THE THRESHOLD OF VISION FOR DIFFERENT COLOURED LIGHTS. 
107 
results were very variable, but corresponded roughly to those obtained at about 
2'5 degrees from the fovea though he was under the impression that correct fixation 
was being secured. During the course of these observations, however, he noticed that 
the light sometimes appeared to vanish, and he satisfied himself that this occurred 
whenever he managed to bring the image on the fovea. If, when using the fixation 
spot and shutter for central fixation the annulus is turned till the light has vanished, 
say in the green, if the eye is moved a little bit on one side of the fixation spot and 
the shutter opened, the flash of light seen appears almost blinding. 
In fig. 9 are collected together all the curves for 1'25 degrees from the fovea, and 
it will be noted how in the case of all observers of class I., there is a marked change 
from the values at the fovea, but they approximate much more nearly to those 
VOL. OCXVI.-A. Q 
