824 



SCIKNGE. 



[N. S. Vol. VIII. No. 206. 



show by a qualitative and quantitative 

 study of this illusion that no factor other 

 than irradiation need be appealed to for a 

 thoroughly satisfactory explanation. The 

 hesitation shown in accepting this explana- 

 tion has been partly due, no doubt, to the 

 fact that in the usual cases of encroach- 

 ment by irradiation the diminished dark 

 areas have retained outlines that are every- 

 where parallel to the original outlines of the 

 figure. One has only to think of the dark 

 square or circle on the light background. 

 In the present case, however, the effective 



■■a 



Fig. 9. 



point of the irradiation is in the corners 

 formed by the adjacent rectangles. The 

 white areas hore into the dark corners, 

 as it were. In this way those portions of 

 any vertical that run along the sides of the 

 various rectangles are deflected towards the 

 corners, both above and below, and the de- 



flections of these several portions give the 

 tilted character to the line as a whole. 

 This irradiation in corners is strikingly 

 shown in Fig. 9, where the point of effec- 

 tive irradiation has been shifted to the cen- 

 ters of the incomplete squares. Here there 

 is no longer a deflection of the vertical, but 

 instead the bars of white that cross the line 

 seem to slope slightly downwards to the 

 right. The result of the qualitative and 

 quantitative investigation above referred to 

 showed clearly that the illusion vanishes 

 whenever there can be secured the impossi- 

 bility of irradiation in the corners situated 

 along the line. The figures devised secured 

 the latter condition while still retaining 

 any factors that this figure may have in 

 common with the Zollner patterns. If, 

 further, the character of the illusion in the 

 regular figui-e was altered by substituting 

 colors for the blacks and whites, or if the 

 character of the illumination was changed 

 by the use of the electric spark, or by the 

 interposition of colored media between the 

 diagram and the eye, the measurements of 

 the illusion disclosed varying changes in the 

 amount of apparent deflection. These re- 

 sults seemed to show the entire sufficiency 

 of the explanation in terms of irradiation, 

 at the same time rendering superfluous the 

 ajjpeal to, or the search for, auj^ further ex- 

 planatorj' principle. 



5. Loeb's Illusion. — In 1895 Professor Loeb, 

 of Chicago, called attention to the follow- 

 ing interesting illusion.* Let 31, Fig. 10, 

 be a fixed vertical line and a a shorter line 

 parallel with M and lying to the right. 

 Placing Jlf in the median plane and steadily 

 fixating some point in it, place a second 

 line b in such a position that it shall be con- 

 tinuous with a. This attempt will prob- 

 ably succeed very well. But now being a 

 third line in the position occupied by c in 

 the figure b will seem to lie too far to the 

 right. That is, b must now be brought to 



* Jacquea Loeb. Plfiiger's Archiv, LX., 516. 



