DESCRIPTION OF A NEW SPECIES OF KROMBEINIUS (HYMENOPTERA: PERILAMPIDAE) FROM THE PHILIPPINES, AND THE PHYLOGENETIC RELATIONSHIPS OF THE GENUS*

The genus Krombeinius (Hymenoptera: Perilampidae) was re-cently described (Bou6ek 1978) to include perilampids with an amalgam of the characters of Euperilampus Latreille and Perilampus Walker. The habitus suggests Euperilampus, and there are two synapomorphies to unite these two genera (Darling 1983): postspiracular sclerite reduced to a narrow triangle, less than one-half as wide as the adjacent pronotum; and pronotum massive, at least one-third the length of the mesoscutum. However, Krombeinius exhibits the wing venation, presence of a marginal rim on the scutellum, and large third metasomal tergite characteristic of Perilampus. plesiomorphic similarities. of Euperilampus, symplesiomorphy.

synapomorphies to unite these two genera (Darling 1983): postspiracular sclerite reduced to a narrow triangle, less than one-half as wide as the adjacent pronotum; and pronotum massive, at least one-third the length of the mesoscutum. However, Krombeinius exhibits the wing venation, presence of a marginal rim on the scutellum, and large third metasomal tergite characteristic of Perilampus. regard these as plesiomorphic similarities. The genus is characterized by the absence of the defining apomorphic characteristics of Euperilampus, i.e., by symplesiomorphy.
In this paper present new information on the structure of the male genitalia and labrum of the type species of Krombeinius. These structures have proved to be of considerable value in defining genera in the Perilampidae (Darling 1983). From this analysis suggest autapomorphies for defining Krombeinius. In addition, describe a new species of Krombeinius from the Philippines, and discuss the affinities of the three included species.
Taxonomic studies of Krombeinius have been hampered by t-he scarcity of material. The type species, K. eumenidarum, was described by Bou6ek (1978) from a series of specimens (2 male, 2 female) reared from the larvae of an eumenine wasp in Sri Lanka. All specimens were prematurely killed and had to be liberated from the pupal cuticles, producing some abnormalities in the type material. Also included by Bouek (1978)  Hymenoptera: Chalcidoidea: Perilampidae (sensu Graham, 1969). Species of Krombeinius can be distinguished from Monacon Waterston, Burksilampus Bou6ek, Steffanolampus Peck and Perilampus Latreille by the narrow postspiracular sclerite, less than one-half the width of the adjacent pronotal panel, and from Euperilampus by having the marginal vein longer than the postmarginal (Fig. 1).
All known species are moderately large, 3 to 51/2 mm long, black without metallic reflections and are restricted to the Oriental region. There are three species currently placed in Krombeinius: K. eumenidarum Boucek, K. megalaspis (Cameron) and K. saunion, n.sp.
A revised key to the species of Krombeinius is not presented. The key of Bou6ek (1978) separates K. eumenidarum and K. megalaspis. An additional character to separate these two species is the inner orbits: costate in K. eumenidarum (Fig. 8), and smooth in K. megalaspis (Fig. 12). K. saunion is readily distinguished from these two species by the prominent spine at the apex of the scutellum (Fig.  1,15). The apex of the scutellum is truncate in the other two species (Figs. 7,11

