6 D. B. CASTEEL 



edges of their sides resting upon the insulation plate. The 

 side toward the screen has an opening at the top sufficiently- 

 large for introducing the hand in baiting the box. The side 

 toward the runway has an opening at the bottom sufficiently 

 large to allow the turtle to enter easily. The boxes fit loosely 

 in the end of the large oblong box and can readily be removed 

 or interchanged by the operator. They are protected from 

 the sides of the large box by cardboard insulation. 



In a few of the earlier experiments the designs or shades 

 used w^ere painted on the boxes, but for several reasons this 

 method proved unsatisfactory and was discontinued. The boxes 

 most generally used are as exactly alike as it was possible to 

 make them. Their distinctive character is given them by the 

 attachment of pieces of cardboard which bear the desired de- 

 signs. Upon each box are placed two of these boards. One 

 completely covers the outer surface which faces the runway; 

 the other covers the inner surface directly opposite the opening 

 through which the turtle enters in search of food. This latter 

 is immediately behind the cork blind which conceals the food 

 and is intended to give continued emphasis to the design after 

 the outer surface of the box has passed from view. These design 

 boards are fastened to the boxes with small clips and can readily 

 be removed. In practice it was customary to change them 

 from one box to another during or between a set of trials to 

 avoid the chance that the boxes themselves, through some 

 unnoticed peculiarity of form or odor, might aid in influencing 

 the choice. 



During most of the experiments the electric current used 

 was taken from a commercial circuit, being reduced to a proper 

 degree by use of a rheostat. The severity of the shock used was 

 varied considerably for different individual turtles. As a rule 

 a shock of very slight degree, one not unpleasantly perceptible 

 to the moistened fingers of the operator, was most effectual. 



GENERAL METHODS AND CHARACTERISTIC REACTIONS 



All of the turtles upon which observations were made were 

 kept in an aquarium and tamed for from three to six months 

 before experiments were undertaken. This is particularly neces- 

 sary when dealing with a reptile. Chrysemys is probably one of 

 the most tractable water turtles, yet some individuals never 



