278 THE ENTOMOLOGIST. 



Si/mpetrum was especially well in evidence, no fewer than five 

 out of the six British species having fallen to our nets. Tlie 

 exceptionally fine weather which characterized the summer 

 extended far into the autumn, and we enjoyed the unusual 

 experience of taking dragonfiies so late in the year as October 

 21st. 



Eepresentatives of seventeen species were obtained, viz. : — 



(1) Pi/rrhosoma nipnvhula. — The taking of immature females 

 of this dragonfly — always our earliest — opened the season on 

 May 13th. It became tolerably common throughout the district, 

 and was last met with on July 22nd. 



(2) Brachytron pratense. — On May 27th we saw at a pond 

 occupying the site of an old forest gravel-pit what appeared to be 

 a newly-emerged dragonfly resting on a rush far beyond our 

 reach. We threw twigs at it, but, instead of causing it to rise, 

 we only succeeded in covering it up. On the evening of the 

 following day we found the insect occupying much the same 

 position, and, by lashing together two or three sticks, we were 

 able to reach and secure the specimen with the net. It proved 

 to be a female of B. pratense, a species new to our collection. It 

 was in good condition save for the undeveloped state of the wings 

 on the right side, a blemish which remained in spite of our 

 keeping the insect alive a day or two longer at home. The 

 circumstances point to the specimen having been bred in the 

 pond where it was found. The species was not met with again 

 during the season. 



(3) Agrionpuella was first found on June 3rd, when some indi- 

 viduals were in a mature state. On June 24th two males bearing 

 a close superficial resemblance to A. pulchellum were taken at 

 one small pond. The U-shaped marking on the second abdo- 

 minal segment was connected by a slender black line with the 

 circlet behind. The base of the marking was thicker than in 

 typical puella, and its posterior margin was not so deeply exca- 

 vated. In a third specimen, taken at the same time and at the 

 same pond, the connection with the circlet behind was not com- 

 plete ; the circlet was crossed transversely by a short median 

 black line, which anteriorly nearly joined a backward process 

 from the U-shaped marking. Another male, with the connection 

 complete, was obtained at a different locality on July 15th. That 

 these specimens must be regarded as A. puella, and not as 

 A. pulchellum, is clear from the morphological characters which 

 separate the two species. It is interesting to compare this aber- 

 ration of puella in the direction ol pulchellum, with the variation 

 of pulchellum towards puella described by Mr. W. J. Lucas in 

 1901 (Entom. xxxiv. 215). On July 1st we had an opportunity of 

 watching at close quarters a pair of A. jniella ovipositing; the 

 female was evidently placing her eggs within the bark of the 

 floating twig upon which she was resting, while the male, 



