232 [October, 



ponds, I observed no specimens either settled there or even flying 

 over them. 



Ovping to a change in the weather collecting had to be postponed 

 for the time being, but quite at the end of August T again visited the 

 locality. After nearly five hours' searching in a gale on August 25th, 

 late in the afternoon a glimpse was caught of a dragon-fly ovipositing 

 in some reeds at the pond where I had first collected on August 3rd ; 

 I got nearly within striking distance, but the insect flew up, hovered 

 a moment — just long enough for me to see the pointed ovipositor so 

 characteristic of *S^. metallica ? — and was off into the trees. It is 

 curious that I should have recorded a female of this species on 

 August 3rd, 1907, at Marburg-on-the-Lahn, in Germany, being 

 frightened in much the same way, and settling in a tree ; very likely 

 the Sussex one also took protection among the leaves oF a high oak 

 close by, and it is just possible that the green colour of the insect is 

 protective. The latter conjecture is strengthened by the fact that I 

 saw comparatively large numbers of Gordulia cenea settling on the 

 tree-tops at Burnham Beeches in June. 



On August 27th I got very close to an insect which looked like 

 another, or was perhaps the same, female ; later in the day I twice 

 saw other specimens, but it may be that I drove the same one from 

 pond to pond. After August 29th, collecting became impossible 

 owing to the storms. All the specimens taken in Sussex at the be- 

 ginning of August were fully mature, but although three of them 

 had damaged wings or legs, they could not be called " worn," and the 

 wings of all, moreover, were only very slightly suffused with brown, 

 which points to the fact that they were not very old. 



The abdomen of Scotch specimens, which I have seen, is notably 

 stouter than that of the Sussex insects, and the wings of the former 

 taken at the same time of the year are much browner than the wings 

 of the latter. 



Whether *S^. metallica has migrated from the continent or not is 

 at present uncertain, but there is no particular reason why it should 

 not be resident. It is true that large numbers were seen within an 

 area of about one square mile, but by my list of other Odonata ob- 

 served in the same place it is clear that the locality is an ideal one 

 for Odonata. 



The following are the species observed near Tunbridge Wells on 

 August 3rd, 4th, and 29th :— 



