﻿certain Gases in Geisskr's Tubes. 407 



essentially at the blue end which is so characteristic of nitrogen, 

 lnenrst observation, which I communicated in 1866 at the May 

 meetmg of the Niederrheinische Gesellschaft fur Natur- nnd 

 Heilkunde was an accidental one. In an experimental investi- 

 gation on the relation between the refractive indices and the 

 densities of bodies, I used in determining the refractive indices 

 he spectrum of ignited hydrogen consisting of the three bright 

 lines. On this occasion a hydrogen-tube, which had previously 

 Seen used for frequent measurements, became one day suddenly 

 altered : its previous beautiful red light became white ; the violet 

 line disappeared in the spectrum; and instead of the former 

 spectrum a continuous one was formed which, especially in the 

 green was most beautifully shaded. The first supposition, that 

 the tube haa begun to leak and had admitted some air, was seen 

 to be erroneous on making a comparison with the spectrum of 

 nitrogen, and in further experiments which had for their object 

 a more accurate investigation of the spectrum; for the tube 

 just as suddenly assumed its previous red colour, and exhibited 

 as spectrum the three well-known bright lines only. Hence it 

 was established that, under circumstances which as yet were not 

 determinable, hydrogen could furnish a continuous spectrum. 



Shortly afterwards I obtained from Dr. Geissler of Bonn a 

 number of new hydrogen-tubes. One of them was found on in- 

 stigation to undergo just^ such a change; it suddenly became 

 white, and yielded the contmuous spectrum observed in the first 

 tl l ' ?u ,' W ^- Ch is an ordina ry one with platinum elec- 

 tees, still shows this spectrum when the induction-current 

 merely passes through it. The spectrum is so rich and beautiful 



Sfi.'SsS? Il 1S mcah to desmbe > a ° d stm ™« diffi - 



At the least refrangible end it exhibits the red line H a, next 

 to which is first of all a perfectly dark field, then, at about 0-15 

 of the distance H «-D from H «, the contmuous spectrum, which 

 reaches froin there not quite to H y -that is, not quite to Fraun- 

 K S Ta i the K , d and yell0W P art abo «t eighteen beau- 

 S,? ed m T\°i leS5 broad bands ™Y ^ discriminated, 

 f I ul u° ne ? ° f Whlch are '' em arkable for their brightness. In 

 the bright bands a quantity of very fine lines may be recognized ; 



L ALTS- yh l USln ?; 3 L Sl ' eater "'^ber of prisms, asllucker 

 has done, this part would be resolved into individual lines, and 



St^ScktttH^ ^«-- d - 



avjahnffen Jubelfeter der UmversitM Bonn. Bona : Marcus, 1868 

 2E 2 



