Birds of Celebes; Artamidae. 
433 
Artamus leucogaster is a species which appears to have recently extended 
Its range, having issued most probably from Australia, where the Artamidae are 
most strongly represented and where A. leucogaster in South Australia and New 
^ outh AVales “would appear to be migratory, visiting these parts in summer for 
the purpose of breeding” (Gould h 3). In the East India Islands it seems to 
e stationary. In Celebes, as Meyer remarks, it is very common at all times 
an eveiy where; it is said by Fischer to occur throughout the year in Ternate ; 
ree s plentifully in Java, breeds also in the Andamans, where specimens 
e been collected in most months of the year — January till July, November 
thprr^^' ’ Y^ile the dates of specimens tend to prove that it is stationary and 
ar ^ breeding bird in other parts. Specimens from the Tenimber Islands 
e istinguishable by their dusky plumage, the broader zone of white across the 
^ump and upper tail-coverts, larger bill, and tail narrowly tipped with white. They 
appear to be the only form of A. leucogaster worthy of specific or subspecific 
IS inction, and have been named A. miisschenbroehi by Meyer. It would be of 
of Th"^ Timorlaut has furnished the first appreciable differentiation 
AustvnV of invaders from 
exam^l! T f"""' mlj^Aurea). Celebesian 
h>iesden larger than others; the largest of those in the 
other adnlf'""^'"'” localities, but 
the latter fldebes are again considerably smaller than some of 
oomm^^t Artamus consists of insects. Its striking habits of flight are 
Swall^^*^- many writers. AVallace remarks that they closely resemble 
whe habits and flight, and Gould compares the present species, 
Bpt lining near the ground, to the House Martin of our own country, 
it having some similarity to a Bird-of-prey’s , as 
its almost without making a stroke on outstretched wings, or alters 
sone raising or lowering one wing or the other. Yet this is 
or nf fl nothing of the rushing haste of the small true Falcons, 
and P ft- ' It® Pci-ch 
at th^ ^ insect-food on the wing, “sometimes maldng only a short flight. 
It f • extended one before returning to the same or another perch. 
^ lequent y descends to the ground to jaick up an insect, and I have at times 
en several seated together on the roads” (near Port Blair, Andaman Islands, 
rJavison c 3). 
By many authors, as Prof. Newton (Diet. B. 1893, 22) remarks, the Arto- 
'mi ae are considered to be the nearest neighbours of the Hirundinidae , making- 
some approach to them in their long wings and habit of catching insects in 
continuous flights. If it be granted from their possessing patches of powder- 
down (Nitzsch, Pterylogr. Engl. ed. 1867, 80) that they should form a sepa- 
ls e family Artamidae, its true alliance must still be guessed at”. 
Meyer * Wiglesworth, Birds of Celeljes (Jfov. Oth, 1897). 
55 
