34 



STELLA BURNHAM VINCENT 



Ti me 



Errors. 



most part eliminated such accidents before the twenty-fifth trial 

 these animals had more falls after this time than before : a total 

 of twenty-eight as against nineteen. 



Still one might have expected a greater number than this on 

 such a maze and under such conditions. But the nose and vib- 



37 



3b 



35 



3^ 

 33 

 32 

 3/ 

 30 

 29 

 2,8 

 27 

 2b 

 25 

 24 

 23 

 22 

 2( 

 20 



/8 



n 



llo 

 15 



li 



/3 

 IZ 

 II 

 10 

 <? 

 8 

 7 

 6 

 5 

 H 

 3 

 Z 



I Z 2> 14 5 (o 1 8 "f^^W 



Fig. 10 — Learning curves, time and error, of blind rats with vibrissae. 



rissae seem to be of more help than eyes to a rat in such cases 

 and as these had to rely almost entirely upon the form of sensory 

 control best adapted to the situation, they were, doubtless, saved 

 many an accident. 



Strangely enough the curve plotted for these rats (fig. lo) 



