 

 line-spectra to band-spectra is also discussed on j). 167. The two other papers, 

 entitled Parts II. and III. of "Flame-Spectra at High Temperatures," were 

 published also in the same volume. 



The continued investigation of this subject has been jointly prosecuted for the 

 last six years by the authors of this communication. Photographs of the spectra 

 have been taken with a four-prism spectrograph. § This instrument gives wider 

 dispersion and better definition than the one-prism instrument used in the earlier 

 part of the investigation. By this means, and by small but important improve- 

 ments in the method of working, finer detail has been obtained, new band-spectra 

 have been discovered, and some of the bands which had been previously photo- 

 graphed have now been resolved so as to clearly show their component lines. 



The spectrum of silver, which has already been completely described in the 

 first paper above-mentioned, is remarkable for the character of the banded spectrum 



* Phil. Trans., vol. 185 (A.), 1894, pp. 161-212. 



■f See p. 166. % See p. 168. § Phil. Trans., vol. 185 (A.), 1894, p. 1047. 



TBANS. EOT. DUB. SOC, K.S. VOL. VII., PAUT SII. 3 ^ 



