184 Dr. E. Goldstein on the Elect 



rie 



This view, then, as I will briefly show, resolves all the 

 former difficulties, and makes the whole of the hypotheses pre- 

 viously required unnecessary. The assumption made does not, 

 however, merely present a simple consistent representation of 

 the numerous phenomena which immediately lead to it ; but 

 there are, further, a large number of phenomena which har- 

 monize with it extremely well, some of which indeed make it 

 appear not only admissible but even necessary. 



Since, according to oft-quoted experiments, the positive 

 light is nothing else than an envelope of the negative, I shall 

 speak of rays of electric light also when referring to the posi- 

 tive light, and understand by it luminous particles lying in a 

 line which represents the direction of propagation from any 

 point on the surface of the layer nearest the negative pole to 

 the other bounding surface of the layer. The following pro- 

 position may be deduced from my experiments: — 



The properties which the discharge shows at any point of its 

 path are not dependent on the relationships of the point itself, so 

 much as upon the relationships of the point from which the 

 ray passing through it takes its origin. Or, somewhat differ- 

 ently expressed: — An electric ray possesses throughout its 

 entire length the properties which the discharge possesses at 

 its point of origin, and which are conditioned by the nature of 

 this point of origin. 



If, for example, two electric rays pass in equally wide simi- 

 larly shaped portions of the same discharge-tube, and so in 

 media of identical chemical and physical nature, their proper- 

 ties are different if the origin of one of them lies within the 

 portion of the tube under consideration, and that of the other 

 at the point of junction of this portion and another of less 

 width. 



It will be understood, from the example already given, that 

 this is the explanation of all the phenomena of the influence 

 of the section of the discharge on the character of the dis- 

 charge as positive or negative. I will endeavour to make the 

 proposition stated plain by means of a striking example. 



In wide tubes filled with air (for example, cylinders of 2 

 centims. or more in width) the stratified positive light has a 

 yellowish-red colour, and when analyzed with the spectroscope 



fives the spectrum of nitrogen described and figured by 

 Kicker and Hittorf, consisting of numerous closely-placed 

 bright bands. 



Narrow cylinders, on the other hand, show with the same 

 pressure a blue light, whose spectrum contains only a few of 

 the bands seen in the spectrum of the yellowish-red light. 

 If, now, two wide cylinders be connected by a narrow tube 



