70 



L. W. SACKETT 



curves of plate III tell a truer, more complete or even a materi- 

 ally different story from those of plate II. They are submitted 

 for scientific completeness and in the hope that others may 

 find more value in them than this study has been able to appre- 

 ciate. The writer's own opinion is that the method of counting 



oaooq 



fOaa^ 



yiATE M. 



Sc^\^\ terror =: $-0 JecoTj^j. 



/S' 2,0 ZS~ 30 -+ 3J 



^o ^5^r~^ 



errors by segments of the retraced path is faulty in that retrac- 

 ing through 10 segments in succession does not represent;- 10 

 successive errors but' rather one error with its logical conse- 

 Cjuences, chief of which are motor factors, among which^may 



