and of Simple Substances. 



171 



3. The same arrangement as the above is used, but with chlo- 

 rine instead of oxygen, and hydrogen instead of coal-gas in the 

 burner. The spectrum which might be formed by the union of 

 these bodies is too feeble to be observed; only a very feeble 

 brightness in the green is perceptible. 



4. Bromine or iodine is evaporated in hydrogen, and this is 

 burnt in air or in oxygen, and the substances volatilized in them. 



5. If the substances to be investigated are combustible gases, 

 they are allowed to emerge out of the middle aperture of the 

 oxyhydrogen burner instead of coal-gas, and can thus be burned 

 in air or in oxygen ; if they are not combustible, they are m-ixed 

 with a combustible gas such as carbonic oxide or hydrogen. 

 The light produced by the combustion of carbonic oxide and of 

 hydrogen is so feeble that it may be quite disregarded in con- 

 sidering the formation of other spectra by its means. 



6. The solid substances are introduced into a tube, one end of 

 which is connected with a Rose's hydrogen-apparatus, or with a 

 gas-holder filled with carbonic oxide. By heating, the substance 

 is volatilized in the gases, and kindled at the other end of the 

 tube. If the substances are volatile, a small bulb is blown in 

 the tube, into which the substance is brought. 



In the above six methods the light is produced by a flame, while 

 in the two following the substance to be investigated is volati- 

 lized by the discharge of a small Stohrer's induction- apparatus. 



7. In a vessel closed by a cork, I caused the electrical spark 

 to pass either between two wires of the metal investigated, or 

 between salts fastened on the ends of wires. The vessel was 

 provided with two glass tubes, by which it could be filled with 

 any given gas and closed, or by which a current of gas could be 

 passed during the discharge. 



8. In order to be able to use liquids 

 as well as solids for electrodes, two tubes, 

 bent as shown in the annexed figure, 

 were fixed in the cork of the vessel and 

 filled with liquid till this reached the 

 lower aperture. Two thin platinum 

 wires in connexion with the induction- 

 apparatus were brought in both tubes 

 to near the aperture. The spark then 

 passes from liquid to liquid. By means 

 of tubes fixed in the vessel, as in the 

 other method, the experiment may be 

 made in different gases. It is observed 

 that if the spark only passes from liquid 

 to liquid, the gases in the vessel are with- 

 out action upon the spectrum, and that 



N2 



