Electron Atmospheres of Metals. 321 



What is required is a semitransparent film with an optical 

 surface equivalent to that of a speculum plate. It is possible 

 that by using a plate of gold which has never been previously 

 used as a cathode, a suitable film may be procured, for it 

 seems possible that the surface may be rendered spongy by 

 disintegration, resulting in the projection of microscopic 

 particles of gold against the film. 



Very good results were obtained by gilding the convex 

 surface of a plano-convex quartz lens of six-foot focus. The 

 central zone of the lens was protected from the deposit by a 

 circular disk of metal 5 mm. in diameter. x\n electrode of 

 silver-leaf was attached to the gold film at the edge of the 

 lens, which was then laid carefully upon a flat of speculum 

 metal. The centre of the system of Newton's rings observed 

 with white light was brought to the centre of the ungilded 

 circular patch by placing a small weight at the proper point 

 on the lens which gave it the requisite tilt. Even at the 

 centre of the system the interference colour was very pale, 

 showing that the air-film was several wave-lengths in thick- 

 ness at the so-called point of contact. By applying pressure 

 it was ascertained that the edge of the zone at which the 

 gilded surface commenced was about 8 wave-lengths from 

 the speculum surface when a first-order colour appeared at 

 the centre. The pressure was now removed causing a still 

 further separation, nevertheless a current easily measurable 

 with the milliammeter was obtained with one dry cell. If 

 the lens was tilted by shifting the weight, the gold surface 

 was brought into closer proximity to the speculum surface 

 on one side of the ring system, and the ammeter indicated 

 an increase of conductivity. This experiment is very easily 

 repeated, and appears to be free from objection, but unfor- 

 tunately it is not well adapted to quantitative work on 

 account of the variable distance due to the curvature of one 

 of the electrodes. 



At this stage of the work I left Baltimore, taking it up 

 again at my East Hampton laboratory. Using a D'Arsonval 

 galvanometer I have obtained currents with much greater 

 separation of the plates. By shaking up lycopodium in a 

 jar and allowing the spores suspended in the air to settle on 

 a speculum plate, I have succeeded in obtaining a very satis- 

 factory series of readings by squeezing the plates together 

 with a micrometer screw, the spores being quite compressible. 

 In spite of the fact that they are very resinous and probably 

 good insulators we have the possibility of absorbed moisture. 



Similar results were obtained when the speculum flats 

 were separated with small flakes of very thin mica. The 



Phil. Mag. S. 6. Vol. 24. No. 140. Aug. 1912. Y 



