212 MESSRS. A. SCHUSTER AND O. HEMSALECH ON 



discharge may pass through the metal vapour, which after a few millionths of a second 

 fills the whole spark gap. The spectrum is modified by the change in the mode of 

 spin-kin::. :md Mr. HEMSALECH is at present investigating this change, but our joint 

 photographs are sufficient to show that the spectrum is not reduced to that of the arc 

 discharge, the double zinc line at 4925, even with large self-inductions, remaining strong, 

 though it thins out a little. We consider the method of investigating spectra a most 

 useful one, and it allows us to distinguish at once what is an air line and what is not. 

 The obnoxious noise of the spark is also much diminished. 



[April 26, 1899.] Professor KAYSER has kindly drawn our attention to a paper 

 by Mr. AUEB v. WELSBACH (' Wiener Sitzungsberichte,' 88, II., 1883), in which an 

 interesting device is described by which strong spark spectra may be obtained 

 without the usual induction coil. It is essentially an arrangement by which the 

 spark at the break of the primary current is used in place of a gap in the secondary 

 circuit. The air lines also disappear under these conditions. In how far the spark is 

 of a similar nature to that used by us is difficult to say without direct comparison of 

 the spectra obtained. 



When the above method is applied to different metals, it is found that the calcium 

 lines H and K of the solar spectrum appear in most, possibly in all, cases. There is 

 perhaps nothing surprising in this, owing to the presence of calcium in the dust of the 

 air, and the probability of its occurring as an impurity in many metals. But what is 

 surprising is the great intensity of these calcium lines, as compared with those of the 

 metal proper when the poles used are silver. We deposited some of the metal 

 electrolytically, and Mr. J. CROWTHER fused the deposit into poles on a block of 

 willow charcoal, which contains only a very small amount of ash, but we must admit 

 the possibility of some traces of calcium having been taken up by the silver in the 

 process. 



The silver so prepared gave not only the H and K lines, but also the line at 4226, 

 with undiminished vigour. We do not offer any explanation as to the possible 

 sources of calcium, either in the metal or in the dust of the air, but the fact is 

 rendered remarkable by the appearance of the lines in question when the spectrum is 

 taken on the rotating wheel. Fig. 27 is a photograph taken on the stationary film 

 without self-induction, and the H and K lines are seen strongest near the centre of 

 the spark. With self-induction, the strongest lines of the spectrum are those of 

 calcium. In fig. 28 the disc was spinning, and there was no self-induction. The 

 lines H and K present a comet-like appearance, K being much the strongest. 

 The head of the curve corresponding to the K line is displaced towards the 

 violet, and if our interpretation of these curves is correct, the photograph seems 

 to prove that the H and K vibrations start not at the pole, but in the centre 

 of the spark, and luminosity begins not when the main initial discharge strikes 

 through the air, but, roughly speaking, about the 200,000th part of a second 

 afterwards, when the luminosity of air as shown in our photographs has completely 



