and of Simple Substances. 175 



It is not possible to obtain the spectrum of the sulphur com- 

 pounds of these metals, which is also the case with those of 

 copper and bismuth, as all sulphur compounds appear to be 

 completely non-volatile. I made the experiment by bringing the 

 metallic sulphide into hydrogen saturated with sulphide of car- 

 bon, or into sulphuretted hydrogen, which burned in an oxygen 

 atmosphere (method 5). In the latter experiment no spectrum 

 was seen; in the former frequently a spectrum which was produced 

 by the decomposition of the sulphur compounds. Nor did I 

 succeed in obtaining the spectra of cyanogen compounds; if 

 the salts of the above metals are brought into burning cyanogen, 

 spectra are formed as in ordinary flame. 



As regards the compounds of lead, gold, iron, and manganese, 

 I could only obtain spectra in a very few cases. Chloride of 

 lead by method 6 (volatilization in the glass tube), or acetate of 

 lead by method 2 in the oxidizing part of the flame, always give 

 the spectrum of the oxide. By chloride of lead or acetate of 

 lead, heated according to method 2 (H and 0) in the reducing 

 part of the flame, the spectrum of the metal could never be 

 obtained free from that of oxide of lead. 



By method 8 (liquid electrodes) it is obtained free from that 

 of the oxides. The lines which are seen in this method are de- 

 noted by the letter e in the spectra given ; it is probable that 

 they are also contained in the spectrum of lead which is obtained 

 by the flame, and that they are only obscured by the brightness 

 of the oxide spectrum. By method 3 (H and CI), the chloride- 

 of-lead spectrum is formed even with the most varied lead com- 

 pounds, but somewhat obscurely. Iodide of lead, bromide of 

 lead and fluoride of lead, examined by method 6 (volatilization 

 in the glass tube), gave always only the oxide-of-lead spectrum, 

 formed by the burning of lead to oxide of lead at this high 

 temperature. 



Chloride of gold, by method 6 (volatilization in the glass tube), 

 gives the spectrum of this compound. In no other way could I 

 obtain a spectrum of a gold compound, neither by using hypo- 

 sulphite of protoxide of gold and soda, nor potassio-iodide of 

 gold or potassio-cyanide of gold. The spectrum of gold itself 

 is prepared with chloride of gold by method 8 (liquid electrodes). 



Of iron I could only obtain the spectrum of an oxide and that 

 of the metal. The former is produced if the chloride or iodide 

 are used by method 6 (volatilization in the glass tube), and very 

 distinctly with sesquichloride of iron. By method 2 (H, or coal- 

 gas, and 0), according as the volatile iron salt is brought into the 

 reducing or oxidizing flame, this spectrum is obtained simul- 

 taneously with that of iron or without it. By method 8 (liquid 

 electrodes), using concentrated solution of sesquichloride of iron, 



