Demoulinia insularis Lugo-Ortiz & McCafferty 1998

(Figs 6–7, 24–25, 38, 44, 57)

Demoulinia insularis Lugo-Ortiz & McCafferty 1998: 360 (larva).

Demoulinia assimilis Gattolliat 2003: 2 (larva), syn. n.

Material. The holotype of D. insularis is deposited in Purdue Entomological Research Collection, West Lafayette, Indiana, USA (Lugo-Ortiz & McCafferty 1998: 357) and is a mature female larva ready to moult to subimago; certain parts of it are slide-mounted (Lugo-Ortiz & McCafferty 1998: 360), and the remaining parts are preserved in alcohol; Luke Jacobus kindly reexamined this specimen and made pictures (Figs 6–7, 24–25, 38, 44). The holotype and the paratypes of D. assimilis are deposited in the Museum of Zoology, Lausanne, Switzerland (Gattolliat 2003: 2); Jean-Luc Gattolliat kindly reexamined certain details, prepared an additional slide of mature male larva and made pictures (Fig. 57).

Synonymy of D. insularis and D. assimilis. The original description of D. insularis was based on a single larva. The original description of D. assimilis was based on 21 larvae (Gattolliat 2003). As the differences between D. assimilis and D. insularis, Gattolliat (2003) reported three characters: the setation of the dorsal surface of the labrum, the shape of the lingua and the number of the segments of the maxillary palp.

Labral setation of D. assimilis was described as «subapical row of about 11 long stout setae subparallel to the distal margin, a few long fine setae scattered over surface» (Gattolliat 2003: 2, fig. 1), while labral setation of D. insularis was described as «numerous long, fine, simple setae scattered over surface» (Lugo-Ortiz & McCafferty 1998: 360, fig. 11). The South African species D. crassi demonstrates individual variability of this character: in some individuals long setae are evenly scattered over surface (left side of Fig. 11), in others they form a more or less distinct subapical row (Fig. 9); this fact allows the assumption that the Malagasy species can have the same individual variability.

The hypopharynx of D. assimilis was described as having «a broad trilobate superlingua, covered only with minute setae; lingua weakly developed laterally, apically with long thin setae» (Gattolliat 2003: 2, fig. 2); here the term «superlingua» is erroneously applied to hypopharynx (= lingua), and the term «lingua» is erroneously applied to superlingua. The hypopharynx of D. insularis was not characterized, and only figured in a resolution which does not show its setation (Lugo-Ortiz & McCafferty 1998: fig. 12). The holotype of D. insularis is a larva ready to molt to subimago, so its mouth parts are represented by empty cuticle; in such a condition the hypopharynx, which has a very thin cuticle, is crumpled: left lobe of the hypopharynx (= lingua) is compressed, but the right lobe retains its shape (Fig. 25); it shape does not differ from that of D. assimilis (Gattolliat 2003: fig. 2) and D. crassi (Fig. 23).

The maxillary palp of D. assimilis was described as «3-segmented» (Gattolliat 2003: 4, fig. 5), while the maxillary palp of D. insularis was described as «two segmented» (Lugo-Ortiz & McCafferty 1998: 360, fig. 15). However, the holotype of D. insularis (which is the only specimen of the type series) has one maxillary palp broken; another palp is preserved, mounted on a slide and has a clearly expressed third segment (Fig. 24) whose shape is the same as in D. assimilis (Gattolliat 2003: fig. 5) and in D. crassi (Figs 21).

Thus, not one of the three reported characters allows D. assimilis to be distinguished from D. insularis. A new synonymy is proposed here: Demoulinia insularis Lugo-Ortiz & McCafferty 1998 (= Demoulinia assimilis Gattolliat 2003, syn. n.).