126 z. bouCek 



<§. 6-5 mm. As in $ of a darker form, with pronotal band reduced to transverse double 

 spot in middle. Gaster black with pale yellow narrow arcuate cross-band in middle; also basal 

 sternites and unsculptured lateral parts of the following sternites (exposed in allotype) pale 

 brown; apex of epipygium red. For shape of gaster see Text-fig. 146. Hind corners of sixth 

 tergite not produced. Sternites narrow, the first and second without unusual features, not 

 punctate, apical margin of the second produced, arched; the third convex, elongate, exposed 

 part not well delimited; fourth to sixth shallowly depressed, fifth nearly twice, sixth 1-5 times 

 as long as broad; last (seventh) sternite depressed only at base, subquadrate with rounded 

 posterior corners. 



Biology. Host unknown. 

 Distribution. Nigeria, Gabon, Zaire. 



Holotype $, Nigeria: Umudike, 4.1.1951 (/. L. Gregory) (BMNH). 



Paratypes. Nigeria: Ibadan, 1 $ (BMNH). Gabon: Ogoue, Lambarene, 1912, 

 1 $, 1 c? (allotype) (R. Ellenberger) (MNHN, Paris). Zaire: Prov. Maniema, Kindu, 

 1912, 1 $ (L. Burgeon) (MNHN, Paris); Lusambo, Sankuru, 1921, 1 $ (J. Ghesquiere); 

 Lulua, Kapanga, iii. 1933, 1 $ (F. D. Overlaet) (MRAC, Tervuren). 



Within the tricolor-group, L. parvula may be recognized by the absence of the 

 premarginal carina on the pronotum, a very coarse sculpture on the raised median 

 part of propodeum, the absence of the mesal carina dorsally on hind coxa, relatively 

 less swollen hind femur and in the female by the relatively long ovipositor with 

 its furrow extending to the base of the fourth tergite. 



Leucospis tricolor Kirby (Paggregate) 



(Text-figs 151-153) 



Leucospis tricolor Kirby, 1883 : 69, <J. Holotype $, South Africa (BMNH) [examined]. 



Under L. tricolor I classify all specimens which I cannot separate as different 

 species on reliable morphological characters, even in the female sex which normally 

 offers more characters than the males. Some of these specimens look very different 

 but, although their diversity seems to match a certain geographic pattern, in most 

 cases intermediate specimens could be examined which suggest an unusual range 

 of variation. The phenotypic populations probably correspond with certain host 

 species in the region or (less probably in this group) with the aculeate species the 

 relevant form mimics. When better known, some of the forms will possibly be 

 separated as subspecies, but at the present stage I do not regard it justifiable to 

 give them nomenclaturally valid names. 



In colour all these forms have the face, sides of thoracic dorsum, usually in a belt from the 

 pronotal shoulder down to dorsellum, then sides of thorax, coxae and femora and, at least 

 partly, base and sides of gaster red ; some of these parts often with yellowish markings including 

 the rest of the gaster in varying pattern. Morphologically, as mentioned in the key, antennae 

 and genae are relatively short, pronotum with distinct though low premarginal carina, median 

 part of propodeum weakly convex with median carina indistinct, upper half of depression of 

 hind coxa very sparsely punctured and dorsal edge with distinct mesal carina; in female first 

 tergite shorter than the fifth, the fourth with a fine median groove instead of the ovipositorial 

 furrow and its hind margin varying from straight to medially slightly produced, fifth tergite 



