( xliv ) 



took place occasionally. Finally the insects rose again and 

 flew round in the same manner as before, similarly alighting 

 two or three times for a few seconds. On one occasion, 

 perhaps the first time after the long rest, they dived deeply 

 into the grass, and I saw them fluttering together far down 

 and almost hidden. They soon emerged, however, and finally 

 alighting again, the male recommenced the rapping move- 

 ments. On this occasion, however, the process was very brief, 

 lasting for perhaps a quarter of a minute ; they rose again 

 and almost immediately darted into the long thick grass. 

 Again I saw them fluttering far below the surface, but in a 

 few seconds they had come to rest side by side on the under- 

 side of a buttercup-leaf. Coitus had already taken place, and 

 the extremities of the two abdomens were bent round into 

 a U. The wings of both were closed and were hanging down- 

 wards. The leaf on which they were resting was some inches 

 below the top of the long grass. They were completely hidden 

 and could only have been found by tracking them as I had 

 done. They were now completely quiescent when the grass 

 was disturbed in order to examine them more closely, when 

 the buttercvip-stem was picked, and even when I removed 

 them, still side by side, between my finger and thumb. In 

 this manner I carried them a few hundred yards and then 

 took them home in a box and placed them in a cylinder with 

 some green leaves, upon one of which they immediately took 

 up a position similar to that which they had occupied on the 

 buttercup-leaf. From 4.15 p.m., when coitus took place, they 

 remained until 11.15 p.m., when I disturbed them in trying to 

 ascertain Avhether pairing had come to end. Under ordinary 

 circumstances it is improbable that they would have separated 

 until the next day. 



" In one respect the observation is incomplete. Courtship 

 had already begun, and had been going on for an unknown 

 period when the buttei'flies were first seen. 



" In spite of this hiatus in our knowledge the observations 

 here recorded throw much light on an obscure part of the 

 natural history of this species. We now know by a direct 

 observation that pairing takes place between hybernated 

 individuals in the spring — a conclusion which was long 



