618 PROCEEDINGS OF THE ACADEMY OF [Sept., 



On August 29, 1903, I placed about fifty workers and three queens 

 of this same species, with a half-teaspoonful of their pupae, upon the 

 base of the same Lubbock nest, in the same place in the same room,^ 

 hoping that they would again build a hut for their young during the 

 ensuing night. I had many tens of times put Stenamma fulvum on 

 this board, with their nest-earth and young, and they had never 

 once escaped by swimming. The board is square, thirty-eight centi- 

 metei's broad and four centimeters high. On its upper surface, 

 about a centimeter from its edge, there is a channel having a flat 

 bottom and vertical sides, two centimeters deep and from twenty- 

 six to thirty millimeters wide. When this channel is filled with 

 water an island about thirty centimeters square is formed by the 

 central portion of the board. The ants were on this island, and 

 their kind had never manifested al:)ility to swim across the sur- 

 rounding channel. These individual ants were habitants of a nest 

 located at a considerable distance from any body of water, and they 

 could have had no previous experience in swimming. I was there- 

 fore astonished on the following morning in finding that the ants, in- 

 stead of building huts for their young, were carrying it to a crevice 

 forty centimeters from the island, between the supporting table and 

 the wall of the room. Two of the queens had gone, and one queen 

 remained with a group of newly hatched callows. The labor of trans- 

 porting the callows continued all day. I was curious to see how the 

 queen would reach the crevice, as no ant travels far on any path befpre 

 untraversed by her. As soon as the queen crossed the channel there 

 was increased excitement among the ants. Several of the larger ones 

 approached the queen, and one of them lifted her free from the edge- 

 way and carried her to the crevice. All the remaining callows and the 

 pupse were likewise carried to the crevice, and at nightfall not an ant 

 remained on the island, nor had any dead ants nor deserted pupte 

 been left in the channel. The exodus had occupied thirty hours. 



When crossing the channel without a burden the smallest ants 

 walked on the surface of the water. Heavier ones clawed the surface 

 with the fore feet, walked with the middle pair and trailed the hinder 

 ones. The degree of submergence greatly varied. When carrying a 

 burden the ant swam with the legs wholly submerged, only the tarsi 

 of the hinder ones being above the surface. The antennae waved 

 constantly and progress was very slow. Ants without burdens were 

 from ten to sixty seconds in crossing the channel, while burdened ones 



^ These experiments were all made in the Marine Biological Laboratory at 

 Woods Hole, Mass. 



