( 296 ) 



the last values, which hold for the faintest magnitudes, show a 

 decrease in the colour figures with regard to the preceding ones. 



This phenomenon may be accounted for by the existence of the 

 colourless perception of faint sources of light. In faint stars we do 

 not see any colour at all ; there the perception of colour disappears 

 almost entirely and there remains only a colourless (i. e. Avhitish) 

 impression of light. With stars which approach this limit, the 

 impression of colour will be mixed up to a high degree with the 

 colourless impression, and therefore they appear paler and will be 

 indicated by a lower figure. As for the redder stars this colourless 

 impression is relatively much weaker, the paleness of colour for 

 these stars occurs only with a much less degree of brightness; in 

 this manner we explain why the b^^^ and 6''^ groups do not show 

 this decrease. Whether in these cases the phenomenon occurs with 

 fainter stars cannot be decided because Maury's spectral catalogue 

 does not contain fainter stars. 



For the practical purpose of reducing the observed colours to one 

 brightness it is about the same which of the two relations is 

 adopted, as long as we keep within certain limits of brightness, for 

 instance between the magnitudes 1 and 5. To facilitate the reduction 

 we have made use of the linear formula given above for the 5 first 

 groups (down to class XV included) while for the redder classes 

 0,20 has been adopted as the coefficient of brightness. 



To explain the long known phenomenon that the colour deepens 

 with decreasing brightness as is shown in the tables on p. 

 Helmholtz in his Physiologisclie Optik has given a theory called 

 "Theorie der kürzesten Linien im Farbensystem". In the diagram 

 of colours in space, where each impression of light is represented 

 by a point of which the 3 coordinates represent the quantities of the 

 elemental colours, red, green, blue, the lines of equal colours are 

 not straight radii through the origin, but curved lines which with 

 increasing distance from the origin bend more and more towards 

 the axes and so diverge more and more from one radius which is 

 straight and represents the "Principalfarbe". Hence in the triangle 

 of colours the points of equal colour diverge the more from the 

 principal colour and run in curved lines towards the sides and the 

 vertices as the triangle of colours is removed farther from the origin, 

 and thus represents a greater brightness. Helmholtz gives as principal 

 colour a certain "yellow-white" to which with extremely great 

 intensity all colours seem to approach. Therefore colours which 

 lie on the blue side of this principal colour must become bluer 

 by fading. 



