^Ir. J. U. ^lallocli on E.votic ]\Iuscari(Uo. 073 



This species runs to antarct'ica in the key already referred 

 to, but is distinguished at ouee by the eohjur of tlie antenna} 

 and |)alj)i, which in the former are entirely l)lack, only the 

 palpi being rarely paler at bases. 



Melatiochelia ienuicornis, sp. n. 



Female. — Very similar to ordiJiota, Ilutton, in the structure 

 of the head and antenn<e. Dilfers very strikingly in colour, 

 the pruinescence on thorax and abdomen l)eing bluish grey 

 instead of brownish grey, the abdomen especially appearing 

 bluish. The pruinescence on orbits and parafacials is brown 

 except at base of antennae where it is white, whereas in ordi- 

 nata it is entirely yellow, and the upper half of the cheeks 

 is rufous instead of yellow. The thoracic vittse are narrowly 

 fused in front of suture, just behind suture, and again beyond 

 middle, tlie median spot on scutellum is as large as the lateral 

 spots and covers most of the disc basally instead of being 

 very small. Each abdominal tergite from 1 to 3 has a pair 

 of moderate-sized subtriangular brown spots, and the fourth 

 lias sometimes a brown spot in centre at base. The tibige 

 are rufous and the mid-pair have an antero-dorsal bristle. 

 Third antennal segment at least four times as long as 

 second ; the longest hairs on arista are noticeably longer 

 than its basal diameter, whereas in ordinata they are not as 

 long, the eyes are hairy, and the frons is at least three times 

 as wide as the slender third antennal segment. The lower 

 calyptra and the knobs of halteres are fuscous, not yellow. 

 Chsetotaxy as in ordinata. 



Length 8-9 mm. 



T^^pe and one paratype, Crofton, New Zealand, 28. x. 

 190A(?F. Wesche), 



This species was compared with the type-specimen of 

 ordinata^ Hutton. 



Genus Exsul, Hutton. 



This genus was erected for the reception of one species, 

 singularis, Hutton. Stein omitted it from his ' Catalogue of 

 the Genera of AutliomyiidtC of the World ^ in 1919, probably 

 because he had some doubt as to its relationships. I have 

 seen the genotype and type-specimen, as well as three other 

 males from New Zealand, and find that it is closely related 

 to Lispodes,^lR\\och. The basal abdominal sternite is bare, 

 the fore and mid femora and tibi?e are armed on their 

 ventral surfaces with short stout spines, and the wings are 

 abnormally broad, being about as wide as long. 



The wing-characters alone serve to separate singularis 

 Ann. cO Mag. N. Hist. Ser. 9. l\>l. xi. -ii 



