SOME SOCIAL WASPS 251 



six minutes, cleaning its antennae with its legs; then fell to 

 the flower-heads below, where it remained. This one did not 

 return the same day, nor the next, but on the third evening 

 at 5 : 50 p. m. it dragged itself wearily into the nest, after 

 an absence of almost two and a quarter days. 



Thus of the four queens that had not made the preliminary 

 flight, three returned to their nests; two of them did it in the 

 same time as some queens used in Experiment I, and one of 

 them in about twice the time. The fact of some evidence of 

 defect in wasp no. 10 practically eliminates her from the 

 experiment. 



This experiment shows that the short flight near home as 

 a preliminary training is not necessary to insure a successful 

 long flight. The evidence also increases with increased ex- 

 perimentation that the wasps do not return during the night. 



Experiment III 



June 21. Six queens were used in this test, four of which, 

 nos. n, 13, 15 and 16, were used for the first time. No. i 

 had been used in the one-eighth-mile flight from the mail- 

 box and had been successful ; No. 3 had done the same and, 

 in addition, had made the return flight from Becks over a 

 distance of one and one-half miles. 



In the previous experiments, the releasing point and the 

 home were at the same elevation. This time the wasps were 

 taken down the hill to an altitude about 1 50 feet lower than 

 the home point, carried along the railroad tracks for a dis- 

 tance of two miles, in darkened wire cages, with many turns 

 of confusion on the way, and released at Kimmswick. They 

 were taken from their nests at 8 : 45 and released at 10 : 30 

 a. m. When they were liberated, a very gentle breeze was 

 blowing toward the south, against which they would have to 

 travel to reach home. If they were to return by the curving 

 railroad track, they would have clear sailing, but if they 

 deviated from this, their route would be most difficult be- 



