ON THOMAS TOUNG S THEOEY OF COLOUR- VISION. 



435 



third class, whereas the persons of the second subdivision are not more 

 numerously represented than the persons of the preceding class. 



A person of this third class on examining the spectrum finds two 

 boundary regions situated similarly as in the preceding class, and we 

 assume, as we did before, that the colour-sensations of these regions are 

 elementaiy sensations. The parts from the boundary regions to a certain 

 distance towards the middle of the spectrum we shall call the ' boundary 

 intervals,' and the remaining part between them we shall call the ' central 

 interval.' The colour-equations have shown that in each boundary in- 

 terval we must assume two elementary sensations, one which ifj the same 

 for both, whereas the other is the elementary sensation of the adjoining 

 boundary region, and similar equations have also shown that any colour 

 of the central interval is the result of the three elementaiy sensations 

 already found, that is, the sensations R, G, and V. I would like to 

 mention here that only such colour-equations could be used, in the case 

 of which the colours as to their hue and saturation could be easily matched, 

 and in whose combination the errors of observation had no great influence 

 upon the results of calculation. 



To obtain the first object whitish colours had to be avoided, and only 

 neighbouring parts of the spectrum had to be mixed ; whereas to obtain 

 the second object the component parts of the spectrum had to be at a 

 considerable distance from each other. This, of course, places the experi- 

 menter in a sort of dilemma, and many thousands of colour-equations had 

 to be produced before the proper ones were obtained. 



The continuous curves R, G, and V (fig. 4) belong to the first sub- 

 division of this third class, and the two dotted curves together with the 

 curve V belong to the 

 second subdivision. I 

 shall henceforth de- 

 note the colour-sensa- 

 tions of the first sub- 

 division as the normal, 

 and those of the second 

 as the abnormal. 



§ 4. Having accom- 

 plished the analysis of 

 colour-sensations with- 

 out the assistance of 

 any hypothesis let us 

 consider whether we * 

 can draw any inferences as to the physiological process which produces the 

 sensations of colours. Following the above-mentioned usual definition, we 

 shall call that sensation, which is caused by a simple process at the terminal 

 of the optical nerve, a ' fundamental sensation.' It is evident that for every 

 person the number of fundamental sensations is equal to the number of 

 elementary sensations, and that we can speak of ' curves of fundamental sen- 

 sation ' just the same as we did before of curves of elementary sensation. 

 We shall employ the following symbols for fundamental sensations : — 



For the first class >§ 



, f first type SOB,, .^, 



„ „ second „ [^^^^^^ \^ 2B,^^£ 



,, . , / normal 3fi, ®, 33 



„ „ tliira „ I abnormal jR', ®', S' 



P F 2 



Fig. 4. 



