526 



PROCEEDINGS OF TUE ACADEMY OF 



[Opt., 



usual by c-d. Twice more she likewise came to the paper, and 

 although several other ants had meantime passed over it, she 

 retraced her steps into c-d, and thereafter made no more excur- 

 sions into m, going always through c-d. 



Dot Six entered the maze at T, made her first round trip, with- 

 out a burden, through vi, then took a false start into c-d, wan- 

 dered there, reached 2, returned to m^ and followed the m run to 

 the nest. Thereafter she made nineteen journeys, going in fifteen 

 times through m and four times through n, and returning once 

 through n, and eighteen times through c-d. I then put a barrier 

 across 2, expecting her to turn back and take the m or n route, 

 but in returning on her trail she discovered what neither the ants 

 nor I had before observed, a small hole in the glass wall at h, and 

 through this hole she reached n. After that her journeys, made 

 with no hesitation, were as follows : 



25 " n-7i-d 



After Dot Six had several tens of times passed through the hole 

 in the wall, I stopped the runs from c-d at 2 and at 3, enclosing 

 two burden-beariug ants and a dropped pupa. The two ants had 

 trails directly to I, and when they found themselves unable to 

 proceed through .2 they turned back and repeatedly explored all 



