464 Proceedings of Boyal Society of Edinburgh, [jult 16 , 
Since the diminution of brightness is supposed to he continuous, 
the colour will he persistent. If, however, the brightness be very 
gradually diminished, this reasoning will not hold. Nor will it hold 
if the brightness be kept constant for a short time, say several 
seconds, before the colour is observed. As I understand the account 
of the observations, the colour was noticed during the change of 
brightness. Whether it was seen at the very maximum, or near 
the minimum, is not specifically stated. If it was only observed 
below and above equality, so that the star which was approaching 
its extinction was blue, and that which was approaching its maxi- 
mum was red, there is another conceivable physiological explanation 
of the bluish phase, depending on the fact that in feeble lights the 
blue preponderates. 
For the star which is increasing in brightness the after effect will 
not tell, because the after effect of a weaker stimulus will not be 
appreciable when a stronger stimulus is actually present. When 
the time of stimulation is increased say from to (fig. 6), 
the curve of red will reach the full height corresponding to the new 
time sooner than the curves of green and violet. With continuous 
increase of brightness the red will, therefore, be continually favoured, 
and the star should appear reddish. I put forward this suggestion 
with much diffidence, because I have not hitherto been able to try 
whether the differences of colour, which theoretically ought to be 
brought out by varying the intensity alone, can actually be observed. 
For monochromatic light Young and Forbes found that it 
required a greater velocity to produce equality with blue light than 
with red. If the explanation of this be a physiological one, it is 
probably connected with the facts, that the minimum difference 
which can be appreciated is not the same for each colour, that it 
further varies with the saturation, and that the degree of saturation 
of any one colour varies with the intensity of stimulation. Two 
similar white lights, which seem equal when looked at through a 
red glass, may not appear equal when viewed through a blue glass. 
Corrigenda. 
P. 444, line 6 from bottom, fur 53,000 read 53,000,000. 
„ „ 5 „ for 93,000 read 93,000,000. 
P. 445, line 6 from top, for thousand read million. 
