244 Scientific Proceedings, Royal Dublin Society. 
photograph had been taken in hydrogen, this gas was displaced by 
air, and another photograph secured. Here the spectrum was a 
continuous one, overlying and obscuring the red and green bands, 
just as if the rays had proceeded simultaneously from a white-hot 
solid substance and the flame of burning calcium. It is not im- 
probable that the bright metallic electrodes had become oxidised on 
the surface, and that a trace of lime was the origin of the continuous 
rays. It was thought desirable that the spark should be examined - 
again in a vacuum; and the result led to a red flame being seen to 
proceed from one electrode, half-way towards the other, upon 
which bright white scintillations were observed. The red flame 
yielded the calcium bands ; the lines in the spectrum evidently pro- 
ceeded from the bright scintillations. hese changes in the spark- 
discharge were observed when the electrodes had become heated. 
The position of the red flame was reversed when the direction of 
the current was reversed. ‘his behaviour of metallic calcium is 
apparently similar to that described by Ramage’ in connexion 
with lithium carbonate fused toa platinum wire ; but it is probably 
connected with the phenomenon described by Horton.’ 
It should be noted here that the red, orange, and green bands 
appear inthe spark-spectrum with Hemsalech’s self-induction coil 
in. circuit, the characteristic spark-lines overlying the bands. On 
renewing the electrodes, and making a further series of experiments 
with a small motor-car sparking-coil, with the discharge between. 
calcium electrodes in an exhausted tube at 5 mm. pressure, it 
was found that there was no direct passage of the spark. The 
metal points glowed with a violet light; bright stationary spots 
of white light were seen on the negative electrode, and a great 
number of scintillations less bright on the positive electrode, 
not at one spot, but all over it. At intervals, as in the previous 
instances, a small flame of red light, lasting only a few seconds, 
passed apparently from the positive electrode. With a more 
powerful coil, the sparks between the electrodes were fairly 
continuous for a time, but still showed irregularities. 
1 <¢Qn the Distribution and Spectra of Metallic Vapours in Electric Sparks.”’ By 
Hugh Ramage, B.a., Proc. Camb. Phil. Soc., 1904, vol. xii., pt. v., p. 849. 
2 “On the Discharge of Negative Electricity from Hot Calcium and from Lime.”’ 
By Frank Horton, D.Sc. A Paper read on January 31st, 1907. Proc. Roy. Soe., 
1907, A, vol. Ixxix., p. 96. 
