242 
BIRDS or BRITISH BURMAH. 
Berlin Museum I observed a specimen from Mozambique, tbe most southern 
locality in Africa where I can state with certainty that this species is found. 
It remains to point out the characters by which P. roseus and P. om- 
crotalus may be recognized. In the first place, P. roseus is a smaller bird, 
notably so with regard to the bill, which in the male never exceeds a length 
of 16 inches and is usually under 15, measured from the forehead to tip ; 
the female has the bill varying in length from I0"75 to 12 inches and 
never exceeding, even in very large birds, 12*5 inches. P. onocrotalus has 
the bill in the male bird exceeding 17 inches and reaching up to 18, and 
in the female it is usually 14 inches and never, so far as I am aware, less 
than 13 inches. In the second place, P. roseus has invariably twenty-two 
rectrices, P. onocrotalus twenty-four. This is the safest and surest test 
when the tails are perfect, and even when imperfect it is seldom difficult 
to arrive at the correct number. Professor Schlegel has suggested to me 
that, like the Geese, the Pelicans may have a varying number of rectrices ; 
but I am in a position to state that, with regard to the two species I met 
with in Burmah, the rectrices are invariably twenty-two in number, neither 
more nor less. All the specimens in museums, which from their size, 
locality, and other characters should be P. onocrotalus, proved on exami- 
nation to have twenty-four rectrices. Many mounted specimens were 
obviously derived from zoological gardens, and in many cases the tails were 
hopelessly imperfect. The same may be said of living birds, of which I 
have examined nearly twenty ; the tail is usually imperfect, and the base 
of the tail so covered with down that the number of rectrices cannot be 
arrived at with any certainty. Yet, notwithstanding these drawbacks, I 
counted a sufficient number of tails to make certain that P. onocrotalus 
has twenty-four rectrices. A third point is the coloration of the imma- 
ture birds. In P^ roseus the young are chestnut below, and a tinge of this 
colour is retained by even old birds ; in P. onocrotalus the lower plumage 
in immature birds is brown of various shades, not rufous or chestnut. A 
fourth point is, that in P. roseus in the breeding-season the forehead is not 
swollen to any appreciable extent ; in P. onocrotalus it is swollen to such 
an extent as to form a remarkable feature in the bird^s appearance. A 
fifth point is one on which I do not wish to lay too much stress, for I may 
be wrong ; it is, that in P. roseus the adult male is not crested. 
P. longirostris , Hume (S. F. v. p. 491), belongs to the P. roseus type of 
Pelican, where the frontal feathers terminate in a point. I have not been 
able to examine Mr. Hume^s type, and it is useless to hazard an opinion 
about it; but I would remark that the length of the bill (18*1 inches) 
is by no means too long for a male P. onocrotalus^ . 
* I take this opportunity of returning my thanks to the authorities of the museunas and 
zoological gardens at Paris, Frankfort, Strasbourg, Berlin, Hamburg, Amsterdam, Leyden, 
Rotterdam, the Hague, and iintwerp, for their courtesy in allowiLg me to examine, in the 
fullest manner possible, all specimens of Pelicans, both living and stuffed, under their charge. 
