452 MR. A. FOWLER ON THE SPECTRUM OF MAGNESIUM HYDRIDE. 



are not so restricted. Since these lines may be obtained in other ways, unaccom- 

 panied by the flutings, without the aid of hydrogen (as in the ordinary spark in air), 

 it is probable that no chemical action is involved in their production in the arc in 

 hydrogen. On the other hand, the flutings in question have only been produced in 

 the presence of hydrogen, and since they may be obtained without direct electrical 

 aid (as in burning the metal in a Bunsen flame), a chemical union of the two elements 

 provides the most satisfactory explanation of the observations. 



One of the chief objections to the above explanation of the flutings is that, 

 although the spectrum may be easily produced, the compound itself has not yet been 

 clearly recognised by chemists. WINKLER* possibly obtained it as part of a mixture 

 resulting from heating magnesium with its oxide in an atmosphere of hydrogen, but 

 no other chemical evidence of the existence of the compound has been brought 

 forward. Further evidence has been sought by an analysis of the deposit formed by 

 the passage of the arc in hydrogen. This consisted in part of thin spherical shells of 

 magnesium, and partly of a fine black or dark grey powder, probably mixed, as a rule, 

 with a small proportion of magnesium oxide. The analysis of this deposit was kindly 

 undertaken by Prof. TILDEX, who reports that the volume of hydrogen given off on 

 dissolving the sample supplied was always less than the amount calculated on the 

 supposition that it was pure magnesium, and that there was no evidence of a hydride. 

 This was, perhaps, only to be expected from the observation that there was no 

 appreciable absorption of hydrogen during the passage of the arc, but there was the 

 possibility that such absorption might be partly counteracted by the liberation of 

 hydrogen from the heated poles. 



In explanation of the spectroscopic observations, on the hypothesis that the flutings 

 are produced by magnesium hydride, it may be supposed that a very small quantity 

 of the, compound is competent to produce a brilliant spectrum. The compound would 

 not be unique in this respect, as the cyanogen flutings are ordinarily well developed 

 in experiments on carlxm compounds even if only a very small trace of nitrogen be 

 present. Or it may be, as suggested by Mr. BROOKS, that the compound is in most 

 experiments decomposed as quickly as it is formed. A perfectly definite conclusion 

 does not at present seem to be possible. 



The, Spectrogra/phs Employed. 



For obtaining a general record of the visible spectrum, a small Littrow prismatic 

 spectrograph, giving the region C to K on a plate 6^ inches long, was employed. The 

 photograph reproduced in Plate 12, No. 1, was taken with this instrument. 



Numerous photographs were also obtained with a larger instrument of the same 

 type, giving the region 6700 to 4700 on a 12-inch plate. A great deal of time was 

 spent in reducing these photographs, using ROWLAND'S wave-lengths, but a concave 



* 'Ber. Deutsch. Chem. Gesell.,' vol. 24, p. 1973 (1891). 



