REVISION OF LEUCOSPIDAE 69 



lower tooth of the mandibles is rather deep (Text-fig. 42) and the hind coxa shows 

 dorsally a low blunt lobe, whilst in the second subgroup there is a distinct tooth. 

 To the second subgroup belong L. egaia Walker, L. coxalis Kirby, L. pulchriceps 

 Cameron, L. colombiana sp. n., L. signifera sp. n., L. opalescens Weld and L. aliena 

 sp. n. The species come from Central and South America; their hosts are not 

 known. 



Leucospis bulbiventris Cresson 

 (Text-figs 42, 64-67) 



Leucospis bulbiventris Cresson, 1872 : 29-30, <$. Holotype <$, Mexico (ANS, Philadelphia) 

 [examined]. 



Weld (1922 : 4) placed this species with a query into the genus Polistomorpha 

 Westwood, but that was not correct. Although only one male is known, L. bulbi- 

 ventris apparently belongs to the egaia-group and within this is nearest to L. 

 manaica Roman. From most species of the egaia-group Cresson's species differs 

 mainly in the aberrant form of the male gaster, the relatively less dense puncturation 

 of the body (mainly on the pronotum and hind femur) and in the relatively longer 

 lower tooth of the mandibles the form of which slightly reminds one of the cayennensis- 

 group; in the latter the notch is semicircular, not triangular. I find similar form of 

 the mandible in the Venezuelan female which I classify as L. manaica Roman. It 

 may be even possible that L. manaica is the female sex of L. bulbiventris, however 

 unusual and great the difference between them may seem. I am unable to resolve 

 the question from the poor material available. 



Biology. Unknown. 



Distribution. Mexico. 



Leucospis manaica Roman 



(Text-figs 68-71) 



Leucospis manaica Roman, 1920 : 9-10, fig. ia, $. Holotype <j>, Brazil: State Manaos, Rio 

 Negro W. of Sitio Cataporanga (NR, Stockholm) [examined]. 



I have had difficulties in understanding the variation of this species, mainly because 

 its holotype is slightly different from the few other available specimens. At one 

 stage I regarded the females from Santa Catarina as a different species, with the 

 gaster distinctly more narrowed at apex, the body more densely punctured and 

 hind femur relatively narrower (Text-fig. 71) than in the holotype of L. manaica 

 (Text-fig. 69), although the latter was relatively well described and figured by 

 Roman. Quite recently I was able to examine another female, from Venezuela, 

 which is rather intermediate in the form of the gaster but is, on the other hand, 

 clearly more pubescent on the sides of thorax, on the propodeum and on the gaster 

 than any other specimen. This seems to suggest that the range of variation is 

 wider than I thought before. 



