^-2 PROCEEDINGS OF THE ACADEMY OF [Sept.. 



2. Of twenty-two ants, without burden, on their wav from the 

 island to the edgeway, seventeen turned themselves in the water and 

 went to the edgeway. Five failed in the effort to turn and went 

 back to the island. 



3. Of five ants, with burdens, on their way from the island to the 

 edgeway, four turned themselves in the water and swam to the 

 edgeway. One, after a long struggle, landed on the island. 



4. A burden-bearer, on her way from the island to the edgeway. 

 was gently turned around by me when she was in mid-channel and 

 caused to land upon the island at her starting-point. She was mani- 

 festly bewildered. Holding on to her charge, she wandered about over 

 the island for twenty minutes, then returned to the place of her landing, 

 laid down her burden, swam across the channel to the edgeway and 

 back again, picked up her burden, crossed the channel again and went 

 to the nest. 



Dead ants, or flecks of wood or yarn of the same size as an ant, when 

 cast upon the water and turned by its swirling in the same way as 

 that in which I turned the ants, did not revert to their former positions 

 The volition of the ant, not the force of the water, was the cause of 

 the semi-revolution made by the ant in turning back upon her course. 

 She can orient herself upon trackless water, as well as upon a tracked 

 solid, but in either case her environment must have been previouslv 

 explored. 



I then undertook tests showing whether the ant could orient herself 

 when removed from the shortest line between the end of her path on 

 the one shore and its beginning on the other shore. 



Test /.—When the ant was on her way to the island, and in mid- 

 channel, I placed the nozzle of a bulb-syringe at a distance from the 

 swimmer and very gently drew the water and the ant several centi- 

 meters out of her line of crossing and in the direction of the length 

 of the channel, without changing the direction of the ant's bod}-. 

 Of twenty-eight ants thus drawn away from their position, all gave 

 marked signs of bewilderment. Some ceased for a long time from 

 swimming; some made a complete re^'olution sidcwise; some 

 plunged deeper; some swam up and down the channel.' That 

 they had lost their bearings was as evident as if thev had been 

 upon a solid, although their movements were more limited in range. 

 Twelve of these ants finally landed on the island and fourte^'en 

 landed on the edgeway. 



Of five other ants crossing with burdens from the island to the 

 edgeway, all were similarly bewildered by being drawn away from 



