and of Simple Substances, 181 



retted hydrogen, sulphide of carbon, sulphide of nitrogen, car- 

 bonic oxide. I never succeeded in recognizing dark or bright 

 lines in the spectra of these compounds. Of the compounds of 

 the metalloids with one another, I could only recognize the spectra 

 of cyanogen and ammonia : the first is already known, but not 

 in a state of purity; the lines of the carbon spectrum which 

 result from the decomposition of cyanogen when burnt in hydro- 

 gen have been assigned to the cyanogen spectrum ; the spec- 

 trum of ammonia has recently, during my research, been depicted 

 by M. Dibbits. It is interesting to recognize again in the 

 spectra of both compounds the properties of the simple sub- 

 stances resembling them. While ammonia always appears with 

 more or less intensity in the spectra of its compounds and thus 

 behaves like the metals, cyanogen, like iodine, bromine, &c, 

 loses in its compounds the property of producing a spectrum ; 

 thus hydrocyanic acid, sulphocyanogen, cyanic acid, &c. give no 

 recognizable spectrum. 



The spectrum of ammonia can best be depicted by method 5 

 (combustion of gases), by the use of oxygen ; it is obtained of 

 feebler luminosity by method 6 (evaporation in the glass tube), 

 from ammonia compounds, or best by urea. 



The decomposition and formation of cyanogen at high tempe- 

 ratures, which can easily be followed by the spectra, is interesting. 

 For if the electrical discharge passes in cyanogen, the gas is 

 decomposed, very dense carbon being deposited in characteristic 

 curves, but if much carbonic oxide is added it remains unde- 

 composed ; but if the electrical spark be allowed to strike for a 

 short time through air which contains a hydrocarbon, cyanogen 

 is formed. Hence at the same temperature the compound is 

 formed and decomposed. 



If hydrogen is brought to burning cyanogen, a change of colour 

 is speedily observed in the flame, and the formation of ammonia 

 is readily shown by the spectrum. If carbonic oxide is admitted 

 to cyanogen, no change, as already observed, is produced. Hence 

 at this high temperature hydrogen has a greater affinity for 

 nitrogen than carbon. 



Of the compounds of the metalloids, I further examined by 

 method 6 (volatilization in the glass tube), or method 5 (com- 

 bustion of gases), protoxide of nitrogen, binoxide of nitrogen, 

 nitrous acid, nitric acid, chloride of sulphur, oxychloride of phos- 

 phorus, and, further, sulphuretted hydrogen and sulphide of 

 carbon mixed with chlorine ; in all cases I only obtained lumi- 

 nosities without shadings. By a decomposition of the com- 

 pounds, selenious and selenic acids give the selenium spectrum, 

 and chloride of iodine the iodine spectrum. 



