204 SOMF CURIOSITIES OF VISION. 



which they assume are rather uncertain and difficult to specify. By 

 far the most striking of the colors exhibited by the top is the red, 

 and next to that the blue. This latter, however, is sometimes described 

 as bluish green. (The top was exhibited as a lantern slide.) 



My recent experiments seem to indicate pretty clearly the cause of 

 the remarkably bright red color, and also that of the blue. The more 

 feeble tints of the two intermediate groups of lines perhaps result from 

 similar causes in a modified form, but these I have not yet investigated. 



In the red color we have another striking example of an exceedingly 

 common phenomenon which is habitually disregarded; indeed, I can 

 find no record of its ever having been noticed at all. The fact is, that 

 whenever a bright image is suddenly formed upon the retina after a 

 period of comparative darkness, this image appears for a short time to 

 be surrounded by a narrow colored border, the color under ordinary 

 conditions of illumination being red. If the light is very strong the 

 transient border is greenish blue. Sometimes both red and blue 

 borders appear together, the blue being inside the red. 1 The color 

 generally seen is, however, red, and it is most conspicuous with good 

 lamplight. 



This observation was first made in the following manner : A black- 

 ened zinc plate with a small round hole in it is fixed over a larger hole 

 in a wooden board; the hole in the zinc is covered with a piece of 

 thin white writing paper. Thus we are furnished with a sharply 

 defined translucent disk which is surrounded by a perfectly opaque 

 substance. An arrangement is made for covering the translucent disk 

 with a shutter which can be opened very rapidly by means of a strong 

 spring. If this apparatus is held betweeu the eyes and a lamp, and 

 the translucent disc is suddenly disclosed by working the shutter, the 

 disk appears for a short time to be surrounded by a narrow red bor- 

 der. The width of the border is perhaps one twenty-fifth of an inch, 

 or 1 millimeter, and the appearance lasts for something like one-tenth 

 of a second. Most people are at first quite unable to recognize this 

 effect, the difficulty being not to see it, but to know that one sees it. 

 Those who have been accustomed to visual observations generally per- 

 ceive it without any difficulty when they know what to look for, and 

 no doubt it would be quite evident to a baby a few weeks old, which 

 had not advanced very far in the education of its eyes. 



The observation is made rather less difficult by a further device. If 

 the disk is divided into two parts by an opaque strip across the mid- 

 dle, it is clear that each half disk will have its red border, and, if the 

 strip is made sufficiently narrow, the red borders along its edges will 

 meet, or perhaps overlap, and the whole strip will, for a moment after 

 the shutter is opened, appear red. A disk was prepared by gumming 

 across the paper a strip of tin foil about one-thirtieth of an inch wide. 



1 I have recently shown that the greenish-hlue border is simply the "negative 

 after-image" of the red one. April 24. 



