1903.] NATURAL SCIENCES OF PHILADELPHIA. 623 



position, and were many minutes in reaching shore. Three of them 

 landed on the island and two of them landed on the edgeway. 



Dead ants or flecks of wood or yarn, subjected to the same condi- 

 tions, did not behave as did the ants, and made no similar movements. 



Test g. — When an ant was returning to the island, and was in mid- 

 channel, I pushed the water wdth the cloth broom, so that the swimmer 

 would be borne on the water to a distance of a centimeter or two off 

 the line by which she would have crossed if the w^ater had remained 

 untouched, but keeping the axis of the ant's body parallel to that 

 line. Of nine ants thus moved five went on to the island and four 

 turned back to the edgeway. The evidences of bewilderment were 

 much less marked than when the ants were drawn to a distance of 

 several centimeters from their position. 



It is plain that the ants did not orient themselves from their own 

 track where no track existed. Far less of manipulation than was 

 spent upon the water will destroy the track of these ants upon a solid. 

 They must have traversed the distance across the channel depending 

 for their orientation on something other than their immediate foot- 

 prints. Their lesser bewilderment when they were moved a lesser 

 distance from their position, indicates that the}- orient themselves 

 by something that is not far removed from them. This may be the 

 remembered odor of objects at very short distances one from another, 

 recognized for guidance from point to point, as is supposed by Forel. 



The ants gave no indication that they had an intellectual concep- 

 tion of the island on which they spent many daj's, nor of its surround- 

 ing channel, whose edgeway they seemed to everywhere explore. 

 They gave no sign of an idea that if they crossed the water at any 

 point they could reach a solid and their destination. They were 

 manifestly guided by what was very minute in size and very near in 

 ])osition. But these minute things must be rememljered, since the 

 ant rightly orients herself by them after absences of considerable length. 



The queens were shoved off the island by the workers. They 

 drifted rather than swam to the edgeway, were hauled out by waiting 

 workers and carried by them to the nest. ^lany of the callows were 

 carried all the way from the island to the nest, while others followed 

 adult workers, all the time touching the leader with the antennae. 

 Occasionally three ants crossed thus in file. 



On the morning of September 9, ten days after the ants had been 

 placed on the island, only one ciueen and seven workers remained 

 there. All the five queens without antennse had been transported 

 to the nest, as had all the young. No loss of life or of young had oc- 



