268 CAPTAIN W. DE W. ABNEY ON THE COLOUR 



By taking colours on each side of the blue lithium line we find that the proportions 

 of blue and red sensations are unchanged, and always fulfil the above equation. 



Having found the percentage composition of all the spectrum in terms of the red 

 and green sensations and of violet, the last is converted into blue sensation and red 

 sensation. The red sensation existing iu the violet is then added to that already 

 found. 



(XII.) Difficulties in making the Observations. 



The description of the nature of the observations may make it appear that they 

 are simple, but the reverse is the case. The labour involved is very great, and the 

 difficulty soon becomes apparent when the work has fairly started. The sensitiveness 

 of the eye to colour varies considerably, and this in itself makes observations hard. 

 On some mornings, when coming fresh to the laboratory, the comparisons are readily 

 made, but those made in the evenings after a day's work are often wild at first, and 

 much more time has to be spent in perfecting them than may be supposed. Before 

 any match can be considered worthy of recording, the eye has to be withdrawn from 

 the light and to look into darkness for a minute at least, when a rapid glance will show 

 if it needs alteration. If not correct, the slits have to be opened or closed, as may be 

 required, and again a rest given to the eye. This procedure may be repeated several 

 times before the match is considered satisfactory. The fatigue of the retina has a 

 good deal to say to the difficulties encountered. 



(XIII.) Order of the Observations. 



It may be as well to record the order in which the observations were made. The 

 first are preliminary, and are as follows : 



(1.) The position of the spectrum in regard to the slit-holder is determined. 



(2.) The scales at the back and front of the slit-holder are compared. 



(3.) The lens with which the apertures of the slits are measured is adjusted. 



The second are those taken for recorded observations : 



(1.) The slits are placed in position. 



(2.) The matches are made. 



(3.) The luminosities of the light coming through the slits are measured and the 



apertures of the sectors noted. 

 (4.) The widths of the slits are measured. 



2 and 4 had to be repeated several times in each series of observations. I have 

 already shown, in " Colour Measurement and Mixture," that a certain percentage of 

 coloured light can be hidden in white without being perceived. In making a match 

 with the white, each slit had to be opened in turn till it was evident that an excess 



