540 CAPTAIN W. DE W. ABNEY AND MAJOR-GENERAL E. R. FESTING 
shows that at this point most of the spectrum has lost its colour. The faint spectrum 
when re-combined will be pale green, mixed, of course, with a fair proportion of 
grey, due to those portions of the spectrum from which the colour has disappeared. 
Fig. 37. 
Diagram of limit of colour vision. 
§ XLVIII .—Extinction of the Light of Different Parts of the Spectrum. 
A preliminary note has already been published by one of us (Captain Abney) on 
this subject. It was expected that the results then given might be liable to correction, 
as in work of this kind it is only after repeated experiment that sources of error 
can be discovered and eliminated. The errors which have since been discovered by 
comparing the results of some hundreds of observations are not serious, though their 
correction alters the extinction curve to some small extent. 
In the first experiments the readings of the scale had to be made by the observer, 
and it has been found that the light used to illuminate the scale, small though it was, 
fatigued the eye sufficiently to vitiate, to a small but sensible extent, the observations. 
This source of error has, in the latter experiments, been eliminated. 
