Birds of Celebes: Asionidae. 
95 
There are now examples in the Dresden, Leyden, British, Tring, and apparently, 
Vienna Museums. It is nearly allied to the Little Owl of the Philippines, N. 
philippensis Bp., which differs from it in having the upper wing-coverts spotted, 
the tail shorter, etc. Another species, showing much similarity in tone of colo- 
ration, but having the under surface marked with cross-bars instead of uniform 
or with indistinct stripes, is the Ninox hantu of Burn, figured by Dr. Sharpe 
on the same plate with N. ochracea. 
\ 
31. NINOX SCUTULATA (RaffL). 
Brown Hawk-owl. 
This species apparently consists of four races. We have not been able to 
investigate those of the Indian countries, but accept the views thereon of Schlegel 
and Dr. Sharpe, with the corrections of Mr. Hume and more recently of Mr. 
Blanford (Bull. B. O. C. 1894 XVIII, p. XLII, Ibis 1894, 526; Paun. Br. 
Ind. B. Ill 1895, 309). The races may be diagnosed as follows: 
1. The typical Ninox scutulata. 
Diagnosis. Size medium, wing 197 — 210 mm; 4^** primary scarcely longer than the 3’’^ and 
5*’\ which are about equal in length (Schlegel a I). 
Distribution. India and Ceylon, south-eastwards as far as Java and Borneo. 
2. Ninox scutulata lugubris (Tickell). 
Diagnosis. “Rather paler, with a greyer head” (Blanford). 
Distribution. “Found in India and Burmah generally, chiefly in the less damp parts of the 
country” (Blanford). The typical form, on the contrary, is believed by Mr. Blan- 
ford to belong to those parts where the rainfall is heavier — Malabar, Ceylon, parts 
of Burmah, etc. (Faun. Br. Ind. B. HI 1895, 311). 
3. Ninox lugubris affinis (Tytler). 
Diagnosis. A small race: wing 168—193 mm (Blanford). 
Distribution. Andaman Islands. 
4. Ninox scutulata japonica (T. & Schl.). 
Diagnosis. Size large, wing c. 218—235 mm; 3’’^ and 4^1* primaries about equally long, 
or the thud a trifle the longest, the 5*^ (according to Schlegel and W. Blasius) 
much shorter and nearly equal to the 2“‘‘. (In our sjpecimens with three exposed 
bands on the tail this seems to hold good for the 5^^ primary, hut in three of our 
four with more rufous tails and four exposed hands, the 5*^ is only about 10 mm 
shorter than the 3’''^, instead of about 20 mm shorter, like the 2"'^; in the fourth 
example the fifth quill is missing. Prof. W. Blasius measurements (i 3) show 
however, that the wing-formula is subject to great variation, the 5^’* quill being in one 
specimen (a female) only 85 mm shorter than the longest, in another specimen 21 mm 
