204 Dr. G. J. Burch. me at [ Mar. 7 
9 
the current. For the same reason I used paper instead of sodium carbonate, 
thus getting rid, by fluorescence, of a great part of the violet rays. 
Fia. 2. 

The experiment consisted in turning the polarising prism D until the blue- 
violet band V, and the red band R, appeared equally bright, and then altering 
the intensity of the light by placing the glow-lamp A nearer to or farther 
from the paper reflector B, and finally, if any change was observed, readjusting 
the polarising prism until V, and R, were again equally bright. From the 
data thus obtained, the relative intensities of the colour-sensations under the 
different degrees of illumination could be calculated. 
Here some explanation is necessary. It may be objected that there can be 
no definite standard of equality of brightness between two different colours. 
That is perfectly true in the sense that not only does the position of the 
prism vary when it it is set by different observers, but the same observer, 
under different conditions of light adaptation, makes very different settings. 
What I here mean by “ equality of brightness” between two colours. is this: 
an observer sets the prism so that the two colours are, in his opinion, equally 
bright. The prism is then displaced by someone else, and the observer asked 
to repeat the setting immediately. He will probably do so with not more 
than 2 per cent. variation in the relative intensity, and generally much less. 
His standard of equality is personal and empirical, but it is constant so long 
as his physiological condition is unchanged. 
