122 



Plate XII.). Occasionally the borders of the rosettes are compound, 

 as in Hcematopota coronata, Austen (Plate XII., fig. 91), while in 

 H. ugandce, Ricardo (Plate XII., fig. 93), the centres of the rosettes 

 are usually filled out with a different shade of colour : in species 

 such as H. fidva, Austen, and H. denshamii, Austen (Plate XI., 

 figs. 84, 85), in which the dark colour in the wing is for the most 

 part confined to the apex and hind border, remains of at least the 

 distal rosette are generally distinguishable, while faint traces of one or 

 both of the other rosettes can often be seen. Other useful characters 

 for the distinction of species presented by the wings are to be found 

 in the markings of the discal cell and of the wing-tip, the nature 

 of the stigma (the more or less dark, elongate and somewhat 

 thickened spot below the end of the first longitudinal vein), and the 

 presence or absence of a large, unbroken, dark blotch beneath it. 

 The markings on the front and hind tibiae (if present) should also 

 be noted, while additional characters are afforded by the shape of 

 these joints (whether swollen or non-swollen), by that of those of 

 the antennae,* and by the shape and coloration of the frontal callus 

 (the raised and shining area on the head between the eyes, above 

 the base of the antennae). 



The majority of species of Hcematopota are somewhat sombrely 

 coloured flies, since, in addition to the dusky hue of the wings, 

 the coloration of the usually narrow, elongate body is generally 

 some shade of brown or grey, often relieved with markings, such as 

 longitudinal stripes on the thorax or a double row of spots on the 

 abdomen (sometimes with the addition of an edging to the abdominal 

 segments and a series of median dorsal triangles), in a lighter shade 

 of grey. The eyes in the living insects are usually shining coppery 



* The shape of the first and third joints of the antennae varies greatly in different 

 species of Hcematopota, for whicli reason it is doubtful whether Parhcetnatopoia, 

 Griinb. (founded, as already stated, on these characters for P. cognata, Grunb. — 

 German East Africa and Zanzibar), can be maintained. Should Parhcematopota, 

 Griinb., be recognised as a valid genus, Hcematopota brunnescens, Ricardo (Uganda 

 and Nyasaland Protectorate), would also have to be assigned to it. Speiser is, 

 however, mistaken in stating (Zeitschrift fiir syst. Hymenopterologie m. Dipterologie, 

 VII. Jaln-g., 1907. p. 360), that H. vittata, Lw., is a Parhcematopota, since in this 

 species, as also in the closely allied H. pulchrithorax, Austen, and others, the first 

 joint of the antonnsB is elongate cylindrical, and its upper angle is not produced, 

 although there is a notch on the upper side just before the tip. 



