206 H. M. JOHNSON 



After each of the blind dogs had learned to open three boxes 

 in this way, the eyelids were parted by an operation. All the 

 animals except Dog 7 showed an opacity of the cornea, usually 

 slight in degree and extent, but varying in different individuals. 

 Dog 7 did not exhibit an opacity. None of these animals showed 

 marked changes of behavior after the operation. I then required 

 Dogs 7 and 8 to learn three more problem-boxes each. They 

 exhibited slightly greater proficiency in acquiring skill than they 

 had shown while they were blind, but the improvement is due — ■ 

 at least in part — to the effect of practice, and is considerably 

 less than that brought about wholly by practice in the case of 

 the normal dog. 2 



These experiments demonstrated that the dog can get along 

 surprisingly well without vision in making complicated adjust- 

 ments. Other experiments indicated that even normal dogs did 

 not react to some objects as if the latter were visually per- 

 ceived, provided the objects were stationary. 



While the foregoing experiments suggested that the dog may 

 make little use of vision as we ordinarily think of it, they left 

 the question open as to the kind and degree of vision which the 

 dog may have. In 1911-12, I made an attempt to test the dog's 

 sensitivity to difference of outline in visual objects. I used a 

 single animal, designated as Dog 5 in the report cited. I made 

 the tests in the psychological laboratory of the Johns Hopkins 

 University, under the direction of Professor John B. Watson, 

 using the Yerkes-Watson standard apparatus 3 and the discrimi- 

 nation-method recommended by these authors. The test-objects 

 used were a circle 6 cm. in diameter and its (approximately) 

 equivalent square. The animal was fed at the food-box under 

 the square. He learned in about 1000 trials to choose the 

 square invariably, but its brightness (and hence its luminous 

 intensity) was four times that of the circle. When the bright- 

 ness-difference was eliminated, the animal immediately ceased 

 to show any preference for the square, and he did not make 

 any consistent improvement in 600 trials. It became necessary 

 to discontinue the work at this point. Shortly afterwards the 



2 The records of dog 6 (normal) and of dogs 7 and 8 (first blind and later with 

 the eyelids separated) are intercomparable. The records of dogs 1 and 2 and of 

 dog 5 are not. 



3 Yerkes, R. M. and Watson, J. B. Methods of studying vision in animals. 

 Behavior Monograph, no. 2, Cambridge, Mass., Henry Holt & Co., 1911. 



