254 OBSERVATIONS ON VISION. 



VIT. OCULAR PHANTOMS. 



The hallucinations of rod vision are best illustrated when a small 

 surface is observed whose retinal image does not exceed the area of 

 most distinct vision and Avhose gradual increase of brightness can be 

 noted, beginning Avitli darkness. T employ the following apparatus: 

 An electrical glow lamp of (>5-volt type is placed in a box with a sin- 

 gle aperture 15 mm. in diameter, and connected to a 110-volt circuit 

 whose current may be gradually varied from zero upward. Behind 

 the opening is })laced a dark shutter, and in front of this a movable 

 brass diaphragm with diiferent holes of 3, (>, and nun. diameters. 

 An enlarged, distinct image of the sharply bounded aperture is then 

 cast on a white screen, and there are placed in the path of the beam 

 several sheets of gelatin stained blue-green, so that the rods are more 

 strongly iuipressed than the cones. 



So long as (h(> brightness is so feeble that the cones are excluded, 

 the light s])()t, seen obliquely, ap})ears colorless and lacking sharp 

 outlines; but it quite disappeai's when gazed at fixedly. This dis- 

 appearance occurs even when the light spot is so large that its image 

 covers the whole of the macula lutea, as shown by the introduction 

 of the larger dia])hragm. To be sure, the disappearance is more 

 difficult to produce, because the slightest nu)venient of the eyes then 

 renders the spot visible, and the waiulering eye involuntarily takes the 

 position Avhere it receives the most light. Hence it happens that the 

 spot floats in and out of the vision, for as soon as we attempt to fix 

 distinctly what we saw indirectly the " ghost " disapj^ears." But 

 when tlie light is made so bright that the cones are impressed this 

 condition of things ceases, and we see the light spot distinctly and at 

 rest with its shar]) outlines and its l)lue-green color. The experiment 

 is less successful when i)erfornied with red stained gelatin, but even 

 in this case the sj)()t may be catised to disapi)ear by proper reduction 

 of the illumination, as in ExiH'i-iment TTl, whei-e the dark-red glow- 

 ing lamp filament becomes invisible Wy direct vision, while yielding 

 the gray-glow appearance with obli(]ue vision. 



I). Kxi'KKniKNTs wnii IbiKiirr Spkciha. 



1 I I.vimllicsis ;is to tilt' ciiusc of color-hliiHliicss. | 



Discriuiiiiiit ion between the rods and cones and actjuaintance with 

 their ditl'erent functions has not yet furnished the answer to the ques- 

 tion why the cones differentiate colors, or what means they emplo_y, 

 ;ind why the rods gi-ow more sensitive in ()V)scui-ity. It is true that 

 th(> cones are destitute of the visual pur])le. so that |)er1ia])s this plays 

 a part in " adaptation to obscurity.'"' 



« See O. Lmniuer, Roilr.iu zur Kliiiuiijx der nonesten Versuclio von R. Bloncllot 

 liber tlie n-rays. Vci'li. der Dontsclicii IMiys. (J('S(>l]s('liiift, f), 418, 1!)0?>. 



