1888 .] Mr G. N. Stewart on Intermittent Light. 
461 
the greenish and reddish phases were not so well marked as in the 
other experiments. This might be because the luminous intensity 
was less ; or the relative amount of excitation produced by sodium 
light in the red and green fibres may be such that the curves are 
more nearly parallel than with white light. If the difference were 
small with the greater intensity, it would be still smaller with the 
less, and this would explain why no difference of colour was got 
with the flame 4 feet away. There were no changes seen at the 
edges, probably because the light was not intense enough to show 
differences of phase in its different parts. The blurred reddish 
appearance could be seen with the mirror at rest, and was apparently 
a diffraction halo. 
Summary of Conclusions. — When the retina is stimulated by 
a succession of very short flashes of white light, the propor- 
tion between the amount of excitation in the three hypothetical 
groups of fibres is not constant. W ith a certain duration of each 
stimulus, the excitation in the violet group preponderates ; with a 
shorter duration, that in the green preponderates ; with a still 
shorter duration, that in the red. This is explained by supposing 
that the curves of excitation have some such course as that repre- 
sented in fig. 4, the curve of red rising at first more steeply and 
then less steeply than those of green and violet, while the steepest 
portion of the violet curve falls later than that of the green. 
The more intense the stimulus is, the shorter must be its duration 
for any given phase. 
Colour Phenomena observed hy Young and Forbes. 
It occurred to me that the appearances observed by Young and 
Forbes, in their experiments on the velocity of light {Phil. Trans. ^ 
vol. clxxiii., 1882, p. 273, &c.), were possibly analogous in nature to 
those above described. The method employed was a modification 
of Fizeau’s. It is unnecessary to describe the details. The source 
of light is placed behind a toothed wheel, which can be rotated at a 
very rapid rate. The ray passes out between two teeth, and is re- 
flected from two distant mirrors placed nearly in the same line, but 
one more remote than the other. When the toothed wheel is at rest 
the observer sees two stars side by side, and of approximately equal 
brightness. When the wheel is in rapid rotation, the same thing 
