366 M. A. Wiillner on the Spectra of some 



to balance the atmospheric pressure, some well-dried gas was 

 passed in; the tube was then exhausted, and this process re- 

 peated until an almost perfect vacuum was obtained; the tube 

 was then left for twenty-four hours in communication with the 

 drying-apparatus of the pump. The experiment could then be 

 proceeded with by allowing perfectly pure and dry gas to enter 

 while the electrical current passed between the two most distant 

 electrodes. When the pressure in the apparatus had become 

 equal to the atmospheric pressure, it was increased by gradually 

 pouring mercury into the long leg of the U-tube, which was 

 provided with a scale to measure the variations in pressure. 



a. Hydrogen. 



At low pressures, the use of a large RuhmkorfPs inductorium 

 with six Grove's elements gave successively three spectra iden- 

 tical with those described in M. Wullner's first memoir. First, 

 when the pressure is the lowest that can be attained with a Geiss- 

 ler's pump, a discontinuous spectrum, or one of the second class, 

 is obtained, consisting of six groups of very brilliant green lines ; 

 then, from about 1 to 2 or 3 millims., a spectrum consisting of 

 three bright lines, Ha, H/3, H7 (the spectrum of the second class 

 described by PKicker) ; and lastly, with a further increase of pres- 

 sure, a continuous (or first-class) spectrum extending from the 

 orange to the blue, with a certain number of bright lines beyond 

 H« and H/3 (which still remain). From M. Wiillner's new ob- 

 servations, this spectrum remains in its full brightness at 200 

 millims. pressure ; thence it becomes paler, but otherwise retains 

 the same appearance up to 400 millims. and even beyond. At 671 

 millims. the space between Ha and the commencement of the 

 green is almost completely dark ; the remainder of the continuous 

 spectrum is still pretty bright; the line H/3 has almost disap- 

 peared, and in its place there is only seen a large band a little 

 brighter than the surrounding parts of the spectrum. 



As the pressure still increases, the spectrum again becomes 

 more luminous, the yellow and orange parts gradually reappear, 

 the line Ha is still very bright, but a little indistinct at the 

 edges; beyond this line a continuous spectrum extends from the 

 orange to the violet, and presents a maximum lustre in the 

 place previously occupied by H/3. The spectrum retains the 

 same aspect up to 1400 millims., but becomes far more luminous; 

 it is particularly brilliant about H/3, which appears to have simply 

 widened; the same is the case in the violet, where a broad 

 bright band seems to have taken the place of the old line H7. 



As the pressure increases, the spectrum of hydrogen appears 

 more and more like the absolutely continuous one of an incan- 

 descent solid body. At 2240 millims. (that is to say, nearly 



