PATTERN-DISCRIMINATION IN VERTEBRATES 347 



with forced ventilation. The floor was kept covered with a layer 

 of sawdust 2 inches thick, in the interest of cleanliness. All the 

 animals, with two exceptions, remained in excellent condition 

 throughout the work. Monkey 1 was active and playful on the 

 morning of his death. The cause of death was not ascertained, 

 but the most plausible guess is that he became overheated and 

 fell to the floor from an overhead steam-pipe, on which he habit- 

 ually slept. Chick 2 was the smaller of the two gamecocks and' 

 sustained severe injuries in fighting the other bird. On February 

 9, while in a crippled condition, he was also attacked by Dog 1 

 and sustained several deep scratches, some of which required 

 several stitches. They were properly treated, however, and 

 healed without infection, and without serious disturbance of 

 the work. 



The learning records of the different individuals appear in 

 Tables 1 to 4, inclusive. Certain noteworthy features of the 

 behavior of the different animals must be noted separately. 



The dog's learning record up to and including January 7, when 

 the problem ordinarily would be regarded as ' learned " is 

 unusual. Up to the introduction of punishment he was not 

 affected by the stimuli. From the day when punishment became 

 effective the learning curve descends abruptly,, showing very 

 short plateaux and no reversals . This, as far as I have observed, 

 is hardly characteristic of acquisition of a discrimination-habit. 

 An earlier (unpublished) experiment of mine on visual form- 

 discrimination in the dog, which I expect soon to describe in 

 another report, and the behavior of some blind dogs which I 

 worked 5 in comparison with normal dogs, tended to make me 

 suspicious of this dog's record. At the time, however, I could 

 find nothing to warrant the suspicion, and in good faith I reported 

 the dog's results 6 as indicating that the method of making such 

 tests is reliable. On January 26, after the test-bands had been 

 reduced to about 40% their original size, I noticed that the dog 

 sniffed several times at the entrance of each alley before he made 

 his choice. Dogs under conditions of excitement often sniff 

 apparently to release surplus energy, but this animal was making 

 perfect records and did not show any other forms of excitement. 



6 Johnson, H. M., "Audition and Habit-Formation in the Dog," Behavior Mono- 

 graphs, No. 8, 1913. 



• "A Method of Testing Visual Acuity and Pattern-Discrimination in Animals," 

 Psychological Bulletin, vol. 11, 1914, pp. 55 ff. 



