PATTERN-DISCRIMINATION IN VERTEBRATES 171 



his eyes in or quite near the plane of the windows in the parti- 

 tion — 60 cm. from the test-field. In all cases I recorded as the 

 alley "chosen" the one into which the animal first stepped. 

 The animals automatically and instantly registered their choices 

 by breaking the circuit through a signal lamp as they stepped on 

 to the floor of the alley. The third change is the introduction 

 of a shallow copper tray into the Yerkes box, between the exit- 

 door and the alleys A 1 and A 2 . This tray contained a moist- 

 ened felt pad 4 on to which the animal had to step before he 

 could enter the alley. This tended to minimize the variability 

 of the resistance of the animal's feet. It is somewhat more 

 satisfactory than the means employed in my earlier work — that 

 of soaking the animal's feet before beginning the daily series 

 of trials. 



RELATIVE EASE OF LEARNING FOR INDIVIDUAL ANIMALS 



The animals used were a young male Capuchin monkey and 

 two Indian gamecocks. They were described in paper No. 2 

 of this series, and designated severally as Monkey 2 and Chicks 

 1 and 2. Since that report was published, Dr. P. W. Cobb 

 refracted the eyes of the birds by skiascopy. The static error 



4 Breed had used a similar device in his work on the chick. Some such precau- 

 tion is indispensable to good results if punishment with electric shock is employed. 

 The resistances of dry feet and moist feet are of different orders of magnitude. If 

 an animal is put into the box with dry feet and hands, and if during the series the 

 latter become moist, as from perspiration or contact with urine or wet food, the 

 increased effect of a steady line current may work great disturbance. In my work 

 on audition in dogs I found this factor troublesome. Miss E. M. Smith reports 

 a similar experience. The unsatisfactory results reported by some other experi- 

 menters may be partly due to failure to take this precaution. Watson (Behavior, 

 p. 60) asserts that some animals " are extremely resistant even to fairly high cur- 

 rents. The rabbit is not disturbed in the smallest degree by a current which is 

 unbearable to the human being." It should be remembered that thick tufts of 

 long hair cover the soles of the rabbit's feet, forming an excellent insulation when 

 dry. When moistened, the water acts as a conductor. The rabbit will then react 

 violently to a current which is not unpleasantly strong when received by the 

 human subject through the moistened finger-tips. I recently demonstrated this 

 fact in a simple experiment on the rabbit. However, even with the best attempts 

 to keep the resistance of the animal's feet constant, the variability is fairly large — 

 as great, I have been told, as 25%. It is impossible to keep the animal's feet free 

 of dirt and grease, which interfere with good contact with the electrodes. It there- 

 fore seems unnecessary to use such extreme care to control the steadiness of the 

 line current as some students have done. I have obtained satisfactory results 

 by passing A.C. from the city circuit through the primary coil of a Zimmermann 

 inductorium, connecting a bank of lamps in series with the coil and in parallel 

 with each other. Small changes in intensity of shock can be made by changing 

 the position of the secondary coil with reference to the primary. This method 

 has one advantage over the use of D.C. with an interrupter, in that the noise of 

 the latter, which is often a source of disturbance, is dispensed with. 