Description"
Head: supraclypeal area smoothly convex, without horn or ridge; scrobal cavity deep, extending to lower ocular line or to middle of clypeus; clypeus and supraclypeal area separated by distinct suture or by faint line; inner orbits carinate; frontal carina separating the median and posterior ocelli; malar sulcus absent; malar region with strong oblique costae; posterior ocellus located high on vertex, POL approximately equal to OOL; labrum with a single narrow stalk, expanded distally with 7 digits, each with a terminal seta, and with pair of sessile setae not associated with digits, strongly excised medially [n 1, K. eumenidarum, Fig. 3].
Mesosoma: dorsum of pronotum smoothly convex, without transverse elevations; pronotum massive, about one-third length of mesoscutum, not narrowed medially; mesothoracic spiracle located between pronotum and sidelobe of mesoscutum; postspiracular sclerite fused to the pronotum but delimited by surface sculpture; postspiracular sclerite less than one-half width of adjacent pronotal panel, with many or a single puncture; notauli distinct; scutellum vaulted, jutting over propodeum and base of metasoma; apex of scutellum"acuminate, or truncate or with a distinct spine; propodeum with median area foveate, or witha short median ridge, sub- Metasoma" petiole short, transverse, the tergum forming a ridge along anterior face of gaster, sternum shifted posteriorly; gaster triquetrous, T2 and T3 fused, covering most of dorsum; T2 without distinct basal fovea; T3 much longer than T2, subquadrate, slightly wider than length along midline; ovipositor ventral, not upturned, sheaths not distinctly exserted; male genitalia with distinct parameres In 1, K. eumenidarum, Fig. 2]. Discussion: The male genitalia of Krombeinius eumenidarum (Fig. 2) are similar to those of species of Perilampus: the parameres are distinct, and strong setae are distributed on these lobes. This configuration occurs throughout the Chalcidoidea (see Domenichini 1953) and is regarded as plesiomorphic. In Euperilampus a derived condition is found (Darling 1983): distinct parameres are lacking, and the basiparamere has a patch of strong setae distributed on transparent areas laterad of the ventral lobe (Fig. 4). The labrum of Krombeinius eumenidarum (Fig. 3) has a narrow central stalk, not found in other perilampid genera (Riek 1966;Domenichini 1969;Darling, unpublished). However, the labrum does share synapomorphies with species of Euperilampus (Fig. 5) including a reduced number of digits (7 or 8), a pair of smaller, sessile setae not associated with digits, and a strong median excision. The narrow stalk distinguishes the labrum of Krombeinius from those of Euperilampus, and is postulated as an autapomorphy of Krombeinius. All other perilampid labra are 10-12-digitate, and not as strongly excised medially. Etymology: The specific epithet is a noun in apposition, Greek for "javelin", and is a reference to the elongate spine on the scutellum of this species.

Diagnosis:
This species can be immediately recognized by the prominent spine at the apex of the scutellum (Figs. 1, 15). The apex of the scutellum is truncate in K. eumenidarum (Fig. 7) and K. megalaspis (Fig. 11 Krombeinius saunion is more closely related to K. eumenidarum than to K. megalaspis. Synapomorphies of these two species are: the stigmal vein making a right angle with the marginal vein (oblique in K. megalaspis and outgroup: Euperilampus and Perilampus); clypeal-supraclypeal margin weak or indistinct (separated by distinct suture in K. megalaspis and outgroup: Euperilampus and Perilampus); lateral pronotal collar suggesting bumpy shoulders (regularly convex in K. megalaspis and outgroup: Euperilampus and Perilampus); and postspiracular sclerite with many foveae (a single fovea is found in K. megalaspis, and in the ancestral species groups of Euperilampus, Darling 1983). Considering Euperilampus as the outgroup, the following similarities of K. eumenidarum and K. saunion are regarded as plesiomorphic: propodeum medially about twice as long as metanotum (equal to metanotum in K. megalaspis; autapomorphy); scutellum, in lateral view, not strongly convex, tapering gradually towards the apex (highly convex in K. megalaspis, Bou?ek 1978: Fig. 2; autapomorphy).

1982]
Darling--New Species of Krombeinius 315 K. saunion and K. eumenidarum also have the inner orbits with strong costae (Figs. 8,9,16,17), whereas the inner orbits of K. megalaspis are smooth (Figs. 12,13). consider the costate inner orbits to be a synapomorphy of Euperilampus + Krombeinius. As such interpret the smooth orbits of K. negalapis as an autapomorphic reversal. A similar reversal in this character is indicated in the Euperilampus cladogram (Darling 1983).
There remain some difficulties in justifying the current composition of the genus Kronbeinius. The numerous characters separating Figs. 14-17. Krontbeinius saunion. 14. Head,dorsal. 15. Mesosoma,dorsal. 16. Head,frontal. 17. Head,lateral. [Scale line mm.] Psyche [Vol. 89 K. eunlenidarunl + K. saunion from K. megalaspis question the inclusion of K. megalaspis. A revised classification would create a monobasic new genus for Perilantpus megalaspis, and would allow Krotbeinius to be defined by the synapomorphies of K. eumenidarutl + K. saunion. Recalling that the proposed synapomorphies of Krombeinius (structure of the labrum; host association) are not known for K. megalaspis, it would not be surprising if this species were to be excluded at some later date. Clearly, more material and associated biological information are essential to re-evaluate the composition of Krombeinius, and any nomenclatural changes at this time would be premature.