16 D. B. CASTEEL 



TABLE 7 



Turtle No. 9, Pattern Discrim- 

 ination 



Result of preference trials: Pattern A, 



19; pattern B, 31. Fed in 



pattern A box 



Test Right Wrong 



1 10 5 



2 7 3 



3 5 5 



4 5 5 



5 4 6 



6 4 6 



7 4 6 



8 5 5 



9 6 4 



10 9 11 



11 9 11 



12 9 11 



13 11 9 



88 87 



LINE DLSCRIMINATION 



The most definite and satisfactory results were obtained in 

 testing for this type of discrirrination. The experiments may 

 be divided into two groups: 



I. Discrimination between two lined boxes on which 

 parallel lines of equal width run horizontally on one 

 box, vertically on the other. 

 II. Discrimjination between two lined boxes on which the 

 parallel lines run in the same direction on both boxes 

 but are of different width. 



In all cases the amount of black and white on the faces of the 



boxes was the same. Great care was taken that the faces of 



the boxes exhibiting the lines should show no other difference 



which might aid the turtle in making a choice. The electric 



shock was used for punishment and preference tests were taken 



in all cases. 



Group I 



Experiments in line discrimination under the conditions 

 described for Group I were conducted as follows: 



The boxes were faced with cardboards bearing parallel lines, 

 black and white, each 8 mm. in width. On one of the boxes 

 the lines ran vertically, on the other horizontally. Fig. I illus- 

 trates the apparatus arranged for this experiment. Preference 

 tests were taken and the box showing the lesser number of 



*?3 



