RELATIONSHIP BETWEEN STIMULUS AND SENSATION 563 



of the same order as that of intensity given above, namely, that a differ- 

 ence in the amount of white light diluting a colour by y^th of the total 

 intensity present can be just appreciated by the eye. 



Difference Threshold of Size 



If two objects are the same distance from the eye, and are close to one 

 another and in similar positions, a difference of one- hundredth the mean size 

 can as a rule be appreciated. If, they are at different distances from 

 the eye, or are far apart, or are not in similar position (e.g. one perpendicular 

 and the other horizontal ), then considerable errors may occur. 



THE METHOD OF COMPARISONS 



In the application of this method three different series of investigations 

 have been carried out. (1) To determine the intensities of three primary 

 colours which when mixed together will match the different spectral colours 

 or white light. (2) To determine the intensities and wavelengths of the 

 complementary colours. (3) To ascertain the intensities and wavelengths of 

 red and green rays which when mixed together match a pure spectral yellow. 



The colour box in some form or other is used for these tests. Abney's 

 apparatus may be described as a typical example. Light from an arc lamp 

 is f ocussed on to the slit of a powerful spectroscope, which consists of a colli- 

 mator, a train of prisms and a telescope. The spectrum thus produced is 

 caused to fall on three slits, one of which corresponds with the red, another 

 with the green, and the third with the blue. The light having passed through 

 the slits falls on a lens which forms an image of the prism faces on a screen. 

 The light from the slits thus recombines on the screen to produce a bright 

 patch, the colour of which alters according to the intensities of the three 

 components. To one side of the patch a second patch of light can be thrown 

 from the arc lamp, and this also may be varied in different ways according 

 to the nature of the experiment. The intensities of the different beams could 

 be modified by altering the widths of the slits ; a preferable method is 

 to employ rotating sectors, the angles between the blades of which can 

 be varied at will (see Fig. 290). 



COLOUR MIXING EXPERIMENTS performed with this apparatus 

 give results that have already been briefly considered in Section I. It is 

 found that, not only do the three primary colours when mixed together in the 

 right proportions form a white light that is indistinguishable from ordinary 

 white light, but they can also be made to match the whole range 

 of colours both of the spectrum and of pigments. It was also described how 

 that certain pairs of colours when mixed in the right proportions are able 

 to match white, and that these pairs are called complementary colours. 

 If the colours that are mixed are further apart than are the complementary 

 colours, then the mixed colour is found to be a shade of purple ; but if nearer 

 together than the complementaries, then the colour formed by the mixture 

 corresponds to an intermediate part of the spectrum. Thus, if the colours 

 mixed are red and green, the intermediate yellow and orange portions of the 



