Birds of Celebes: Meropidae. 
251 
than those from New South Wales”; nevertheless North’s subsequently recorded 
measurements of Australian specimens (25) do not bear out his former statement: 
New Britain 21.6 x 19.5 mm, a set from Buldery (N. S. W.) average ca. 
21.6x18.3 mm. Count Salvador! (12) was led by North’s remarks doubt- 
fully to refer the New Britain birds to Merops salvadorii M.QjeT^ a species, whose 
validity is still to be proved. 
Turning to Celebes and the Moluccas, it appears that this Bee-eater occurs 
here almost but not quite, entirely in the months when it is absent in 
Australia — that is, in the Australian winter. Meyer (11) writes: “M. orriatus 
Lath, is only numerous during the east monsoon (April — October). Near Ma- 
nado in May, on the Togian Islands in August”. Fischer (16^'^) says: “this Bee- 
eater does not occur as a stationary bird, but appears in Ternate in the months 
March, April and May”. Nevertheless Count Salvador! (12) was able to re- 
cord two specimens killed in Halmahera in December, though the remaining 62 
of his series from New Guinea and the Moluccas are dated from March to 
September. It would thus appear that a few remain behind in the islands 
throughout the year; whether they breed there or not is not known. 
Notwithstanding that this species is such a thorough migrant it appears 
very possible that it is undergoing differentiation into an eastern and a western 
race. Mr. Dresser has the following remarks on its variations: “Specimens 
which I have examined vary somewhat in intensity of coloration and especially 
as regards the amount of blue in the plumage. A specimen in the Tweed- 
dale Collection from Port Albany has the crown, scapulars, and back slightly 
tinged with turquoise-blue; and one in the British Museum, a female from 
Dorey, has the abdomen washed with blue and is labelled Merops caerulescens . 
[See our remarks, antea^ on the probable breeding plumage of the female.] 
Another example in the British Museum, a female from Lombock, is peculiar 
in having a tolerably broad band of blue below the black pectoral band . . . 
I can endorse Mr. Wallace’s remarks [5] that specimens from the Sula Islands 
agree with those from Ternate in having more brown on the head and less blue 
on the breast than the Timor and Lombock specimens”. Meyer [15) has since 
separated specimens from Sumba as a subspecies, M. ornatus sumhaensis. Two, 
apparently second-year birds, have more blue on the throat below the black 
cross-band than any of over 60 other specimens in the Dresden Museum from 
Celebes, Sangi, the Moluccas and New Guinea; in two others the amount 
of blue is inconsiderable. Hartert (33) enumerates specimens from Sumbawa, 
Sumba, and Lombock, some with and some without blue on the throat. This 
character consequently seems not to be of a positive nature. There can be little 
doubt that the Sumba birds correspond with those from Timor and Lombok, 
whose differences struck Mr. Wallace. It would be of much interest to know 
whether specimens from West Australia have more blue below the black collar 
than those of New South Wales, or other differences, for if M. ornatus sumhaensis 
32 * 
