﻿696 Prof. R. W.Wood on Selective Reflexion, Scattering 



molecule. It appeared to ms that the most direct way of 

 investigating this question was to take the intensity of the 

 cone of light as the measure of the intensity of the primary 

 beam, for there appears to be no doubt but that the intensity 

 of the resonance radiation is proportional to the intensity of 

 that particular frequency in the exciting light which is 

 capable of setting up resonance. I accordingly made a very 

 large number of photographs with different times of ex- 

 posure and different vapour densities, and measured the 

 photographic density of the image at different distances from 

 the point where the light entered the vapour. This gives us 

 a measure of the rate at which the vapour cuts down the 

 amplitude of the exciting frequency as the wave moves 

 through the medium. The method by which these measure- 

 ments were made w T ill be described presently, and for the 

 present 1 need only point out that the usual precautions 

 necessary in photometric work by photography were token. 

 The mercury vapour in this case was not contained in the 

 quartz tube, for it was impossible to get a good image of 

 the excited region close up to the point at which the beam 

 entered on account of the irregular refraction by the wall of 

 the tube. A rectangular box of brass with windows of 

 crystal quartz plates, which I shall describe presently, was 

 used, and by paying attention to adjustments, it was possible 

 to get beautifully sharp images of the excited region close 

 up to the surface of the window through which the light 

 entered. I made measurements on fully a dozen plates 

 taken on different days, and found that the intensity of the 

 primary beam was reduced to one-half of its value after 

 travelling for a distance of 5 mm. in mercury vapour at a 

 pressure of 0*001 mm. The smallest value found was 4 mm. 

 and the largest 6 mm. for a reduction to half intensity, a 

 variation due without doubt to the circumstance that the 

 room temperature varied by several degrees from day to day. 

 This variation in the temperature of the room was just about 

 sufficient to account for the slight differences observed, and 

 w r e can safely assign the value of 5 mm. for a temperature of 

 22° G. Measuring the intensity along the path of the beam 

 it was found that the ordinary law of absorption was very 

 closely followed, i. e. a 10 mm. layer reduced the intensity 

 to 1/4, and a 15 mm. one reduced it to 1/8 of its original 

 value. It appeared to me, however, that the reduction of 

 the intensity did not follow this law exactly, but that the 

 intensity was a trifle greater after 15 mm. had been traversed 

 than the intensity calculated on the assumption that every 



