1 102 VISION. 



although much importance cannot be attached to this, owing to the fact 

 that our knowledge of the rapidity of the changes in visual purple is 

 derived chiefly from observations 011 cold-blooded animals. 



The difference between the absolute and the chromatic thresholds for 

 coloured light would be due to the sensitiveness of visual purple at low 

 intensities being greater than that of the chromatic apparatus ; and it is 

 in accordance with this view that there should be no photochromatic 

 interval for red light, which does not act appreciably on visual purple. 

 Purkinje's phenomenon (p. 1078) would be due to the greater action of 

 light of short wave-length on visual purple. 



If vision with the dark-adapted eye at low intensities depends wholly 

 on visual purple, the curve of luminosity of the spectrum should corre- 

 spond with the curve of absorption of visual purple. In the absence of 

 observations on the spectrum of human visual purple in a pure condition, 

 this point is at present undecided. 



The similar spectrum of total colour-blindness leads to the assumption 

 that in this condition visual purple is the only substance present, and 

 that the totally colour-blind individual is dependent for his vision wholly 

 on this substance. The features of total colour-blindness, which have led 

 to its being regarded as a pathological condition, namely, photophobia and 

 diminished visual acuity at ordinary illumination, become quite natural 

 if vision depends only on visual purple. Further, there should be an 

 absolute central scotoma corresponding to the f ovea, and this has been in 

 several cases recorded, and may exist in all, probably varying in size in 

 different cases. The nystagmus which is a frequent symptom may be 

 consequently explained by absence of a place of most distinct vision, and 

 therefore of steady fixation. It is in relation to this anomaly that this 

 theory of the action of visual purple may be capable of objective verifica- 

 tion ; for, if true, the retina of a totally colour-blind eye might show 

 absent or ill-developed cones. There is at present no record of the 

 anatomical condition in this defect. 



The departure from the law of colour mixture with adaptation to low 

 intensities would also be ascribed to the influence of visual purple, a new 

 factor coming into play as the intensity is reduced. In order that these 

 deviations may be explained on this hypothesis of the function of visual 

 purple, it is necessary that that side of a colour equation which becomes 

 the brighter at low intensities should have the greater action on visual 

 purple (the greater rod-value) ; and, according to v. Kries, 1 this is the case. 

 To give one instance : in a match of a mixture of spectral red and green 

 with homogeneous yellow, the mixed field becomes relatively brighter on 

 diminution of intensity, owing to the especially high rod-value of the 

 green, both red and yellow having relatively very low rod-values. 



Another phenomenon, which is referred by v. Kries to visual purple, 

 is the recurrent image. He ascribes it to the greater inertia of visual 

 purple, so that this substance reacts to light appreciably later than the 

 other photochromatic substance or substances, the latent period becoming, 

 however, shorter as the intensity is increased. The fact observed by 

 both Bidwell and v. Kries, that there is no recurrent image with red 

 light, is very much in favour of this view, and so is the influence of 

 adaptation on the phenomenon. In a case of night-blindness, v. Kries 

 found that no recurrent image was perceived, although other equally 



1 Ztschr. f. Psychol. u. Physiol. d. Sinnesorg., Hamburg u. Leipzig, 1896, Bd. ix. 

 S. 81 ; and Bd. xii. S. 1. 



