244 
Annals of the Transvaal Museum 
and one-fifth (120 per cent.), the graduation of the feathers equal to nearly 
half the length of the wing. 
In East Africa a longer-tailed group has evolved, apparently from 
Fiscus, but differs therefrom in having a bunch of stiff, curved, bristles 
on the forehead, well illustrated in the Journal fur Ornithologie, 1869, 
PL III. The first primary is about two-thirds (66 per cent.) of the length 
of the second, which is shorter than the seventh, while the fourth, fifth and 
sixth are subequal and longest ; the tail is very long, nearly one and a half 
times the length of the wing, the graduation of the feathers equal to more 
than half the length of the wing, and narrow as in Fiscus. For this sub- 
group I propose the subgeneric name of Neofiscus, genotype Lanius 
caudatus Cabanis. 
Enneoctonus collurio represents a very distinct genus, differing in colour 
and structural characters from the true shrikes. Its tail is much shorter 
than the wing (about 83 per cent.), the feathers much narrower than in 
Lanius, the outermost being in breadth little more than half of and the 
graduation less than the length of the culmen. Although it migrates for 
about the same distance as Lanius minor, its diet is more strictly insecti- 
vorous, and its habits calling for less speed these are reflected in the shape 
of the wing, the first primary being about one-third (33 per cent.) of the 
length of the second in those individuals which migrate farthest south. 
I have not seen a specimen of L. guhernator of West Africa, which some 
authorities have associated with this genus; but I suspect that it is more 
closely allied to L. minor than Enneoctonus, having perhaps developed in 
the same way as L. vittatus of Southern Asia. The Wood Shrike apparently 
represents a distinct genus, Phoneus, which is confined to north of the 
Equator. 
The Asiatic Shrikes, which do not occur in South Africa, are some- 
times separated from Lanius under the name of Otomela, typified by 
L. cristatus. I have examined specimens of the genotype and another 
species of the genus and the following characters were noted : first primary 
more than half (55 per cent.) the length of the second, which is about equal 
to the sixth, while the third and fourth are subequal and longest ; the tail 
is of about the same length as the wing, or a little longer, the outermost 
feather in breadth about equal to the length of the culmen, the tips less 
rounded than in Fiscus and the graduation nearly equal to the length of 
the tarsus. As one would expect after comparison of species in other genera 
and families, these characters indicate local migration or evolution from 
a migratory ancestor. Three species are said to occur in North-East Africa, 
whence they would seem to have radiated from Asia. In East Africa there 
is a resident species which has been allocated to a genus by some authors, 
Corvinella corvina, which I have not been able to examine, that may prove 
to have evolved from Otomela in the same way as Neolanius and Neofiscus. 
It has a very long tail, more than one and a half times the length of the 
wing. Sclater, in Shelley’s Birds of Africa (v. 239) has placed Lanius souzae 
of Angola in the genus Corvinella, apparently on colour grounds, although 
the tail is much shorter than in the type species and it can hardly be 
correctly allocated in the genus. All the species of Otomela found in Asia 
and the islands to the south should be more carefully studied and it will 
be found that the local species invariably have longer tails than the 
