344 Prof. J. Trowbridge on the Spectra of 



stage of exhaustion to occlude hydrogen to such a degree 

 that the state of exhaustion in the spectrum-tube is apparently- 

 raised, and no discharge can be forced through it until it is 

 subjected to external heat. This occlusion or destructive 

 electrical dissociation is very much in evidence when the 

 terminals of a battery of ten or twenty thousand cells are 

 connected to an #-ray tube through a resistance of several 

 megohms. If this tube is properly exhausted, that is by 

 repeated heating and by the use of electrical discharges, no 

 current can be forced through it unless the tube is repeatedly 

 flashed with a powerful Bunsen burner. Suddenly, however, 

 under the action of heat a blue cloud arises in the tube ; the 

 cathode beam appears ; the anti-cathode grows white-hot, and 

 the exterior resistance must be speedily increased to save the 

 tube. The pressure has apparently been greatly diminished ; 

 but the experiment of Dr. William Rollins with two connected 

 .r-ray tubes, one of which is heated and the other not, the 

 heated one carrying a current while the unheated one resists 

 even an eight-inch spark, shows that this is not the case, and 

 that we have to deal with electrical dissociation and not with 

 a change of pressure. Let us follow this experiment further. 

 Presently the blue cloud in the tube grows smaller ; the 

 cathode-beam disappears, and the cloud slowly sinks into the 

 positive electrode. At that moment the #-rays flash out with 

 great brilliancy ; but in a few seconds the light in the tube is 

 totally extinguished, and a further heating from an external 

 source is necessary before a current can be again forced 

 through the tube and the phenomena can be repeated. It 

 can be repeated, however, as often as desired, although the 

 apparent resistance of the tube is undoubtedly raised by a 

 more or less permanent occlusion of what I believe is water- 

 vapour. This experiment strongly recalls the familiar one of 

 the glowing of platinum under a stream of non-ignited 

 hydrogen. The electrical dissociation of the aqueous vapour 

 evidently produces an intense heat at the anti-cathode, and 

 the x-Ttiys are emitted during this dissociation. 



If we suppose that there are material bodies circulating 

 about the sun which are charged negatively while the sun is 

 charged positively, we might conceive of a similar action of 

 a difference of potential on rarefied aqueous vapour which 

 might be competent to produce a corona. 



The white spectrum of hydrogen produced by the hydrogen 

 coming from palladium is, in its main features, similar to that 

 obtained from electrolytic hydrogen which has been passed 

 through the drying apparatus I adopted. There are, how- 

 ever, some bands which need examination to determine 



