410 Dr. F. P. Kerschbaum : Interference 



(iii.) That in spite o£ this rapid diffusion the velocity of 

 this diffusion is very small compared with the velocity of the 

 condensation of the vapour by liquid air through tubes of a 

 cross-section large compared with the cross-section of the 

 -slit. 



The Experiments. 



With the interference apparatus as described above the 

 •following experiments were carried out : — 



(I.) The quartz vessel, shown in figs. 1 and 3, was 

 •evacuated by means of working the Gaede pump only, till a 

 partial pressure of the remaining air was about 0'002 mm. 

 -After this the special Hg arc-lamp which has to excite the 

 Hg-vapour to resonance radiation in chamber (R) was 

 brought up to the apparatus opposite the window (w) as 

 • close as possible. A parallel beam of light enters the vacuum 

 without, however, raising the temperature of the chamber 

 ^appreciably, because the light of the arc-lamp has first to 

 pass a layer of water which is cooling the lamp. An 

 exposure of the photographic plate * is then made of two 

 hours' duration. After developing, no trace of any light 

 action can be detected on the plate. Only an exposure of 

 about ten hours' duration produces a few very faint fringes. 

 But these fringes are not sharp, and they are distant from 

 -each other about twice as much as they should be for light 

 X = 2536. In fact, these fringes are only due to the stray 

 light entering through the slit (s). Indeed, fringes due to 

 the resonance radiation of the Hg-vapour in the chamber (R) 

 •cannot be expected in this experiment, as the whole chamber 

 (V) is filled with saturated Hg-vapour. The resonance 

 light, coming through the slit, is totally absorbed in passing 

 through this chamber. 



(II.) The experiment is repeated at the same partial 

 pressure of air, but before starting the exposure the quartz 

 condensation- vessel (1, fig. 3) is submerged in liquid air and 

 kept there during the exposure. Ninety minutes'' exposure 

 now give about twenty fringes, as shown in fig. 1 (PI. VI.). 

 They are sharp fringes showing the proper distance which is 

 to be expected for light of \ = 2536. Owing to the short 

 exposure no traces of fringes due to the stray light can be 

 detected. 



From the relative times of exposures in these experiments 

 •we can roughly deduce that the intensity of the resonance 



* Lumiere plates " Violet Label " were used throughout. 



