40 Proceedings of the Royal Society 
towards B, a point h is reached at which the lamp rays of B begin 
to shorten, £.e., the light of A affects the iris. Getting nearer still 
to A, a point V is reached, where the distant light B appears to have 
lost its rays. 
The average of a dozen experiments gave as the value of bA... 
14 yards, and as the value of V A... 8 yards. Squaring these 
numbers, it appears that about one-third of the light competent to 
contract the pupil very markedly is sufficient to start its movement. 
At present, I abstain from comment, as I am continuing these 
experiments. 
2. On a New Variety of Ocular Spectrum. By 
John Aitken. 
If we look for a short time at any object, and afterwards turn 
the eye in another direction, we see a spectral image of the form of 
the object first looked at. 
Again, if after we have looked at any coloured object we turn the 
eye in another direction, we see a spectral image of a colour compli- 
mentary to that first looked at. 
In addition to these spectral forms and colours, I find there is 
another and distinct kind of ocular spectrum, which we may call a 
motion spectrum. It is seen when we look first at a body in motion 
and afterwards direct the eye towards an object at rest. The object 
at rest, when seen under these conditions, seems to be in motion, 
and the direction of its apparent motion is the opposite of that of 
the moving body first looked at. 
I first observed this motion spectrum when looking at the surface 
of a river where it was flowing rapidly, the eye being afterwards 
directed to a gravel bank. The first effect seemed to be an indis- 
tinctness of vision, but, on carefully repeating the experiment, I 
was much astonished to observe that, after looking steadily at the 
stream and then at the gravel bank, a narrow spectral stream of 
gravel seemed to flow steadily through the middle of the gravel 
bank, the direction of its motion being the opposite of that of the 
