and of Simple Substances. 179 



salt unites with copper and forms a chloride-of-copper spectrum, 

 since chloride of copper is not decomposed at this temperature, 

 — and further, that if oxide of copper is mixed with chloride of 

 sodium, the free chlorine would also partially combine with cop- 

 per, as cupric salts investigated by method 3 (H and CI) give 

 the chloride-of-copper spectrum. These decompositions do in 

 fact take place. If, for instance, chloride of sodium is fused 

 upon a polished copper plate at a high temperature, the bright 

 copper is attacked ; and if it is more strongly heated, chloride of 

 copper is readily observed in the flame by the spectrum-apparatus. 



Chemical processes at high temperatures may be studied by 

 means of the spectra. The most beautiful method is by the 

 sunlight. For compounds which give a more continuous spec- 

 trum, such a study would be more difficult, but very easy for 

 the haloid compounds of barium, strontium, and calcium. I 

 think of shortly making such a set of experiments. 



As I supposed in my former memoir, we may by these decom- 

 positions compare, in an interesting manner, the temperatures 

 which give rise to developments of light of various kinds — thus, 

 for instance, the temperature of electrical discharges with the 

 temperature of light produced by combustion, or with that of 

 the solar light. It follows from the experiments adduced, espe- 

 cially from those with copper compounds, that the electrical 

 sparks have a lower temperature when they come from liquid 

 electrodes, that this temperature is about that of the oxyhydro- 

 gen blowpipe, but that the temperature resulting from the pas- 

 sage of sparks between solid electrodes is much higher than that 

 of any flame. 



The spectra of the metalloids and of their compounds with one 

 another are not very numerous. 



The spectra of hydrogen, oxygen, nitrogen, and chlorine only 

 result from the electrical spark ; they are known, just as are 

 those of iodine and bromine, which can be prepared by the elec- 

 trical spark or by the absorption of white light. 



If iodine is examined by method 6 (volatilization in the glass 

 tube), the spectrum depicted is obtained ; but if very much 

 iodine is volatilized, the absorption spectrum is produced. I 

 shall recur to these phenomena subsequently. I have not suc- 

 ceeded in obtaining a spectrum of chlorine or of bromine in a 

 similar manner, not even with bromine when hydrogen, in which 

 bromine vapours were volatilized, burnt in oxygen ; a continuous 

 brightness was all that could be perceived. 



The spectrum of sulphur was prepared by method 7 (solid 

 electrodes), those of selenium and tellurium by method 6 (volati- 

 lization in the glass tube). 



