TACHINIDAE OF ORIENTAL REGION 75 



Possibly allied to Torocca, and certainly sharing many characters with it, is the 

 genus Zambesa Walker, of which the affinities remain rather uncertain. Nothing 

 is yet known of the hosts of Zambesa, which will be specially significant in determin- 

 ing the phyletic relationships of the genus. Because of the 'closed' fully sclerotized 

 posteroventral declivity of the thorax, and a generally similar facies, both Townsend 

 and Malloch (various publications) regarded Zambesa as a genus of Cylindromyiini 

 (Phasiinae), but Verbeke's (19626) study of the curious male genitalia showed that 

 Zambesa is undoubtedly not a Phasiine and is probably related to Acemyini, Thrixion 

 Brauer & Bergenstamm or Polygastropteryx Mesnil. External resemblance between 

 Zambesa and Polygastropteryx is specially close, as is evidenced by the fact that 

 Mesnil inadvertently redescribed his P. bicoloripes as Z. setinervis (Mesnil, pers. 

 comm., has informed me of this synonymy), and it appears most appropriate at this 

 stage of knowledge to place Polygastropteryx and Zambesa in the same tribe. On 

 practical grounds these genera fit best, as an interim measure, in the Thelairini 

 near to Torocca, but such a placement might prove erroneous when the early stages 

 and host relations of Polygastropteryx and Zambesa become known. Mesnil (1966 : 

 888) proposed the monogeneric subtribe Zambesina for Zambesa but it seems doubtful 

 whether this is warranted, at least at present, and I do not accord status to this 

 segregate in the present work. 



Zambesa contains two known valid species (with some names as synonyms) 

 to which Malloch (19326 : 329) provided a key. Not all the characters given by 

 Malloch are fully satisfactory, and the following new key is given to differentiate 

 the two species. 



Wing cell i? 5 open or closed about at the wing margin. Cross-vein r-m situated distad 

 of the apex of Sc. Palpi clear pale yellow. Abdominal T3 of 5" longer than its 

 breadth, in lateral view conspicuously longer than Ti + 2, and with the paired 

 yellow spots nearly twice as long as wide. Abdominal sternite 5 of <J very weakly 

 bilobate and without a median process (fig. 11 in Malloch, 19326) 



Z. claripalpis Villeneuve 



Wing cell R b closed and conspicuously petiolate, the petiole about equal in length to 

 r-m. Cross-vein r-m situated slightly but clearly basad of the apex of 5c. Palpi 

 dark tawny brown to blackish. Abdominal T3 of <J broader than long, in lateral 

 view subequal in length to Ti + 2, and with the paired yellow spots not much 

 longer than wide. Abdominal sternite 5 of <J deeply bilobate and with a long fine 

 pointed process between the lobes (fig. 10 in Malloch, 19326). Z. ocypteroides Walker 



Finally regarding Zambesa it is noted that Townsend's (1938 : 174) inference that 

 Malloch misidentified Z. ocypteroides Walker is incorrect. 



Key to Oriental Genera of THELAIRINI 



Scutellum with three or more pairs of marginal setae (basals, subapicals and apicals, 

 sometimes laterals in addition). Posteroventral declivity of the thorax mem- 

 branous medially ............ 



Scutellum with two pairs of marginal setae (laterals and subapicals), basal setae 

 absent or hair-like and apical setae absent (the lateral and subapical setae enormous, 

 divergent and subequal in size: Text-fig. 74). Posteroventral declivity of the 

 thorax often fully sclerotized across its width ....... 



