348 H. M. JOHNSON 



It is clear that sniffing for odorous food-substances could serve 

 no useful purpose, as food was present in both food-compartments. 

 It occurred to me that the dog might be testing the punishment 

 grills for charge in some way, and choosing the alley whose floor 

 was not charged. I therefore introduced an additional key in 

 the secondary circuit, so as to leave both grills uncharged until 

 the dog should step on to the wrong one, when I would close the 

 k circuit by hand. As the record shows, the suspicion was well 

 grounded. The animal's subsequent records weaken the suppo- 

 sition that the animal had ever been discriminating on a visual 

 basis. I now changed the wiring so that the circuit through 

 the primary is left open until the animal closes it automatically 

 by stepping on the wrong grill. Throughout the rest of the work, 

 on each individual, I took the additional precaution of testing 

 the secondary circuit with the fingers before each trial, to make 

 sure that the primary circuit was open. The tungsten-mercury 

 contact would occasionally fail to break. 



The nature of the dog's test for charge is an interesting question. 

 It is improbable that he was touching the electrodes with his 

 nose. That happened once or twice in the preliminary stages of 

 the work, and the reaction was negative and violent. At this 

 stage the signal-lamps did not nicker save when he stepped on 

 to the grill. Unipolar stimulation of the nose is possible, but it 

 is hard to suppose that he could control his movements accurately 

 enough to avoid bipolar stimulation while seeking unipolar 

 stimulation. The current used was not strong enough to be 

 markedly unpleasant to the moistened human fingers, but the dog 

 usually howled when he received it through his wet foot. His 

 nose is always damp, and is apparently more sensitive to shock 

 than the foot. I ascertained that his vibrissae are too short for 

 him to touch the electrodes with them without bringing the nose 

 or lips uncontrollably near. They or the hair about the lips 

 might have been so affected by the electric field as to give a 

 sensory response, but in this connection it is well to note that 

 they were usually wet. Dr. Lorenz of this laboratory was kind 

 enough to go over the situation with me when I consulted him 

 regarding the possibility of ozone or some odorous gas similarly 

 generated being in the charged field. He informs me that with 

 the difference of potential which in this case existed between the 

 charged electrodes, ozone in minute quantities would certainly 



