336 M. E. Goldstein on Spectra of Gases. 



The discharge of a Ley den jar does not generally show any 

 stratifications, even if the positive light without the jar shows 

 them. If we introduce into a circuit containing a tube showing 

 stratified light a spark in atmospheric air, and look at the image 

 of the discharge in a rotating mirror, we shall see that the 

 image begins with a band which we cannot widen by increas- 

 ing the number of rotations of the mirror; and this band is 

 not stratified. If we do away with the -spark, the image of 

 this band will again be stratified. 



If a Leyden jar is introduced into the circuit, the resist- 

 ance of which is somewhat large, the dark space which is ge- 

 nerally seen between the positive and negative light disap- 

 pears, and the whole tube is filled with light of the colour of 

 the positive light. Introducing a spark instead of the jar, we 

 observe light in the dark space, the intensity of which varies 

 with the strength of the spark. 



The images given by the rotating mirror are characteristic 

 in this case. The blue light in the negative cylinder of a 

 spectral tube of the common form is widened into a field of 

 light, the intensity of which becomes so small, if the velocity of 

 rotation is sufficiently great, that the eye cannot distinguish it; 

 the dark space extends from the negative light to the capillary 

 part of the tube. On the other hand, feeble light fills the posi- 

 tive tube, which can be made to sink below the limit of visi- 

 bility in the rotating mirror. We thus get a good image of 

 the discharge from the capillary part only, showing a long band, 

 the width of which is equal to its length. The image just de- 

 scribed corresponds to the continuous discharge. When a 

 spark is introduced into the circuit, both cylinders are illumi- 

 nated, and we see the complete image of the whole tube flashing 

 in. The difference between a continuous and discontinuous 

 discharge is here seen in a striking way. Spark and jar pro- 

 duce here again the same effect. 



The discharge of a Leyden jar in thick sparks is affected in a 

 curious way by the magnet. The discharge shows every where 

 the colour of the positive light, and is torn into two parts which 

 run along opposite sides of the tube. Each part, however, has 

 a different character according as it emanates from the positive 

 or negative end of the tube. An explanation of this fact by 

 means of oscillating discharges does not seem possible. 



The positive light of a continuous discharge, on the other 

 hand, is deflected as a continuous thread towards one side. If 

 we introduce a spark in air sufficiently long, the discharge is 

 torn again into two parts, and the whole phenomenon is the 

 same as if a Leyden jar were introduced. 



I think I may consider it to have been proved by these and 



