PATTERN-DISCRIMINATION IN VERTEBRATES 345 



compartment introductory to the alleys, and placed them so 

 that the animal could not bring the eye nearer than 60 cm. to 

 the test-object without overstepping the stop and stepping on 

 to the punishment grill. The setting of the stop used was different 

 for each animal. This rendered the accommodation-factor the 

 same for each animal. 



All the animals were punished for incorrect choices by placing 

 inductive charges on the electrodes of the punishment grills. 

 The current was taken from a 110 volt A. C. power circuit, a 

 100 watt lamp placed in another room being connected as resis- 

 tance in series with the primary coil of a Zimmerman inductorium. 

 The position of the secondary coil with reference to the primary 

 could be varied at will, the setting being shown on a scale belong- 

 ing to the instrument. There is much less disturbance with this 

 arrangement than where D. C. is used with an interrupter. In 

 previous work I found the noise from the interrupter quite 

 disturbing, and the interrupter also worked uncertainly at times. 

 In the first few weeks of this experimentation I charged the grill 

 under the negative test-field before releasing the animal from 

 the home-box. The circuit through the primary coil was kept 

 closed throughout the daily series of trials. Later this arrange- 

 ment was found quite unsatisfactory, as will appear in remarks 

 on the behavior of the dog. I then changed the wiring so that 

 the circuit through the primary was kept open until the animal 

 closed it by stepping on the grill and thus depressing it. This 

 prevents the animal from testing the two grills for charge, and 

 acquiring a discrimination-habit on that basis, rather than on a 

 visual one. The shock under the latter condition is quite 

 different in character from that under the earlier conditions. 

 Under the latter, there is a pronounced ' inductive kick ' at 

 the instant the primary circuit is opened or closed. This is 

 disturbing at first, but has one advantage: the animal tends to 

 leave the alley at once, rather than remain on the grill and receive 

 more punishment. The charge, of course, must be made con- 

 siderably lighter than under the former set of conditions. Before 

 each daily series of trials, I soaked the animal's feet for several 

 minutes, so that they remained moist throughout the series. 

 I could not use a wet pad in the box, owing to the character of 

 the wood, which warps very easily. But the shock conditions 



