1 76 Prof. Mitscherlich on the Spectra of Compounds 



the spectrum of the metal is obtained almost quite free from 

 that of the oxide ; it is only obtained perfectly so by method 7 

 (solid electrodes), using iron and iron compounds. 



Chloride of manganese, investigated by method 6 (volatilization 

 in the glass tube), gives the spectrum of the oxygen compound 

 pure ; it is obtained much more beautifully simultaneously with 

 the spectrum of the metal by method 2 (H andO). By the 

 beauty of its colours and the sharpness of its lines, this spectrum 

 is about the most beautiful of all. By this method I could only 

 detect one line which belonged to the metal, probably because 

 the bright parts of the spectrum of the oxygen compounds 

 obscure that of the metal ; investigated by method 8 (liquid 

 electrodes), the manganese compounds give moreover a few 

 bright and several feebly luminous lines, of which the bright 

 are depicted. 



The spectrum of an oxide of tin, obtained by method 2 (H, or 

 coal-gas, and 0), using oxide of tin, and that of an oxygen or 

 chlorine compound of chromium, using chloride of chromium, by 

 method 8, I have not drawn, as both were too obscure. The 

 spectra of metallic tin and chromium were prepared by method 

 2 (H, or coal-gas, and 0) and by method 8 (liquid electrodes). 



From the investigations adduced, it follows, as I have already 

 expressed in an earlier paper, that every compound of the first 

 order which is not decomposed, and is heated to a temperature 

 adequate for the production of light, exhibits a spectrum peculiar 

 to this compound, and independent of other circumstances. 



The decomposition of compounds may be caused by the gases 

 of flame, or even by the high temperature alone, independent of 

 the gases : thus, for instance, chloride of bismuth, if its solution 

 is used instead of the electrodes iu the electrical discharge in 

 chlorine, is decomposed by the high temperature of the electrical 

 discharge alone. Individual compounds resist even this tem- 

 perature — as, for instance, bromide of copper and iodide of copper, 

 and others, which show the spectrum of the metal together 

 with that of the compound. But if, instead of the solutions 

 which produce a diminution of temperature by their rapid 

 evaporation, the salts themselves are used as solid electrodes, at 

 the then higher temperature most salts are decomposed (as, for 

 instance, these copper salts), and but few compounds remain 

 undecomposed. 



With a great number of metals such a decomposition takes 

 place even under the temperature at which a luminous appearance 

 is observed ; hence in this case it has as yet been impossible to 

 observe a direct spectrum of the compounds ; and therefore a 



