Studies in Visual Sensation. 



473 



The valuable and important researches of Dr. Augustus Waller,* on 

 retinal stimulation and electrical response, seem to indicate that near 

 its lower limit the curve becomes sigmoidal. The stimulus has to 

 overcome a certain amount of physiological inertia before the normal 

 sweep of the curve is established. 



For a time I believed that I had obtained evidence of such a 

 sigmoidal flexure near the origin of the curve in my experimental 

 work on shading. But further observation led me to the conclusion 

 that if it exist within the limits of my curve, the method of investiga- 

 tion is not sufficiently delicate to establish its influence. 



Apart, however, from the experimental evidence which Dr. Waller 

 adduces in support of the sigmoidal curve, and apart from the general 

 considerations which he suggests in favour of such a change of sign, 

 some such assumption seems to be well-nigh necessary if we are to 

 attempt to give a complete curve, which near the threshold of sensation 

 does not land us in the maze of difficulties arising from the asymptotic 

 character of a wholly logarithmic curve. There is nothing therefore 

 extravagant in the assumption that the origin of the completed 

 sigmoidal curve should be placed in round numbers 20 per cent, below 

 the arbitrary zero of sensation, and that this amount should be added 

 to the terms of the sensation series on the arbitrary scale in order to 

 convert it into an absolute scale of physiological response. This is 

 indicated in fig. 5, which represents the hypothetical continuation of 



the curve, on the assumption of sigmoidal curvature, beyond the limits 

 of sensory observation. The part of the graph blocked off by dotted 

 lines shows the lower part of the empirical curve, the absolute zero 



Fia. 5. 



YT] 



o 



* See ' Brain,' vol. 23, Part 1, p. 30 (1900). 



VOL. LXVIII. 



2 L 



