Discharge in Rarefied Gases. 185 



of about 1J millim. diameter, as in fig. 9, all the positive 

 layers in both cylinders are yellowish-red, and the light of 

 the narrow tube is blue. But from the end of the narrow 

 tube turned towards the anode secondary negative light 

 radiates into the wide cylinder, whose rays in the prolongation 

 of the narrow tube show precisely the same blue colour and 

 the same spectrum as the mass of the light of the narrow tube 

 from the end of which it springs. 



If with increasing exhaustion the rays of secondary nega- 

 tive light lengthen, the prolongation shows always the same 

 blue colour ; and so blue light with its peculiar spectrum may 

 appear at each part of the tube previously occupied by the 

 yellowish-red light to which the secondary negative rays ex- 

 tend. The neighbouring first positive layer shows yellowish- 

 red light. 



If several equally wide cylinders, lying in a straight line, are 

 connected by narrow tubes of varying width each projecting 

 into the wide cylinders, the blue wdiich the narrow tubes show 

 when the density is small possesses varying depth according 

 to the width of the tube, whilst yellowish-red mixes with the 

 blue with increasing width. From each narrow tube into 

 the wide tube joining it on the side of the anode there issues 

 a complex formation of secondary negative light, the middle 

 portion of which in the prolongation of the narrow tube 

 possesses exactly the blue (throughout its entire length, which 

 increases with the exhaustion of the tube) corresponding to 

 the narrow tube from which the secondary negative rays 

 issue. On the other hand the positive layers in all the 

 cylinders show precisely the same yellowish-red colour. 



It must be confessed that these, with numerous similar phe- 

 nomena, produce the impression that each secondary negative 

 bundle represents a motion which, excited at the 'point of origin 

 of the bundle, is transferred to the surrounding medium ; hence 

 each particle affected , as far as the excitation is propagated, 

 assumes the characteristic form of motion which is produced 

 at the point of origin of the rays ; whilst a comparison of 

 the discharge at any point with conduction in metals and 

 electrolytes can afford a guide only for the relationships at 

 che point itself. 



The narrower the tubes interposed between the wider tubes 

 are made, the purer, as already mentioned, does the blue be- 

 come, and the more nearly do all the bands in its spectrum 

 disappear with exception of the four definite bands, in which 

 all the light is concentrated. 



It is now intelligible why in a uniformly wide tube, of which 

 the positive light is throughout yellowish-red, the kathode is 

 surrounded by blue light. 



