282 THE ZOOLOGIST. 



Before the experiments began, a note was made of the 

 general habits. The Wasp always used the tunnel in going to 

 and from the nest. She returned at top speed from the feeding- 

 grounds, and bolted faultlessly into the tunnel, save on rare 

 occasions, when she erred by a few centimetres. The "Wasp 

 then hovered for a second or two, at a short distance from the 

 face, until she found the tunnel. Locality studies were not 

 made on departure from the tunnel. The periods of absence 

 ranged from three or four up to twenty minutes. 



The method used in the experiments consisted of blocking the 

 entrance to the tunnel with a plug of clay during the absence of 

 the Wasp, and then recording the time elapsing from the moment 

 of the Wasp's arrival in front of the plug until the moment 

 (approximate) she passed through the entrance of the rabbit- 

 burrow at each trial. The times were taken with a stop-watch, 

 excepting at the first three trials. The plug did not prove to be 

 of a very suitable material, the clay crumbling to pieces sooner 

 or later, when a new plug was required. The defect had a real 

 bearing on the results, but nothing else was available at the 

 time. Trials were given on four days within a period of eight 

 days, the last series being a test for retention. It was hoped to 

 form a new habit of going and coming by the rabbit-burrow b}' 

 retaining the plug constantly in the tunnel, but an hour was lost 

 at the first attempt, while the Wasp, as it subsequently turned 

 out, was engaged in breaking down the plug from the inside of 

 the tunnel. The plug was, therefore, removed from the tunnel 

 at each trial as soon as the Wasp passed into the rabbit-burrow. 

 The Wasp discovered the road to the nest through the rabbit- 

 burrow at the first trial with the plug. At the second trial she 

 succeeded in breaking up the plug, and entered the nest in the 

 normal way by the tunnel. At the third and subsequent trials 

 she was compelled to reach the nest by way of the rabbit-burrow. 

 The times recorded are shown in the table. 



The first ten trials were the most interesting of the series, 

 as within them nearlj' the whole of the learning process was 

 contained. The ten minutes of the first trial were consumed by 

 locality-studies of the immediate neighbourhood of the plug 

 (these studies are here called locality-studies, as they in no way 

 differed from the ordinary locality- study made by the Wasp on 



