174 Prof. Mitscherlich on the Spectra of Compounds 



only obtain, free from those of their compounds, in method 

 2 (H, or coal-gas, and 0), by using a very small quantity of 

 fluoride of barium or fluoride of strontium. The spectrum of 

 calcium I could never obtain by the flame quite free from those 

 of the oxide or of the haloid salt. By method 7 (solid electrodes) 

 the latter is obtained free from admixtures by the use of hydrogen 

 and metal ; but at this high temperature there occur, along with 

 the lines which are produced in the flame, a great number of 

 more feebly luminous lines. The lines which could only be 

 obtained by this method are not depicted in the spectra given. 



In the case of barium, I could free the spectrum of the oxide 

 from that of the metal in the same way as with bismuth ; I even 

 succeeded, when I investigated nitrate of baryta or iodide of 

 barium by method 1 (wick of platinum wire), in producing the 

 spectrum of baryta in individual parts of the flame. In order 

 to obtain the spectra of the oxides of the other alkaline metals, 

 which indeed can usually only be obtained simultaneously with 

 the spectra of their metals, 1 have used in method 1 a mixture 

 of nitrate of ammonia with the nitrates or a solution of the 

 iodides ; the latter are decomposed at high temperatures, and 

 give the spectra of the oxides very beautifully. In these three 

 metals I succeeded in obtaining the spectra of the haloid salts 

 free from the spectra of the oxides, and but seldom mixed with 

 those of the metals. The chlorine compounds have been investi- 

 gated by method 1 (wick of platinum wire), and by method 3. 

 (H and Cl). It is to be observed that in the hydrochloric flame 

 itself they give no light at all, not even if the compound volati- 

 lized by the hydrogen burning in air enters the flame. This can 

 only be explained by assuming that the temperature produced 

 by the union of hydrogen and chlorine is not high enough to 

 bring chlorine compounds to luminosity. The spectra of these 

 compounds are obtained pure if there is more hydrogen than is 

 necessary for union with chlorine, which free hydrogen burning 

 in the air produces a higher temperature. 



The spectra of bromine compounds were prepared by method 

 4 (Br and H), using the bromides in solutions, which were in- 

 troduced into the flame by method 1. 



The iodine compounds are decomposed, as stated ; the iodide- 

 of-barium spectrum is only obtained if the iodine compound is 

 volatilized in a hydrogen flame which contains much iodine 

 vapour, — although truly it is then not free from that of the 

 oxide, so that the feebler lines of the iodide-of-barium spectrum 

 cannot be recognized ; but the lines depicted occur with all their 

 sharpness. 



Fluorine compounds were investigated by method'2 (H, or 

 coal-gas, and 0), using a mixture of the oxide with fluoride of 

 ammonium in abundance. 



