1878.] on the Reversal of the Lines of Metallic Vapours. 495 



described, and not otherwise, we could only ascribe their origin to the 

 mixtures as distinct from the separate elements. It became a question 

 of interest, then, whether we could find the conditions under which the 

 same mixtures would give luminous spectra, consisting of the lines 

 which we had seen reversed. On observing sparks from an induction 

 coil taken between magnesium points in an atmosphere of hydrogen, 

 we soon found that a bright line regularly appeared, with a wave- 

 length about 5,210, in the same position as one of the most con- 

 spicuous of the dark lines we had observed to be produced by vapour 

 of magnesium with hydrogen in our iron tubes. This line is best 

 seen, i.e. is most steady, when no Leyden jar is used, and the rheotome 

 (the coil we used has an ordinary self-acting one) is screwed back, so 

 that it will but just work. It may, however, be seen when the coil is 

 in its ordinary state, and when a small Leyden jar is interposed ; 

 but it disappears (except in flashes) when a larger Leyden jar is used, 

 if the hydrogen be at the atmospheric pressure. This line does not 

 usually extend across the whole interval between the electrodes, and is 

 sometimes only seen near the negative electrode. Its presence seems 

 to depend on the temperature, as it is not seen continuously when a 

 large Leyden jar is employed, until the pressure of the hydrogen and 

 its resistance is very much reduced. When well dried nitrogen or 

 carbonic oxide is substituted for hydrogen, this line disappears entirely ; 

 but if any hydrogen or traces of moisture be present it comes out 

 when the pressure is much reduced. In such cases the hydrogen lines 

 C and ¥ are always visible as well. Sometimes several fine lines 

 appear on the more refrangible side of this line, between it and the 

 b group, which give it the appearance of being a narrow band, shaded on 

 that side. We have used various samples of magnesium as electrodes, 

 and they all give the same results. We have also used hydrogen pre- 

 pared and purified in different ways : hydrogen prepared by the action 

 of zinc on dilute sulphuric acid, purified by an acid solution of bichro- 

 mate or permanganate, and by potash, and dried by sulphuric acid ; 

 electrolytic hydrogen ; hydrogen from dry formiate of soda and soda 

 lime; hydrogen occluded by sodium and expelled by heat ; and hydro- 

 gen occluded by palladium and expelled by heat. In the last two cases 

 the whole apparatus was connected by fusion, and a Sprengel pump, 

 also connected by fusion, employed to remove the air. In all cases 

 the phenomena were the same. 



In addition to the above-mentioned line, we observed that there 

 is also produced a series of fine lines, commencing close to the most 

 refrangible line of the h group, and extending with gradually dimi- 

 nishing intensity towards the blue. These lines are so close to 

 one another, that in a small spectroscope they appear like a broad 

 shaded band. We have little doubt that the dark absorption line, 

 with wave-length about 5,140, shading towards the blue, which 



