out, and the remainder are combined in the resulting sensation. But inasmuch as the optimum period of the cycle is different for each colour-sensation, the use of pigment colours leads sometimes to apparently divergent results, which, however, are easily accounted for when the spectra of the pigments are taken into account.
In manipulating the apparatus when both flashes are monochromatic a farther clue to the explanation is obtained. To produce the maxi- mum Exner-effect, the second stimulus must be much less intense than the first, though it may last longer. This accounts for the difference of the appearance when the direction of the rotation is reversed, since, unless the relative intensity and duration as well as the colours of the flashes are interchanged, the conditions under which the phenomenon occurs are no longer observed.
I fail to see why an independent white sensation must be postulated. If the first stimulus is green and the second white, then, on Young's hypothesis, the second stimulus consists of red, green, violet (and blue according to my own experiments). Accordingly, the green, which is common to both stimuli, vanishes, and the eye perceives that mixture of red, blue, and violet so familiar in purple pigments and flowers. It appears to me that Dr. Shelford Bidwell's black-spot experiment adds farther confirmation to this explanation, since it shows con- clusively that the response to the first stimulus is inhibited if that word may be borrowed by a second stimulus falling upon a neighbour- ing part of the retina. For the second stimulus, white, must neces- sarily include the same physical stimulus, green, as the first, and this component of it is by itself, as my experiments show, quite competent to produce the observed result, namely, the inhibition of the response to the first stimulus. Moreover, a second stimulus which contains all the elements of white save green does not inhibit the response to green, and the same is true of the other colours. If the first stimulus is red and the second blue, or if the first is green and the second purple, or vice versd, i.e., if neither stimulus includes any of the com- ponents of the other, the effect is nil. It is difficult to see why this should be the case if the phenomenon is a function of a white sensa- tion. To me it seems more simple to explain the whole phenomenon as follows :
It is granted that a diminution of the intensity of a sensation occurs whenever the stimulus is prolonged. This diminution may occur so suddenly under suitable conditions that it becomes difficult to regard it as the mere result of using up material previously stored. It resembles rather those reflexes by which the organs of vision are protected from a sudden light, namely, the winking of the eyelids and the contraction of the iris, and is probably a provision for preventing unnecessary waste of material.
The question then arises whether such a function would be located
