DISCRIMINATIVE ABILITY OF THE TURTLE 



17 



entrances . was chosen as the food box. Discrimination was 

 then established by a series of tests using difference in the direc- 

 tion of the parallel Hnes as the criterion of selection. If dis- 

 crimination was satisfactorily established with the lines 8 mm. 

 in width, the cardboards bearing these lines were then removed 

 and boards bearing lines but 4 mm. wide were substituted on 

 both boxes. If discrimination was fairly maintained after this 

 reduction in the width of the lines they were then further re- 



Sinm. 



f 77* 77^. 



i mm. 





UntTn. 



imtri. 



Fig. IV — Graphic representation of the widths of hnes used in the experiments 

 on Line Discrimination. The sides of the boxes facing the runway were covered 

 with hnes of the widths shown. 



duced to 2 mm. on both boxes. No lines narrower than 2 mm. 

 were used in the Group I experiments. At each reduction the 

 number of black and white lines on the faces of the boxes was 

 doubled though the amount of black and white surface remained 

 the same. Fig. IV illustrated graphically the various widths of 

 lines used in the Line Discrimination experiments. 



The histories of turtles Nos. 4 and 10 illustrate the results 

 of the horizontal-vertical line experiments. No. 10 showed 

 much greater speed in acquiring the association than No. 4. 

 This latter turtle became sulky upon reduction in the width 

 of the lines to 2 mm. No. 10 was in every way my most satis- 

 factory animal. It learned the problem fairly rapidly, showed 

 perfect memory for 8 mm. lines two weeks after learning them, 



