Ventral part of male terminalia, which evolved from a fusion of S9 and the gonocoxites. The tips of the U-shaped hypandrium articulate with the epandrium. (Cumming, Sinclair & Wood 1995)
In the stem-species pattern of the Chloropidae family-group, as represented in the Acartophthalmidae, Carnidae, and Milichiidae, the hypandrium is U-shaped and the subepandrial sclerite forms a single plate, connecting the arms of the hypandrium anteriorly.
In the stem-species pattern of the Chloropidae, the hypandrium is probably fused anteriorly into a ring according to Andersson (1977) and Nartshuk (1987). However, according to both these authors, it is also possible that the ring-shaped hypandrium evolved within the Chloropidae. 
The stem-species pattern condition of the subepandrial sclerite has not been studied. It either forms a single plate, or apomorphically it forms two lateral rods leading from the articulation of the hypandrium with the epandrium to the inner side of the surstyli.
Within the Milichiidae, in Stomosis, the hypandrium is fused into a ring, and the subepandrial sclerite forms two rods which are connected anteriorly by a narrow sclerotised band.
If there is a ring-shaped hypandrium present in the stem-species pattern of the Chloropidae, as suggested by Andersson and Nartshuk, then it is possible that this holds true for the Milichiidae too. If this is the case, the shape of the hypandrial complex as in Stomosis would represent the stem-species pattern of the Milichiidae. This would mean though that a reduction of the ring has taken place four times. It is more parsimonious to assume that in the stem-species pattern of the Milichiidae the arms of the hypandrium are free and were apomorphically fused in Stomosis. In this case the ring-shaped hypandrium evolved convergently in Stomosis and in the Chloropidae.
In my opinion there is possibly a functional correlation between the shape of the hypandrium and the shape of the subepandrial sclerite. A connection between both tips of the hypandrium is probably important for the stability of the genitalia. This connection is ensured either by lengthening and fusion of the hypandrial arms into a ring, or by a connection of the hypandrial arms with the anterior margin of the subepandrial sclerite. In cases where a ring is present, as in Stomosis and Apotropina ?fuscipleuris (Chloropidae), the subepandrial sclerite can be split into two rods, which may only be connected by a narrow band. This hypothesis has to be checked for Chloropidae. (ex Brake 2000)