Annals or the Transvaal Museum. 
255 
seen the specimens vindication is unnecessary. The fact that the two 
species are both found in Kaffraria counts for nothing, since C. superciliosus 
and C. burchelli are said to occur side by side north of the Zambesi. 
At p. 421, Grant has indicated two errors on my part in the following 
terms : — “ Attention must be drawn to the fact that Mr. Austin Roberts 
recently recorded a pair of Centropus grilli caeruleiceps from Sabi, 
eastern Transvaal (cf. Annals of the Transvaal Museum , Yol. IV, 1914, 
p. 175) ; considerable confusion is entailed by taking a name, and that 
only a subspecific (geographical) one, from north-east Africa and fixing 
it on to a South African bird, thus defeating the laws governing 
geographical forms, and creating unnecessary synonyms. 
“ Mr. Roberts has remarked that the $ differs from the in being 
banded ; this is, of course, the character of the young and immature birds. 
When adult the sexes can only be distinguished by the female being slightly 
larger and not by the markings, as is shown by nine sexed birds in the 
British Museum Collection from Natal and Nyasaland in black and rufous 
plumage, six of which are $$ and three A A” 
Perhaps it is as well that I stated the § was banded, as we now know 
what is the difference between the sexes when adult, as Grant has carefully 
pointed out, a difference in size pointed out by me, which has apparently 
previously been missed ; while Grant himself states that the type of 
Centropus grilli caeruleiceps is “ very closely matched by a bird from Port 
Natal.” Why then all this furore ? As a matter of fact the paragraph 
reflects upon the work of Grant’s colleague, for whom it was apparently 
not intended, as I find at p. 286, in dealing with the single specimen on which 
Grant founds his name of Irrisor erythrorhynchus ruwenzorae , he states 
that : “ In the Trans. Zool. Soc., Yol. XIX, 1910, p. 432, Mr. Ogilvie-Granfc 
has recorded this specimen as I. viridis Licht.” It cannot therefore be said 
that I am the only sinner in recording a race incorrectly, more especially 
having regard to the fact that Ogilvie-Grant himself stated nine years 
previously that Irrisor viridis is confined to the Cape Colony. It is surely 
a graver mistake to describe a new race of Centropus grilli on a single 
immature specimen, as he has done at p. 420, when, as he admits, adult 
specimens are not available from the type locality ; would it not have been 
better to have recorded this specimen as Centropus grilli caeruleiceps , seeing 
that it so closely matches the type ? What would he have said had I 
described the Sabi specimens as a new race ? And had he rejected it, 
he would have been justified in doing so only on the grounds that the 
“ phases ” of plumage are not well known enough to warrant the separation 
of the species into races. Moreover, Grant is himself not above reproach 
in the “ creation of unnecessary synonyms,” as, for example, in his 
rejection of species, such as Lophoceros damarensis and subspecies of which 
he has not seen the specimens or series on which they were founded, or, 
if he has seen them, has neglected to examine them with proper care. 
I am puzzled as to what Grant means by “ laws governing geographical 
forms.” Has he discovered the laws by which forms vary geographically ? 
If so, why has he not published his views thereon, for this important 
question has not yet been satisfactorily answered. But I fear, judging 
by his work, that he referred to some unwritten custom which he is now 
