rouse: the visual perception of distance. 



1^3 



It will be seen that the average estimations generalh' vary in- 

 versely as the size of the object observed, i. e., as the box used is 

 larger, the distance judged is shorter, and vice versa. E. g., for 3 

 meters in (I), the smallest box (i) was thought to be 3.75 meters 

 distant, and the next larger ones, (2), (3) and (4), 2.60, 2.45 and 

 2.35 meters, respectively. This is a common illusion, and it is 

 natural that it should be shown here. To this tendency there is 

 but one exception in (I) and one exception in (II), while in (III) 

 there are a half dozen exceptions, and in (IV) the illusion almost 

 wholl)' disappears, showing that m unassisted binocular vision there 

 is a slii^Ii/ tendency to overcome the mistake of judging a larger 

 object to be nearer, and a smaller object to be farther, while in bi- 

 nocular vision assisted in different ways the illusion, in a great 

 measure, disappears. A comparison of the sums of the averages 

 for each box in different columns will show the same relation more 

 plainly. Observing the same order of boxes, the smallest first, we 



find the sums as follows: 



Table II. 



By averaging (1), (2), (3) and (4) of each of the four columns 

 of table No. i, the final average estimations (with different kinds 

 of vision) appear as follows: 



Table Hi. . > 



Table III seem to indicate that, within a scope of 15 meters, dis- 

 tance is nearly always iiudercstiniatid, and appears less to the hft 

 eye than to the right eye, and less to botJi eyes (unassisted) than to 

 the left eye, as is more plainly shown by the following diagram. 

 The perpendicular lines represent on a small scale the actual dis- 

 tances indicated at their lower extremities; and the length of these 

 perpendiculars from the horizontal base line to where they are cut b}'- 

 the different curves, shows the respeetive estimates of these "actual 

 distances," - 



