FUNCTION OF VIBRISSAE IN BEHAVIOR OF WHITE RAT 35 



resembles no curve so much as the curve plotted for the normal 

 rats in the same maze with sides up in Experiment I. In general 

 outlines they are nearly the same and also in the relation of the 

 time and error curves. The time curve for the blind rats is more 

 irregular but this is probably due to the necessity of avoiding 

 accidents in a maze without sides. Why does not the error 

 curve possess the same irregularities w^hich that plotted for the 

 normal rats in the same maze has (Exp. II)? As has been said 

 before these rats must have been thrown back almost entirely 

 upon the cutaneous and kinaesthetic sensations for control. 

 Even the contact sensations are greatly lessened in a maze like 

 this. There is no contact from the sides or from above and the 

 only contact sensations they get are those coming from the floor 

 through their feet, noses and vibrissae. Hence the curve par- 

 takes more of the nature of the curve for Exp. I, where the type 

 of maze used was that which, as has been proved by others, is 

 chiefly learned through kinaesthetic sensations. This curve is 

 more irregular and longer in reaching its level because of the 

 more difficult conditions. 



Experiment IX. Blind rats without vibrissae; maze without 

 sides. Five blind rats whose vibrissae had been removed were 

 next put upon the maze. These rats were taken at random 

 from the group of blind rats and were in just as good condition 

 as the other blind rats with vibrissae which we w^ere using at the 

 same time on the maze. 



These rats were put on the maze Aug. 26, and were kept at 

 the problem until Oct. 18. They had then been given over 150 

 trials, ten times as many as the normal rats needed and five 

 times as many as the other blind rats required on the average to 

 learn the maze. None of these rats had learned it at this time 

 and although one of them seemed to be cutting our errors and 

 probably all would eventually have conquered the coordination 

 the work was discontinued with them. Our interest lay not 

 so much in the actual time taken to learn the maze as in 

 the comparative difficulty of learning it under these conditions. 

 (Fig. II.) 



The curve is not plotted as the others, that was impossible, 

 but an error curve is made by averaging the errors by groups 

 of twenty-five in order to show the slight lessening of errors 

 after the first fall. 



