Bacon — Phenomena Observed in Crooked Tubes. 311 



the liquid condition. Subsequent molecules of hydrogen 

 would suffer like losses (though less rapidly, owing to occa- 

 sional impacts on hydrogen instead of platinum) until a film of 

 hydrogen molecules should form a more perfectly elastic coat- 

 ing, and equilibrium would be established (with higher tem- 

 perature in the film than in the surrounding atmosphere) 

 when impact on the platinum of a fresh molecule of hydrogen 

 raised the temperature of the film to the point of expelling 

 another molecule from its surface. Evidently such a film 

 would give off molecules when heated, and knowing platinum 

 black to have this. action on gases in higher degree than glass, 

 we can see how the vacuum might be increased by the plati- 

 num coating, and why heating the walls of the tube should 

 have the effect of partially restoring its activity. It is also 

 interesting to note that by this theory the higher the vacuum 

 the more important relatively this action becomes, as the 

 exposed surface of imperfect elasticity remains constant, while 

 the number of gaseous molecules diminishes, and thus the 

 ratio increases of impacts resulting in loss of velocity. 



This may sufficiently explain the recrudescence on heating 

 the walls of the tube, but does not explain the emission of 

 rays, and hardly seems sufficient for the continuing gradual 

 increase of vacuum. Why should not these be accounted for 

 on the hypothesis of a resolution of atoms into electrons or 

 emanations under the discharge in a Crookes tube, and radia- 

 tion of these (to which the glass walls might be pervious, more 

 or less, as to the ether, though entirely impervious to molecu- 

 lar matter) of different polarity from the different electrodes ? 

 Of these rays we know very little, except that they differ 

 materially from the forces with which we are accustomed to 

 deal. The X-rays, emanating from the anode, are absorbed by 

 mass apparently much in the way that light is absorbed by 

 partially transparent substances, as is shown by the X-ray pic- 

 tures, and they have been considered (though perhaps on 

 insufficient grounds) to have also the velocity of light, which 

 would tend to suggest their relation to the electric current 

 streaming also from the anode ; but no ordinary reflection or 

 refraction of them is possible. On the other hand, about all 

 that we know of the cathode rays is that they can be deviated 

 by a magnetic field, thus showing some of the characteristics 

 of mass, and, moreover, they seem to move more slowly than 

 light. 



We are not reduced here to consideration of the electron 

 solely. Possibly we may have to do with one or both of the still 

 more ethereal forms of matter postulated to account for the 

 curious changes attending the transformation of radium to 

 helium, but, from present indications, is it not probable 



