CONSTKUCTING BKIDQES. 245 



1 now increased tlie length to 16 feet, and watched 

 her while she made thirty journeys backwards and 

 forwards. She also brought during the time seven 

 friends with her. 



It surprised me very much that she preferred to go 

 so far round rather than to face so short a drop. 



In illustration of the same curious fact, I several 

 times put specimens of L. niger on slips of glass raised 

 only one-third of an inch from the surface of the nest. 

 They remained sometimes three or four hours running 

 about on the glass, and at last seemed to drop off 

 accidentally. 



Myrmica rugiiiodis has the same feeling. One 

 morning, for instance, I placed one in an isolated 

 position, but so that she could escape by dropping one- 

 third of an inch. Nevertheless at the same hour on 

 the foUowiag morning she was still in captivity, having 

 remained out tweuty-four hours rather than Jet herself 

 down this little distance. 



Again I filled a saucer (woodcut. Fig. 11, s) with 

 water and put in it a block of wood (w), on the top of 

 which I fastened a projecting wooden rod (b), on the 

 end of which I placed a shallow glass cell (a) containing 

 several hundred larvae. From this cell I allowed a slip 

 of paper (p) to hang down to within J^- of an inch 

 of the upper surface of the nest. At one side I put 

 another block of wood (c) with a lateral projection 

 (d) which hung over the cell containing the larvae. 

 I then made a connexion between D and A, so that antg 



