﻿and Absorption by Resonating Gas Molecules. 705 



the room contained enough mercury vapour to reduce the 

 intensity of the light from the resonance lamp by nearly one 

 half as it traversed the length of the box. 



This is not so surprising when we remember that the earlier 

 experiments showed that the primary beam was reduced to 

 one half of its intensity {i.e., the frequency capable of exciting 

 resonance was) by traversing 5 mm. of the saturated vapour 

 at room temperature. The trouble was overcome b}^ opening 

 the windows and thoroughly ventilating the room before each 

 experiment. One is reminded of the trouble experienced in 

 carrying on certain investigations in laboratories which have 

 become infected by radium ! 



One of the photographs obtained with the double cell is 

 reproduced on PI. XI. tig. 3. In this case one compartment 

 was free from mercury vapour, while the other contained it. 

 This particular cell was, however, only 8 mm. in thickness, 

 and gives a good idea of the powerful absorption of the highly 

 homogeneous light from the resonance lamp by a shallow 

 layer of the vapour at room temperature. The other photo- 

 graphs showed that the energy diverted from the primary 

 beam was the same for mercury in a high vacuum and in air 

 at 3 cm. pressure, so that the calculation of the absorption- 

 scattering ratio which I gave provisionally was justified. 



It was found, however, that if air at atmospheric pressure 

 was admitted to one compartment the absorption was decreased 

 by more than one half. This is just the opposite of what I 

 found three years ago, when I discovered that I could photo- 

 graph the 2536 absorption line by passing the light of a 

 cadmium spark through a tube 3 metres long containing 

 mercury vapour at room temperature, if the tube contained air 

 at atmospheric pressure, while no trace of the line appeared if 

 the tube was exhausted. There is, however, no real dis- 

 crepancy, for the effect of the air, as I have shown previously, 

 is to broaden the absorption line. This circumstance was 

 discovered independently by Angstrom in the case of C0 2 . 

 The present experiments appear to show as well that, while 

 the line is broadened, the intensity of the absorption at the 

 centre of the line is materially reduced. I have observed the 

 same thing with iodine vapour, the lines becoming fuzzy and 

 less black when air is admitted to the tube. 



The failure to obtain the absorption line in the earlier ex- 

 periment with the three-metre tube exhausted was ot* course 

 simply due to the insufficient resolving power of the spectro- 

 graph employed. 



In the experiments with the double cell which I have just 

 described, I was unable to find any distinct evidence that 



Phil. Mag. S. G. Vol. 23. No. 137. May 1912. 3 A 



