318 



K. S. LASHLEY 

 TABLE IV 



Discrimination Between One and Two Circles of Equal Area. Per- 

 centage of Error in Each Hundred Trials 



EXPERIMENT 5 



This experiment was similar to the preceding, but instead of 

 circles the forms used were two illuminated lines, 2 by 60 milli- 

 metres, one horizontal and one vertical. Rat No. 2, at the age 

 of seven months and after about 2000 trials in other experi- 

 ments, was given this problem After the experiment with 

 interrupted light she had been trained to react to differences in 

 brightness so that, at the beginning of this experiment, she 

 attended to the illuminated forms. After 150 trials she began 

 to show evidence of discrimination, hesitating and apparently 

 looking at the forms before choosing one of the passages. 



Table V gives the daily records for 760 trials. After 300 

 trials there was constantly more than 80 per cent of correct 

 choice, and since control series in which both lines were placed 

 horizontally gave 50 per cent of error, this was taken as suffi- 

 cient evidence of discrimination. Control series in which the 

 relative brightness of the forms was altered did not affect the 

 reaction, while any great change in the shape of the forms 

 themselves caused a complete failure. 



EXPERIMENT 6 



When the reaction to the lines was firmly established an 

 attempt was made to determine the animal's ability to dis- 

 tinguish lesser differences. For this purpose two of the standard 

 forms were provided with sliding metal plates, adjustable in 

 two directions, so that the openings could be varied from a 

 narrow line to a square or rectangle. These were first set as 

 lines, 20 by 60 millimetres, and substituted for the horizontal 

 and vertical lines used in the preceding experiment. By short- 

 ening and widening the lines the contrast between them could 

 be reduced until both became squares, equal in area. 



