﻿Phenomena in Photometry. 131 



the results with the Joly photometer and the Flicker photo- 

 meter were exactly the same. The two lamps were now 

 screened with red and green glass respectively; and the ratio 

 of the red to the green, obtained by the Flicker photoineter, 

 was about 1*34 ; but with the Joly photometer it was found 

 possible to get readings from 0'7 to 1*4 for distances of the 

 eye up to half a metre. Agreement between the photometers 

 occurred when the eye was about 40 centimetres away from 

 the blocks of the Joly. 



One difficulty met with in this investigation was as follows. 

 As the effects observed were comparatively small, it was 

 advisable to use lights widely differing in colour. On the 

 other hand the position of minimum flicker, though sufficiently 

 sharply defined in most practical cases, is far from being so 

 when such colours as red and green are observed. In such 

 cases, the method followed was to note the position of the 

 photometer in which a flicker was just visible on either side 

 of balance, and take the mean. There is, however, another 

 method of judging the position of balance for these two 

 colours. 



When the photoineter is too near the red light, the field of 

 view in the photometer appears reddish in tint. (Similarly, a 

 greenish tinge shows that the green illumination is the 

 stronger. When the illumination of the two surfaces, illumi- 

 nated by light of these two complementary colours, is the 

 same, an intermediate greyish tinge is produced. To the 

 writer's eye the transition from red to grey to green was 

 sharper than the disappearance of the flicker, and a series of 

 tests showed that the result was the same in each case. The 

 method, however, is only applicable to complementary 

 colours. 



Some experiments were first made to discover whether a 

 difference of reading could be produced by altering the tele- 

 scope, as in the case of the Lummer-Brodhun photometer, 

 but no distinct difference was observable. Next the telescope 

 was removed, and readings were taken with the eye about 

 20 centimetres from the aperture. A brass tube was then 

 inserted which allowed the aperture to be inspected from 

 a distance of 60 centimetres from the eye. It was then 

 observed that, as the eye was withdrawn, the field became 

 distinctly redder, and readings taken by the " disappearance 

 of flicker " method also showed a change of relative candle- 

 power in favour of the red. The following table exhibits 

 some of the results obtained : — 



K 2 



