164 



CORA D. REEVES 



The experiment. — Both A and B formed a rather persistent 

 habit of going to the left hand box so that the first day's records 

 of per cent of food choices have no significance. The days fol- 

 lowing, this habit was weakening and the correct choices in- 

 creased. The preceding curves, fig. 3, show that A was slow 

 in response as was also C, while B and D were habitually quicker 

 in response. The curves show also that there was learning 

 where the rats went to the moving stimulus, but with the others 

 less high percentages of correct choice were made. Some ten- 

 dency to decrease the average time for the day appeared. 



On the eighth day, after two days of perfect record, the rat 

 C was terrorized by a stranger with a dog entering the labo- 

 ratory while this rat was on the table. An attempt to get 

 again two days' perfect records caused the work to be prolonged 

 for two months. After about two weeks (160 trials for each 

 rat) the number of trials per day was increased to 20. There 

 was an immediate increase in the average time per trial as shown 

 by the following curve, fig. 4, which is made up from the time 

 records of the four averaged for successive trials. The later 

 trials for each day were slower. Evidently though only small 

 portions of food were allowed each time the hunger stimulus 

 was lessened after a few successful trials. 



Seconds 



^vjeT^geJ of z.o fr'ial* 



AuerAQet of ioo trials 



Figure 4. — Curve showing time in seconds for response to stimulus for four white 

 rats while discriminating between still and moving lights during successive 

 trials. 



