490 JOHN LINCK ULRICH 



to establish " sensory association" or " sensory motor con- 

 nections." 



The development of interaction in learning the inclined- 

 plane problem may be graphically represented by curves. Three 

 curves delineating this development are presented in curves I, 

 II, and III and IV. Points on the abscissa represent trials, and 

 those on the ordinate, the reaction time for a trial. A length 

 of 2 mm. on the abscissa represents one trial, and 2 mm. on the 

 ordinate 2 seconds. Curve I, graphically produces the records 

 of a rat which possessed fairly well developed reflex thrusts; 

 curve II a rat with these thrusts undeveloped; and curve III, 

 one with both undeveloped reflex thrusts and mechanism for 

 the production of reflex excitability. 



A description of these curves must not be a general one, for 

 now there is need for greater details. If learning is the develop- 

 ment of interaction of reflex body parts, the three curves must 

 graphically represent the results obtained from learning with 

 rats with these parts fairly well developed, and with others hav- 

 ing these parts undeveloped. The general outline of each curve 

 must be different in accordance with the degree of development 

 of interaction of parts for learning. Every altitudinal point, 

 the upslopes, represent the things that prevent a ready develop- 

 ment of interaction, and signify an increase in time to plunge the 

 plane; and every descending point, the down-slopes, in the 

 direction of the abscissa, indicate that interaction is becoming 

 effective and there is a decrease in time to plunge the plane. 

 A decrease in the height of the majority of the up-slopes, from 

 the first to the last trial, shows that the development of inter- 

 action for learning is progressive. 



Curve I is constructed from the records presented in table 11, 

 which reveal in a rat the reflex extensor thrusts fairly well devel- 

 oped and reflex excitability manifest. In the first trial, the reac- 

 tion time is short and the first point in the curve is close to the 

 abscissa and the second is still closer to this line. In fact the 

 second trial is perfect. There exists therefore, for these two 

 trials not great altitudinal points, but in the next trials these 

 points greatly increase in height with equally great descending 



