950 REPORT OF THE HARVARD AFRICAN EXPEDITION 
rowed apically, rather suddenly ending in an appendage which is three times less 
wide and nearly five times less long and consists of the three terminal divisions, 
closely packed together.”’ At first Surcouf included in it only two Indian species, 
H. pachycera Bigot and H. rubida Ricardo. Later he also referred to it H. cilipes 
Bigot, although this has the third antennal segment less swollen; but a specimen 
of that species, at the U. S. National Museum, shows that it could not possibly 
be placed in Potisa: the antennae are unusually long, the first segment being 
longer than the second and third together and a little incrassate in the basal 
half, while the basal division of the third is elongate, not at all disciform, and 
merges gradually in the three terminal divisions. On the other hand, H. singu- 
laris Ricardo, of Annam, which I have seen at the U. S. National Museum, has 
antennae like those of H. pachycera. Some other Oriental species, such as H. 
lata Ricardo, H. splendens Schuurmans-Stekhoven, and H. conflwens Schuurmans- 
Stekhoven, form the passage between Potisa and the more usual type of Hae- 
matopota. 
Obviously if the shape of the antennae of the Haematopotini were to be 
accorded more than specific value, many more generic or subgeneric names would 
have to be added to those proposed by Griinberg, Surcouf, and Enderlein. 
Key to THE Species or Haematopota or THE BRLGIAN Conco 
It is very difficult to devise a satisfactory key to the many species of African 
Haematopota, several of which are quite closely allied or superficially alike. In 
the following attempt, I have covered thirty-two species of which I have had 
specimens before me and I have added H. cruenta from the description. I have 
not seen or not recognized in my material H. pellucida (Surcouf), H. laverani 
Sureouf, H. brunnipennis Ricardo, and H. torquens Austen, which I have not been 
able to include in my key from the descriptions alone. Their affinities will, how- 
ever, be discussed in the sequel. 
I have thought it advisable to mention as many characters as possible of each 
species and not merely to give a “‘skeleton key,” as is now the fashion in entomo- 
logical work. This is the more necessary because the territory covered possesses 
several additional species which have not yet been properly recognized and some 
of which may be undescribed. I have in my collection some half-dozen such 
unnamed species. 
The key is entirely based upon females and should be used only for perfectly 
preserved specimens. ‘The length is measured from the frontal callosity to the 
tip of the abdomen and does not include the antennae. 
Haematopota vandenbrandent was a manuscript name applied to a species of 
Sankisia, Katanga, which was found infected with an intestinal flagellate, Crithi- 
dia tenuis Rodhain, Pons, Vanden Branden, and J. Bequaert (1913, Bull. Soc. 
Path. Exot., VI, pp. 182-184). This species has never been described and, as 
specimens are no longer in existence, it should be eliminated from the list of 
African tabanids. 
1. Femora and tibiae with dense, long hairs; those of the hind tibiae longer than the width 
of the tibia and forming a conspicuous fringe on the outer margin; tibiae wider than 
