THE THRESHOLD OF VISION FOR DIFFERENT COLOURED LIGHTS. 
105 
The differences obtained at the fovea between observers belonging to the two 
classes could be explained without supposing any distinction as to the distribution of 
rods if we suppose that the observers belonging to class II. did not obtain correct 
foveal fixation. There is, however, strong evidence that this is not the correct 
explanation. In the first place observers A. and B. are very expert in making this 
10 20 30 40 50 60 
SSA T 
Fig. 8. • 
kind of observation. They can repeat their measures day after day even after many 
months and obtain constant values. Had there been any faulty fixation the 
measurements obtained would not remain constant but would be of the same nature 
as those illustrated in fig. 2. Further evidence that in the case of A. and B. there are 
rods at the fovea will be given later. 
cD 
