Principles underlying the Flicker Photometer. 67 



Dicussion of Results shown in Figs. 3 and 4. 



Let us now examine the results described in the two pre- 

 vious sections in greater detail. The main result of both 

 experiments is that the Purkinje and Yellow-spot phenomena 

 seem to be very much weaker when the flicker method is 

 used. The flicker curve in fig. 3, in fact, seems to resemble 

 that which is obtained by using an ordinary equality of 

 brightness instrument under such conditions that the effective 

 value of the rod-organs has become relativelv insignificant 

 in comparison with that of the cones. And it will be noted 

 that, just as the curves in fig. 1 seem to approach a limiting- 

 shape in which a certain variation in the results with altering 

 retinal area still occurs, even at high illuminations, so, in the 

 case of the flicker photometer, some variation, as the eye is 

 withdrawn from the photometer, seems to be suggested. 



In both the experiments described above, therefore, the 

 effect which seems so well explained by the struggle for 

 predominance between the rods and the cones, is apparently 

 much less marked in the case of the flicker photometer. The 

 natural inference to draw would therefore be that one of the 

 sets of competing organs is in this case less influential; and, 

 for reasons which will be more fully explained shortly, the 

 author is inclined to suggest that in the case of the flicker 

 photometer the rods on the retina may be less influential in 

 determining the point of balance. In an extreme case we 

 might even suppose that the position of balance of an equality 

 of brightness instrument is determined by combined rod and 

 cone vision, while in the case of a flicker instrument it might 

 be affected by cone-vision only. 



The rods, as we have seen, are believed to differ essentially 

 from the cones as regards light and colour-perception. Now 

 the author is inclined to suggest that they may also be credited 

 with a difference in the time taken to record and preserve 

 a luminous impression. If, for instance, it were found that 

 the speed of the flicker photometer, as ordinarily operated, 

 was adjusted to the persistence of vision of the cones, but 

 was too high for that of the rods, the above phenomena would 

 be fairly well accounted for. Under these circumstances, we 

 should obtain a flicker-sensation due to the cones and, super- 

 imposed over this, an impression of luminosity only, due to 

 the fusion of rod-impulses, which might be supposed to 

 succeed each other too rapidly to produce any flicker 

 sensation. The rod-impulses therefore might not, under 

 these circumstances, materially affect the place of balance 

 of the photometer. 



F2 



