REWARD AND PUNISHMENT IN HABIT FORMATION 241 



ject chose the light box it found toasted corn flakes soaked in 

 cream in the food receptacle. The rat was given about ten 

 seconds in which to eat, then the experimenter made it pass on 

 through the alley into the nest box ready for another trial. If the 

 animal chose the dark box it found the door closed and had to 

 return into the entrance chamber and pass out through the light 

 box. 



Timing of subject. The experimenter kept the time in seconds 

 that the subject took from the moment it entered the entrance 

 chamber until it entered one of the electric boxes. 



Selecting of subjects. In order to eliminate family differences 

 each litter of rats was divided equally; and one group trained 

 with hunger and the other with electric shock. 



PRELIMINARY EXPERIMENT 



At the beginning of this experiment certain preliminary ques- 

 tions still remained to be answered. The most important of 

 these were (1) The advisability of giving light-dark preference 

 series and (2) the earliest age at which rats are sufficiently well 

 developed to undergo periods of starvation as long as forty- 

 eight hours without being reduced to physically unfit conditions 

 for experimentation. 



Subjects used for this series of experiments were fifty-six days 

 old on the day the experiment began. Each rat was given one- 

 half hour a day for five days preceding the experiment proper in 

 the control box with doors opened, with the cover of the electric 

 box removed and food in the receptacles. This gave the animals 

 an acquaintance with the box and the place where food might be 

 found in case of hunger. 



Special conditions of series I. A series of experiments as here 

 used includes all subjects, both hunger and electric shock, which 

 were trained at the same time. The electric shock used for the 

 first series was seventy-five units (6) ; and the length of the period 

 of hunger was twenty-four hours. The difficultness of discrim- 

 ination was the same in all experiments being fairly easy. 



PSYCHOBIOLOGY, VOL. I, NO. 3 



