180 Prof. Mitscherlich on the Spectra of Compounds 



The spectrum of phosphorus obtained by method 6 (volatili- 

 zation in the glass tube) is only visible if very much hydrogen is 

 burned with traces of phosphorus ; if more phosphorus is used, 

 much phosphoric acid is formed, which separates as ignited 

 solid substance and, like all such substances, gives a continuous 

 spectrum, which by its brightness conceals the spectrum of phos- 

 phorus. This experiment gives a clue to the affinity at high 

 temperatures of hydrogen to oxygen as compared with that of 

 phosphorus to the same substance; for phosphorus first burns 

 with a green flame, probably to phosphorous acid j subsequently 

 hydrogen burns to water, and phosphorous to phosphoric acid. 



Carbon compounds, examined by methods 5 (combustion of 

 gases) and 6 (volatilization in the glass tube), give different spec- 

 tra according to the nature of the body combined with carbon. 

 Hydrocarbons and chlorides of carbon always show the well- 

 known spectrum of coal-gas flame, which arises from the carbon 

 as such ; by method 7 (solid electrodes) the spectrum of hydro- 

 gen is simultaneously seen, and for the most part a separation 

 of carbon. But if the carbon is combined with oxygen or sul- 

 phur, as in carbonic oxide and sulphide of carbon, a continuous 

 brightness is observed on burning, in which I did not succeed 

 in discovering dark or bright lines. The metalloid which is 

 united with carbon (for instance chlorine, bromine, iodine, and 

 sulphur) can never, with very few exceptions, be recognized by 

 the flame when investigated by methods 1 to 6. With nitrogen 

 this is not the case ; if the compound is not very rich in oxygen, 

 it can be recognized by the formation of a spectrum of ammonia. 



I obtained the spectra of silicon and fluorine by effecting the 

 electrical discharge in silicofluoride of hydrogen and in hydro- 

 fluoric acid. The spectrum of fluorine which I obtained with 

 that of hydrogen alone, was deducted from that of fluoride of 

 silicon, and thus the latter recognized. Both spectra consist of 

 individual lines. These spectra, like the rest which are only 

 formed by the passage of the sparks from dry electrodes, I have 

 not depicted. Silicon and fluorine, investigated by method 5 

 (combustion of gases), using fluoride of silicon with hydrogen, 

 give only luminosities in which no shadings are perceptible. 



The spectrum of boron was prepared with the use of boracic 

 acid by methods 6 (evaporation in the glass tube) and 8 (liquid 

 electrodes), and with fluoride of boron by method 7; in both 

 cases I obtained the same spectrum. 



Spectra of the compounds of the metalloids with one another 

 I could only observe in small number. Most compounds give a 

 spectrum which, from the small intensity of its light, cannot be 

 investigated — thus, for instance, hydrogen and hydrochloric 

 acid; others give a continuous one, as, for instance, sulphu- 



