THE 

 LONDON, EDINBURGH and DUBLIN 



PHILOSOPHICAL MAGAZINE 



AND 



JOURNAL OF SCIENCE. 



[FOURTH SERIES.] 



NOVEMBER 1861. 



XLI1. Chemical Analysis by Spectrum-observations. — -Second 



Memoir. By G. Kirchhopf and R. Bunsen*. 



[With a Plate.] 



IN our first memoir on this subject, a translation of which ap- 

 peared at page 89 of the 20th volume of this Magazine, we 

 showed that the bright lines observed in the spectra of the incan- 

 descent vapour of certain metallic compounds may be employed as 

 the most sure and delicate tests for ascertaining the presence of 

 these metals. The analytical method founded upon such obser- 

 vations is of special importance when applied to the examination 

 of groups of substances either occurring in very small quantities 

 or possessing nearly identical chemical characters, because in 

 these special cases this mode of examination introduces a whole 

 series of most delicate distinctive reactions, hitherto wholly un- 

 known. Considering how much more delicate the spectrum 

 reactions are than the ordinary chemical tests, it appeared to us 

 that this method would serve especially well for discovering the 

 presence of substances which might have been overlooked by the 

 rough methods previously employed, either because the bodies 

 occurred in very small quantities, or because they were not 

 distinguishable by the ordinary tests from other well-known 

 bodies. This assumption was verified by the first appeal to ex- 

 periment ; for we have succeeded in discovering the presence of 

 two new alkaline metals in addition to potassium, sodium, and 

 lithium, notwithstanding that they both give all the character- 

 istic precipitates of the potash salts, and occur in such minute 

 quantities that, in order to obtain a few grammes of these sub- 

 stances, or a sufficient amount for investigation, we had to ope- 

 rate upon 41,000 kilogrammes (about 40 tons) of the mineral 

 water of Durkheim, and upon 180 kilogs. of Lepidolite. 



* From Poggendorff's Annalen, No. 7, 1861. Communicated by Pro- 

 fessor Roscoe. 



Phil. Mag. S. 4. Vol. 2.2. No. 148. Nov, 1861. Z 



