﻿132 



Mr. J. S. Dow on Colour 



Nature of Lights 

 compared. 



Eye 20 cms. 

 from aperture. 



Eye 60 cms. 

 from aperture. 



Percentage 

 change. 



Ruby-Red to Signal- 

 Green. 



™ =1-52 

 ureen 



Red 

 G7een =1 ' 70 



12 



Ruby-Red to Light 

 from unscreened 

 glow-larap(white). 



Methven Gas Stan- 

 dard to Incandes- 

 cent Mantle. 



Eerl 

 Whit-e= ' 70 



Red nnz 

 White" "' 5 



6 



Metlw._ . 2s 

 Inct. 



f et f bV '=029 

 Inct. 



3 



The effect is very noticeable when observed in the follow- 

 ing way: — Supposing we are comparing red and green, and 

 have placed the photometer so as to secure balance, Move 

 the photometer until a distinct flicker can just be seen owing 

 to the green being too bright. It will be found that as the 

 eye is removed the flicker gradually disappears. But if the 

 photometer is put out of balance on the red side, the flicker 

 does not disappear but becomes if anything more distinct as 

 the eye is removed. 



It appears therefore that this Flicker photometer is affected 

 by the distance of the eye, but, apparently, not to a sufficient 

 extent to be noticeable when the telescope is used,, as in 

 ordinary work. 



It is difficult to see why these effects should be so much 

 less noticeable than with ordinary photometers. It almost 

 seems as though, when two differently coloured objects are 

 placed side by side, any change in their relative illumination 

 becomes exaggerated, thus creating a different impression to 

 that received when they are viewed alternately (as in a 

 Flicker photometer). 



An attempt was also made to discover whether the Flicker 

 photometer was subject to the Purkinje phenomenon. Now 

 it is well known that the speed required to just make the 

 flicker disappear depends upon the illumination, and the writer 

 has found it impossible to judge the point of disappearance 

 of the flicker with any certainty at the extremely weak 

 illuminations necessary to produce the Purkinje effect. 



The plan of " colour-reading " referred to above was there- 

 fore adopted. It may be objected that in doing so the real 

 question at issue is avoided altogether. But the method, 

 while admittedly not so satisfactory as the " disappearance of 

 flicker " method from this point of view, has been shown to 

 give the same results at ordinary illuminations and, to the 



