232 
Annals of the Transvaal Museum 
Myrmecocichla which inhabits wide stretches of open country. I propose 
to place it, therefore, in a new genus bearing the name of Sciocincla, 
genotype Saxicola arnotti Tristram. Ogilvie-Grant (Ihis, 1908, p. 299) has 
shown that Sciocincla arnotti differs from nigra in having the crown white 
in the male and the throat white in the female, instead of being entirely 
black in both sexes, and in having the upper wing-coverts white tipped 
with black, instead of pure white; he places as synonyms of arnotti the 
forms about which Reichenow was in doubt, shelley, leiicolaema a.nd collaris ; 
but more recently Hartert (Bull. Brit. Orn. Cl. xxxvii. 42, 1917) has 
revived the name of leucolaema for the East African birds which he regards 
as a subspecies of arnotti, differing in size only and not occurring within 
our limits. Hartert gives the range of Sciocincla arnotti as extending as far 
south as Griqualand East ; but this must surely be an error, as it does not 
occur south of the Lydenburg, Transvaal, low country. 
The Melsetter Robin, described by Shelley (Bull. Brit. Orn. Cl. xvi. 125, 
1906) under the name of Erythracus (= Erithacus) swynnertoni, differs in 
several important respects from all the northern species commonly associ- 
ated in the genus, some of which have already received generic names, and 
it may therefore be placed in a new subgenus bearing the name of Swyn- 
NERTONIA, in honour of Mr C. F. M. Swynnerton, who is not only responsible 
for the discovery of the species but has also contributed very largely to 
our knowledge of our tropical species of birds. The wing formula of the 
genus is : second primary much shorter than the secondaries and the tenth 
primary, the third a little shorter than the seventh and the fourth, hfth 
and sixth primaries subequal; the bill is black and the legs and feet are 
pinkish; three conspicuous bristles are present in the specimen examined 
and the tail is relatively short as compared with other species. 
The other Robins of Africa are in need of careful revision, and although 
the eggs of some species are variable, even the variability is limited to 
colour or shade of colour, and this character will probably prove useful in 
studying the phylogeny of the species. In Cossypha, for example, the two 
yellow species, hicolor and natalensis, which have pale legs, have uniform 
eggs varying from greenish or olive to dark brown, whereas C. cajfra and 
Tarsiger stellatus have hnely spotted eggs that are hardly distinguishable, 
though their nests are distinct. There appears to be no description extant 
of the eggs of C. humeralis, but the species is so entirely different that we 
may safely allocate it in a distinct genus, for which the name of Bessonornis 
A. Smith is available. According to Mathews and Iredale (Austral. Av. Rec. 
III. 43, 1915) the name of Cossypha should be replaced by Bessonornis, 
as it is preoccupied by Cossyphus, a procedure which it may not be necessary 
to follow; and according to the same authors the name of hicolor Sparr- 
mann (1790) must give place to dichroa Gmelin (1789). I propose to change 
the name of Cossypha to Caffrornis for the typical species with dark legs 
and feet; and consider it advisable to separate Cossypha dichroa and C. 
natalensis, applying the generic name of Hyloafdon to the former and 
placing the latter tentatively in the same genus. The species having dark 
legs and feet and a white eyebrow occupy different classes of country, 
C. caffra occurring in the south and west, and in the north at higher levels 
only, C. heuglini occurring in the tropical country only; and in the same 
way Hyloaedon dichroa occurs only in the south of the Tropic of Capricorn, 
