VISION IN THE MOUSE 577 



the edge, is not a search for the nearest point of the surface below, 

 but merely the activity resulting from a kind of agoraphobia; and 

 that the slight height of the disk from which the animal would not 

 jump at all is due, not to any visual short-sightedness, but to the 

 fear of jumping in general. Undoubtedly both of these con- 

 ditions may obtain in the mouse when on the disk, yet the de- 

 creasing series of times elapsing before jumping, as recorded, 

 corresponds in such a way with the decreasing height of the disk 

 as to make it clear that vision does come into play in the estimat- 

 ing of distance. That this perception is not due to some other 

 sense which receives the vibrations from near objects is shown by 

 the poor results when glass was used. The animal seems to be 

 unable either to see the glass or to be affected by any object on 

 the other side of it, except when the object is very close to the glass. 

 This is what we should expect, considering the apparently short 

 range of vision of the mouse and its lack of perception of sharp 

 visual outlines. The shiny surface of the glass is evidently con- 

 nected with the estimated distance of objects on the other side. 

 The black paper when covered with the glass appeared more glossy 

 than the white. This doubtless made it appear less certain as a 

 place to jump to than the white. 



Turning the table, or the disk, or both, through an angle of 180 

 degrees, to act as a check against choice according to absolute 

 position, changed the results only slightly. 



PROBLEM 5. PERCEPTION OF THIRD DIMENSION IN OBJECTS 



This problem is very nearly related to the preceding one. It 

 has reference to the third dimensional nature of the object perceived 

 rather than to the distance of the object from the animal. It is 

 treated separately on account of the entirely different method and 

 apparatus made use of. 



Apparatus : A box measuring 42 cm. x 23^ cm. and 7 cm. deep, 

 with openings at both ends, was used (fig. 7). At one end is 

 attached the triangular enclosure A, and the other end of the box 

 opens into the nest, N. Leading into the box from A is a narrow 

 passageway 10 cm. long and 3.5 cm. wide. B and C are adjust- 

 able partitions which barely pass each other when moved back and 



