﻿130 Mr. J. S. Dow on Colour 



The mean of one set of determinations gave : — 



Coefficient of reflexion, 78*8 per cent, for unscreened lamps. 

 79-0 „ „ red light 



77*8 ,, „ green light. 



Repetitions of the experiment gave slightly differing results, 

 but it was found that the values for the three different colours 

 did not differ among themselves by more than 2*5 per cent. 

 Moreover, no connexion could be traced between these 

 differences and the colours, for the difference was sometimes 

 in favour of the red light, and sometimes in favour of the 

 green. 



We may safely conclude therefore, that in all ordinary cases, 

 where the difference in colour is less pronounced than that 

 adopted here, the effect is inappreciable. 



In any case, the adoption of adjustable photometers of the 

 Simmance-Abady type will probably render mirrors un- 

 necessary in obtaining curves of distribution of light, in the 

 future. 



The only really important effect, under ordinary working 

 conditions, therefore, is that described under the second 

 heading. It has been shown that this may, apparently, give 

 rise to a perceptible discrepancy even when lights so similar 

 in colour as the Harcourt 10-candle Petane standard and the 

 Fleming glow-lamp standard are compared. 



Flicker Photometers. 



The interesting question now arises whether photometers 

 of the Flicker type are also influenced by these colour- 

 phenomena. 



If, as is claimed, the disappearance indicates that the two 

 surfaces appear equally illuminated to the eye, one would 

 suppose that all the conditions which determine this impression 

 must also determine the point of disappearance of the flicker. 

 On the other hand Mr. T. C. Porter, and others, have come 

 to the conclusion that the flicker, at ordinary illuminations, 

 is independent of colour. 



Through the kindness of Messrs. Everett & Edgcombe the 

 author has been able to make some experiments on the points 

 referred to, with a photometer of this kind. 



It is clear, in the first place, that when comparing lights of 

 different colour, the readings of an ordinary photometer and 

 a Flicker photometer may not agree even if the latter be 

 independent of all colour effects. For instance, when two 

 lamps giving light of exactly the same colour were compared, 



