730 



Prof. R. W. Wood on Hydrogen Spectra 



subjective yellow, which, added to the real yellow, produced 

 a resultant tint very close to that of D 3 (helium). 



The tube is of ordinary soft glass, about 2 metres long with 

 a bore of 5 mm., bent as shown in fie:. 1. The wall of 



bent as shown in fie;. 1. 

 Fisr. 1. 



the tube is blown out into a thin-walled bulb at the bends 

 as shown, so that clear vision is obtained through the central 

 portion; this portion is used "end-on-" for observing and 

 photographing the spectra. The electrodes are cylindrical 

 in form, and are made of sheet aluminium of about the 

 thickness of writing-paper. They fit snuiily within the 

 cylindrical bulbs, being supported by the walls, and the 

 sealing-in wire is twisted with a short strip of aluminium, 

 which forms a continuation of the cylinder. .This strip is 

 about 5 mm. wide and 2 em. long, and should lie along the 

 axis of the bulb. It is on this strip that the width of the 

 Crookes dark-space is observed and the degree of vacuum 

 judged. As all data regarding pressure in the present paper 

 will be given in terms of Ihe dark-space, it will be advan- 

 tageous in repeating or extending the work to operate with 

 electrodes of this shape. The bulbs were 2'5 cm. in dia- 

 meter and 12 cm. long. 



Hydrogen is supplied to the tube either from an electro- 

 lytic generator through a long and very fine capillary, no 

 drying-tube being used, with the pump withdrawing the 

 gas continuously at the other end, or through a palladium 

 tube. The first method is preferable for an extended study 

 of the lines of the Balmer series, as the suppression of the 



