from Long Vacuum Tubes. 737 



the secondary spectrum appears first with the Balmer series 

 very weak, but that, in a small fraction of a second, the 2nd 

 spectrum fades out and the Balmer series brightens. With 

 no resistance in circuit with the transformer, which resulted 

 in a current of over 50 amps., the flashes were fiery purple 

 in colour and the spectrum was practically identical with 

 that obtained with a continuous exposure with only 1 amp. 

 in the transformer (PI. XXII. spectra b) ; there was no trace 

 of the new spectrum, or the group of three lines at 4050, 

 which always appeared with continuous operation of the, 

 transformer with currents of 3 amps, or more. This is of 

 importance as showing that the emission centres responsible 

 for the new spectrum and the lines at 4050 require time for 

 their development. 



This at first appeared to be a clue to a method of still 

 further extending the Balmer series, but a close inspection 

 of the plate revealed the presence of a continuous back- 

 ground of about the same relative intensity as on the plate 

 made with 1 amp. of current acting continuously. 



It seemed possible, however, that some gain might be 

 made by employing a high-potential direct current, for 

 the periods of rest between the half-cycles of the alternating 

 current might very possibly allow the gas to partially 

 recover the condition necessary (or the appearance of. the 

 secondary spectrum, as occurred completely when the switch 

 was opened and closed rapidly by hand. 



To test this point a battery of high-potential dynamos 

 giving 10,000 vol Is was substituted for the transformer. 

 No difference in the general appearance of the discharge 

 could be detected either with the green filter or the direct- 

 vision prism. This was to be expected, perhaps, as in 

 the earlier work it was found that a very short inter- 

 ruption of the current, with the tube in the black stage, 

 does not renew the '"' flash," the tube remaining dim alter 

 the interruption. 



Suppression oj the Balmer Series and the White Stage. 



All of the effects described thus far are obtained with 

 hydrogen containing water-vapour, delivered by a capillary 

 tube in direct communication with the electrolytic generator. 

 If a palladium tube is used to supply pure dry hydrogen 

 to the discharge tube, no change is noticeable for the first 

 few hours of operation, even if the tube is repeatedly 

 exhausted to a non-conducting vacuum and freshly charged 

 with hydrogen. This is undoubtedly due to the tact that 



