E. B. Ferry — Persistence of Vision. 197 



maining for a long time in a dark room. And it was also 

 found that the sensitiveness of the eye might have changed 

 by several per cent without producing any noticeable differ- 

 ence in the duration of the retinal impression. This shows 

 two things : first, that the personal equation, or more properly 

 the personal error, in this sort of investigation is less formid- 

 able than ordinarily supposed ; and secondly that a compara- 

 tively large change of sensitiveness of the eye is required to 

 produce a marked change in the duration of the retinal 

 impression. 



An examination of this curve (8, fig. 1) shows that the 

 retinal persistence is very different for different parts of the 

 spectrum. As in the curves published by Dr. Nichols, in the 

 paper already cited, the minimum duration is near the D line, 

 and from this point the duration steadily increases toward each 

 end of the spectrum. The observations were carried relatively 

 farther into the blue than in the red, which largely accounts 

 for the apparent unsymmetrical form of the curve. The curve 

 is of the general form of a parabola with its apex approxi- 

 mately at the D line and the two branches becoming parallel 

 to the ordinates of the ends of the spectrum. 



Duration of Retinal Impression for light of different Intensities. 



In the early part of this century Plateau* noticed that there 

 was an intimate connection between duration of retinal im- 

 pressions and the intensity of the light producing them. One 

 of the principal objects of this investigation was to determine 

 the law connecting these quantities. To do this, values of 

 duration of retinal impression were obtained for monochro- 

 matic light of different intensities. The light intensity was 

 varied by changing the width of the collimator slit according 

 to Yierordt's method. The plan followed was to obtain a 

 spectrum of a certain brightness, and measure the duration of 

 retinal impression at sufficient points in it to be able to plat a 

 curve showing the relation between the retinal persistence and 

 wave-length. Then changing the width of the collimator slit 

 by a definite amount, so as to obtain a spectrum of a brightness 

 in known proportion to that of the preceding spectrum, to 

 determine the duration of impression for the same points as 

 before. In this way the following values were obtained for 

 the duration of retinal impression of monochromatic light of 

 different color and different intensity. 



* Dissertation sur quelque proprietes des impressions produites par la lumiere 

 sur l'organs de la vue. (Liege 1829.) 



