314 



SCIENCE. 



[N. S. Vol. VIII. No. 193. 



sensations dependent upon tliem, the phe- 

 nomena might probably be explained in an 

 indefinite number of ways, and the differ- 

 ent methods of explanation should be re- 

 garded rather as examples of ingenious 

 speculation than as real contributions to 

 the advancement of science. 



To such a category belong many of the 

 later theories of vision. They incline to 

 Helmholtz or to Hering according as their 

 point of view is chiefly physical or psycho- 

 logical, for the standpoint of these two theo- 

 ries is fundamentally different. 



Helmholtz, showing that all colors can be 

 compounded from three, and that white 

 may be also compounded, assumes that 

 three color-sensations are sufficient, and 

 that white may be regarded as a compound 

 sensation. Hering, relying more upon the 

 direct deliverances of consciousness, denies 

 the compound nature of the sensations of 

 white and of yellow, whatever their phys- 

 ical composition may be, and says explicitly 

 that ' ' the entire separation of the optical 

 nature of a light from the sensation which 

 it arouses in us is one of the most neces- 

 sary prerequisites to a clear handling of the 

 theory of color." Along the lines of these 

 two theories, then, new hypotheses move, 

 and will move, since each of them stands 

 for something real, and has its own distinct 

 advantages. 



Upon a somewhat dififerent basis rests a 

 theory, hardly so much of color as of light- 

 sensation, which was hinted at by various 

 observers, but most clearly worked out by 

 von Kries. This supposes that we possess 

 two entirely distinct kinds of visual appa- 

 ratus, one dependent upon the cones of the 

 retina, the other upon the rods, and the 

 visual purple connected with them. 



Max Schultze, so long ago as 1866, mainly 

 on anatomical grounds, suggested that the 

 rods were probably the most important 

 organs of vision in faint light. Animals 

 which prey by night, as cats, moles, owls, 



etc. , possess retinas rich in rods, but with 

 cones either few or absent. Our own eyes 

 perceive faint light more readily with the 

 peripheral portions of the retina, where rods 

 are numerous, than with the central por- 

 tions, where they are few. 



Helmholtz * pointed out the fact that if 

 the visual purple is actually connected with 

 vision it must have to do with peripheral 

 rather than central vision, since it is ab- 

 sent from the fovea, and suggested that it 

 might have to do with the perception of 

 faint light. 



In 1894 Konig studied the absorption 

 curve of the visual purple, finding it sub- 

 stantially identical with the curve of bright- 

 ness for the spectrum of low luminosity, 

 von Kries, combining these and other sug- 

 gestions, considers the visual purple in the 

 rods to be, in the human eye at least, the 

 active agent for the perception of faint light. 

 He shows that the phenomena of adaptation 

 point in the same direction. In strong light 

 the visual purple is soon bleached. An eye 

 ' adapted for brightness ' is very deficient in 

 power to perceive faint light. If it is now 

 kept in darkness for about half an hour this 

 faculty is enormously increased. But in 

 about the same period the visual purple is 

 practically restored. The essence of adap- 

 tation is the recovery of the visual purple. 

 Eed light, which does not act upon this 

 substance, does not destroy the sensitive- 

 ness to faint light in an eye which has 

 been exposed to it for even a considerable 

 time. 



If vision by faint light depends, wholly 

 or partly, on the decomposition of the visual 

 purple, and if light of long wave-lengths 

 does not effect this decomposition, blue light 

 when faint should appear much brighter 

 than i-ed, and Purkinje's phenomenon is 

 thus easily explained. But in the fovea, 

 where the rods and the purple are not pres- 

 ent, this sensation of colorless faint light 

 * Physiol. Optik, 2d. ed., p. 268. 



