n. I CAUSE : EFFECT 



Let us consider the three cases : 



y y y 



37 



FlG. 17. Straight. 



The effect Y varies as its ' 

 cause X. 



Equal increments of cause 

 produce equal increments of 

 effect. The cause /effect curve 

 is a straight line. 



Concave. 



The effect Y varies more 

 rapidly than its cause X. 



Equal increments of cause 

 produce increasing incre- 

 ment of effect. The cause / 

 effect curve is convex to its 

 abscissa OX. 



Convex. 



The effect Y varies less 

 rapidly than its cause X. 



Equal increments of cause 

 produce diminishing increments 

 of effect. The cause / effect 

 curve is concave to its abscissa 

 OX. 



And now turn to the case in point, where the cause X is the 

 intensity of a light, and the effect Y the magnitude of a galvano- 

 metric deflection (which we shall assume to measure magnitude 

 of electrical change, and therefore magnitude of retinal change). 

 With a light of suitable strength, varied on an arithmetic 

 scale constructed in conformity with the law that luminosity 

 varies inversely as the distance squared, we find the following 

 series of results from the given series of stimuli : 



Stimulation by light . 2 4 6 8 loX 

 Retinal change . . 208 248 274 296 326 Y 



which, plotted as described above, gives the curve figured below 

 as that of the retinal effects of medium illumination. 



^O 4O 



fa eo 100 



Strong. 



FlG. 1 8. Frog. Retinal response to illumination of constant duration and of 

 varying strength. The unit of light is a standard candle at a distance of lo feet from 

 the eyeball. 



Now this curve, concave to its abscissa OX, with the effect 

 Y increasing less rapidly than its excitant cause X, is in 



