Experiments in a Highly Rarefied Gas, 411 



light is about twenty times the intensity o£ the stray light, 

 measured by their photographic activity. This despite the 

 fact that the resonance light is weakened by absorption in 

 passing through the slit (s) which contains Hg vapour 

 necessarily. 



(III.) In order to get some idea of this amount of Hg- 

 vapour in the slit and the chamber (V) during the experi- 

 ment just described, I proceeded as follows : — 



The cap (R, fig. 1) was removed and the apparatus 

 thoroughly washed with C0 2 from a cylinder. Now a 

 resonance lamp as used in the diffusion experiment* is brought 

 right up in front of the slit (s), which in this experiment 

 is free from traces of Hg-vapour, as well as chamber (V). 

 Then the resonance lamp is excited. An exposure of eighteen 

 minutes gives fringes as shown in fig. 2 (Pi. VI.). They are 

 practically as intense as the fringes shown in fig. 1. Bearing 

 in mind that a layer of about 5 mm. thickness of saturated 

 Hg-vapour reduces the intensity of the resonance light to the 

 half value, the ratio 5 : 1 for the times of exposure in expe- 

 riments (II.) and (III.) indicates that the light in experi- 

 ment (II.) has to pass a vapour layer of about 12 mm. 

 thickness if saturated. Now the slit (s) is itself 15 mm. long. 

 We have therefore to imagine that, when the charcoal is 

 condensing the Hg-vapour, very probably only the slit (s) is 

 filled with this vapour, and that in the chamber (V) the partial 

 pressure is under 0*00001 mm. For a column of 500 mm. 

 length of Hg-vapour of this pressure would correspond to a 

 layer of 5 mm. thickness of saturated Hg-vapour, and would 

 therefore cause absorption, which could be detected. 



Anyhow, from experiments (II.) and (III.) we can draw 

 the conclusion, that the light emitted from the resonance 

 chamber (R) in experiment (II.) has. after leaving this 

 •chamber, only to traverse a layer of Hg-vapour equal to a 

 layer of 12 mm. thickness of saturated vapour, before 

 reaching the double slit. 



A light cell emitted in the resonance chamber (R), where 

 the centre of the bundle of the exciting rays passes — i. e. in 

 about 9 mm. distance from the slit (s) — has therefore to pass 

 a Hg-vapour layer of about 21 mm. thickness and of 

 O001 mm. partial pressure. If the partial pressure of the 

 air is small enough to be neglected for our purposes, this 

 means that not more than 16 per cent, of such emitted light 

 •cells can possibly hit Hg-molecules, i. e. matter, before fchey 



* Giving resonance light of the same intensity as the chamber (10 if 

 •excited with the same special Hg arc-lamp. 



