413 M. H. Helmholtz on Sir David Brewster's 



is, however, indicated by the circumstance, that the limit of the 

 green extends much further when the eye is permitted to wander 

 over the different colours of the spectrum than when it is per- 

 sistently fixed upon the green portion. In the first case, the 

 yellow colours strike the retina on places which before were acted 

 upon by the shining red, and therefore tend to generate the com- 

 plementary blue-green ; in the second case, the excitation of the 

 subjective colour upon the contiguous portions of the retina is 

 much feebler. That the pha^nomenon is due to a subjective 

 illusion is immediately shown when the colours are isolated 

 according to the method which I have above described, and then 

 looked at through layers of the above-named brown bodies of 

 different thicknesses ; they then appear totally unchanged, and 

 without the slightest tendency to green. 



Looked at through thicker layers of the brown fluids, green, 

 yellow, and a portion of the orange disappear from the spectrum. 

 At the edge of the red which remains, a weak rim of green is 

 observed, even near the line C, where the red has scarcely the 

 appearance of orange. The green rim is too weakly luminous 

 and narrow to permit of its light being isolated and singly ex- 

 amined. That, however, weak orange light beside strong red 

 may appear green, is easily shown by sticking a small disc of 

 paper coloured red by vermilion upon a plate of red glass, and 

 holding the latter against a very bright ground, the bright firma- 

 ment for example, while the disc is only weakly illuminated. 

 With a suitable strength of illumination it appears green*. 



It further appears to me, that the violet colouring of the blue 

 to the vicinity of the line F, through absorption by yellow fluids, 

 olive oil, sap of the Coreopsis tinctoria, &c., belongs to these sub- 

 jective complementary pheenomena. I have repeated the expe- 

 riment with olive oil, and have plainly seen the violet between 

 the lines F and G nearly as far as F, but only when this portion 

 of the spectrum was very feebly illuminated. 



The oil does not sensibly change the brightness of the red, 

 yellow and green ; it weakens the blue considerably, and almost 

 extinguishes the violet. When I permitted the light of bright 

 clouds to enter through the slit, the first-named colours appeared 

 bright, the blue feeble, and the violet was not at all to be seen. 

 When, however, direct sunlight was passed through the slit, the 

 portion between the lines F and G appeared brighter and lost 

 its violet appearance. When isolated from the other colours of the 

 spectrum in the manner before indicated, the blue appears in its 



* The subjective colour is very strikingly exhibited in this experiment. 

 A red wafer answers the purpose perfectly ; the observer stands in a weakly 

 illuminated place and looks at the sky, the wafer appears a vivid green or 

 blue-green. — J. T. 



