340 Prof. J. Trowbridge on the Spectra of 



appreciably absorb the ultra-violet rays until the wave-length 

 1800 was reached ; and this absorption did not affect my 

 conclusions, since my work was confined to the portion of the 

 spectrum studied by astrophysicists. The diameter of the 



Fig. 1. 



narrowest portion of the tubes was about 1 centim. I employed 

 still another form of tube, shown also in fig. 1 . At one side 

 of this tube there was an adjunct consisting of a palladium 

 tube P, joined on the outside of the spectrum-tube to a 

 platinum wire which was inserted in the walls of the tube. 

 A glass vessel filled with dilute sulphuric acid slipped over 

 the palladium tube. At one end of this glass vessel was 

 inserted a platinum wire. The palladium tube was made the 

 cathode, and this platinum wire the anode of a battery. In 

 this way hydrogen is liberated on the surface of the palladium, 

 is occluded, and can be made to pass into the spectrum-tube*. 

 The tubes which were not provided with palladium tubes 

 were filled with hydrogen or nitrogen through a series of 

 tubes filled with caustic potash— and through a number of 

 tubes filled with phosphoric pentoxide provided with partitions 

 of glass-wool. After many attempts, I adopted the following 

 arrangement of these tubes which proved satisfactory. In the 

 first place, I found it necessary to discard all gums or other 

 adhesive material and to use ground-glass joints and mercury 

 * A device due to Dr. William Rollins, Boston, U.S. 



