80 
Birds of Celebes; Falconidae. 
abdomoii marked with drop- and arrow-liead spots of black; under wing-coverts 
white or ochraceous buff spotted like the under surface. 
Adult male. Like the female, but both above and below less plentifully spotted and banded 
with black; tail clear smoke-grey, crossed only by the broad subterminal black band. 
“Iris brown; bill greyish blue; cere and skin round the eyes yellow; feet deep 
citron-yellow, claws black” (M. 2). 
Young. Appears to differ little from the female. 
Measurements. Wing 210—240 mm; tail 160; tarsus 43—44; culmen from cere 15—16. 
Attention was called as long ago as 1866 by Schlegel to local deviations 
of coloration, when it was pointed out that the supposed specimen from Borneo 
and two from Timor and Flores respectively agreed with those from Celebes, 
while the more strongly coloured race of Halmahera was also represented in 
the Leyden Museum in specimens from the neighbouring islands of Morotai, 
Ternate, Mareh, Tidore and Batchian. 
In a good series intermediate specimens may occur in localities indicated 
for the pronounced races. For instance, we have one from the Minahassa, N. 
Celebes, wMch by its dark tawny brown tints approaches three from Halmahera, 
but the under wing-coverts are lighter. 
A male from Timorlaut (Nr. 6693) appears to have a somewhat greyer 
head and the spots on the back smaller than in any others, but, as the speci- 
men is a bad one, it is difficult to form a good judgment of its characters. This 
may be a hidden subspecies. Schlegel (c II) mentions that the Flores specimen 
also has smaller and fewer spots on the U23perside. 
In Borneo the dark Japanese form of the Common Kestrel ( Tinmmcidus 
alaudarius) occurs in winter (Everett, List B. Borneo, in J. Str. Branch E, A. 
S. 1889, 186), and the same species — probably also the same local form of 
it — has been obtained near Luzon at sea (Finsch & Conrad, Verb. Z. B. 
Ges. Wien, 1873 June 4). The statement, therefore, of Horsfield (Tr. %. S. 
1821, XIII, 135) that F. tinnunculus belongs also to Java is likely enough to 
be true, although his note has always been taken as referring to T. mohiccensis. 
The food of the Moluccan, like that of the Common Kestrel is pretty 
varied. In Celebes Salomon Muller noted that it fed on “mice, small lizards 
and birds, grasshoppers etc.”; v. Eosenberg sjoeaks of it as “ein HaujDtinsekten- 
jager”; Meyer observed that it j)reyed on little birds. It is 230ssible for a bird 
with so varied a diet as this to remain stationary in a tro 2 )ical island. As the 
numerous dates recorded by Schlegel and others show, the S 2 )ecies occurs all 
the year round in Celebes, where it is extremely 2^1entiful, and from a note 
of Dr. Fischer, recorded by Mr. Pleske, it is 2 ^i^esent also throughout the 
year in Ternate. Consequently, it is not suiqo rising that in these two neigh- 
bouring localities distinguishable subs 2 recies have arisen. 
