552 M. W. R. de V. GRAHAM 



stouter than the pedicellus, nearly cylindrical ; first funicular segment i-i to 1-5 times as long 

 as broad, the following segments slightly shorter, the sixth quadrate or very slightly transverse ; 

 clava hardly twice as long as broad, slightly longer than the combined length of the two pre- 

 ceding funicular segments ; sensilla usually numerous and in two irregular rows on at least the 

 proximal segments of the funicle, in small females sparser and in only one row. 



Thorax like that of albipennis, but with the pronotal collar a little longer and rather more 

 abrupt in front ; and propodeum different. Pronotal collar medially from slightly more than 

 one sixth, to slightly more than one fifth, as long as the mesoscutum. Propodeum (Text-fig. 

 377) medially a little less than half as long as the scutellum, rather more strongly produced 

 posteriorly than in albipennis ; median area 1-75 to 1-9 times as broad as long, its panels rather 

 dull, mainly finely and rather uniformly reticulate, with some longitudinal carinulae at the base ; 

 costula hardly indicated ; plicae rather more distinct in their posterior half than in albipennis ; 

 median carina often irregular, weak, or even absent ; callus finely and lightly reticulate. Fore 

 wing much as in albipennis ; row of hairs on lower surface of costal cell widely broken medially ; 

 marginal vein 1-4 to 1 -6 times as long as the stigmal vein ; apical margin of the wing ciliate. 



Gaster short-ovate, about as long as, or slightly shorter than, the thorax, about as broad as 

 the latter, 1-2 to i-6 times as long as broad, acute but not acuminate apically ; basal tergite 

 occupying one third to two fifths of the total length ; last tergite slightly to very distinctly 

 shorter than its basal breadth ; ovipositor sheaths projecting at most very slightly beyond the 

 top of the last tergite ; hypopygium extending about half way along the gaster, but usually 

 overlapped by the tergites and not visible. 



o*. Differs from the female as follows : 



POL i-6 to 1-65 OOL. Antennae with combined length of pedicellus and flagellum 1-3 

 to 1-35 breadth of head ; pedicellus about 1-5 times as long as broad, from half to two 

 thirds as long as the first funicular segment ; flagellum cylindrical, distinctly stouter than 

 the pedicellus ; first funicular segment 1-7 to 2 times as long as broad, following segments 

 slightly shorter though all longer than broad, the sixth about 1-5 times as long as broad ; clava 

 nearly or quite three times as long as broad, hardly longer than the combined length of the 

 two preceding funicular segments ; hairs of flagellum standing out at an angle of about 45 , 

 their length about half the breadth of the segments that bear them. Propodeum slightly 

 longer, its median area less transverse. Fore wing with marginal vein 1 -35 to 1 -4 times as long 

 as the stigmal vein. Gaster oblong-obovate, slightly shorter and narrower than the thorax, 

 with a ventral plica. 



The female of this species is similar to that of albipennis (Walker) in many respects, 

 but differs from it particularly in its much shorter gaster ; other differences are 

 mentioned in the description. 



The male of tripolii differs from that of albipennis in having the flagellum rather 

 longer, the funicular segments relatively more elongate. 



Holotype £. England : Kent, Stone Marshes, 18.vii.1935, emerged from 

 Paroxyna plantaginis (Hal.), in flower-heads of Aster tripolium L. (M. Niblett), in 

 Hope Department, University Museum, Oxford. 



Paratypes. England : Same data as holotype, 3 $, 18.vii.1935, in Hope Dept., 

 University Museum, Oxford ; Dorset, Lodmoor, near Weymouth, <$<$, $$, swept in 

 a salt-marsh, 13. ix. 1962 {Graham), in Graham collection. 



Ireland : Co. Dublin, Swords Estuary, 3 $, 26.viii.1954 (Stelfox), North Bull, 

 several $9, 13.viii.1962, collected in the salt-marsh (Mrs. Healy) ; Co. Wicklow, 

 The Murrough, 2 $, 25.viii.1955 (Stelfox), in Graham collection. 



Aster tripolium was seen in the other localities mentioned, and it is probably the 

 main host-plant of tripolii. 



