688 Professor Dewar [June 10, 



either at the same time or by merely varying the discharge by means 

 of a Leyden jar. 



Experiments made in the same way with carbonic oxide instead of 

 nitrogen led to precisely the same results. 



2. Observations on the SparTc between Magnesium Points in Hydrogen 

 at reduced Pressures. 



A tube, similar to those employed with nitrogen and carbonic 

 oxide, was attached at one end to a Sprengel pump and mercury 

 gauge, and at the other end to an apparatus for generating hydrogen. 

 Dry hydrogen was passed through for some time, and the connection 

 with the hydrogen apparatus closed. On sparking with the hydrogen 

 at the atmospheric pressure, the line at 5210 and its attendant series 

 were visible, and were still visible when a small Leyden jar was used 

 with the induction coil, but disappeared almost entirely when a large 

 Leyden jar was used. When the pressure of the hydrogen was 

 reduced to half an atmosphere, the line at 5210 was seen faintly when 

 a large Leyden jar was used, but not the series of fine lines. When 

 the pressure was reduced to 180 millims., the series of fine lines began 

 to show when the large jar was used. By still further reducing the 

 pressure the whole series was permanently visible when the large jar 

 was used ; but when the exhaustion was carried still further they 

 grew fainter, and almost disappeared. On gradually readmitting 

 hydrogen, the same phenomena recurred in the reverse order. 



3. Observations on the Arc ivith Magnesium and Hydrogen. 



The line at 5210 is not seen in the arc in a lime or carbon 

 crucible when magnesium is dropped in without the introduction of 

 hydrogen. If, however, a gentle stream of hydrogen or of coal gas 

 be led in through a perforation in one of the electrodes, the line at 

 5210 immediately makes its appearance, and, by varying the current 

 of gas, it may be made to appear either bright or reversed. However 

 small the current of hydrogen be made, the line can be detected as 

 long as the current and the supply of magnesium continue, but 

 disappears very quickly when the current of gas ceases. 



4. Observations on the Flame of Burning Magnesium. 



The line at 5210 may often be seen in the flame of magnesium 

 burning in air, but both it and the series of fine lines which accom- 

 pany it come out with greatly increased brilliance if the burning 

 magnesium be held in a jet of hydrogen, of coal gas, or of steam. 



The experiments above described, with nitrogen and carbonic 

 oxide at reduced pressures, are almost if not quite conclusive against 

 the supposition that the line at 5210 is due merely to the lower 



