PATTERN-DISCRIMINATION IN VERTEBRATES 349 



be present. It therefore seems possible that the dog was reacting 

 to olfactory stimulation from ozone or some other odorous gas 

 formed in the charged field. Bearing in mind his pronounced 

 sniffing behavior, this seems the simplest explanation. 



The sniffing responses persisted for a long time and were 

 never entirely overcome. After this change in the wiring in 

 over 67% of the times when the animal sniffed during a trial 

 the choice was wrong. 



The evidence obtained in this work is not conclusive against 

 the dog's possessing sensitivity to visual detail. The fact that 

 this individual learned to control his environment for a time by 

 the use of smell may have unfitted him in part for visual dis- 

 crimination. And yet the other animals acquired it more or 

 less easily. I have obtained another individual which is to be 

 put to work on the same problem, along with a cat, in order to 

 throw light on the question whether the carnivora have poor 

 vision, or whether the test made on this dog was unfair to him. 

 I may add that the present gratings are too fine for satisfactory 

 work with the dog. They do not permit test-bands wider than 

 4 mm. to be used satisfactorily. Above that value, parallax is 

 present, and the angle of rotation is very small, so that a very 

 small change in the setting of the micrometer screw produces a 

 very large change in the width of the test-bands. If an animal's 

 threshold lies in this region the steps which can be taken are too 

 large for convenience. 



Both the monkeys used were very active in the experiment-box. 

 When I introduced punishment in the tests on Monkey 1 he had 

 been maintaining an accuracy record of 80% for several con- 

 secutive daily series of trials. His responses were made without 

 comparison, however. Punishment was not satisfactory. At 

 this time the grill forming the floor of the alley under the negative 

 test-field carried an induced charge all the while the stimuli 

 were being presented. In one sense the charge was not unduly 

 heavy. The animal would frequently sit on the charged grill 

 for several minutes, chattering, and picking at different parts 

 of his body. Occasionally he would touch the grill with his hand, 

 give a sudden, though not loud, exclamation, and put his fingers 

 into his mouth. But he showed a decided aversion to touching 

 the grill while choosing, and developed numerous methods of 

 getting through the alley without doing so. I had not sue- 



