218 
Birds of Celebes: Ouculidae. 
on the subject of this troublesome species; to wit, Davison’s Tenasserim 
specimens are included under C.javanicus, but Hume’s remarks relating thereto 
are jjlanted in the synonymy of C. hengalensis; in the sketch of the distribution 
of C. bengalemis the author draws a line at the Burmese countries, though 
including in the synonymy records by Hume and Davison, Bingham, and 
Kelham of its plentiful occurrence in Tenasserim, the Malay Peninsula and 
Singapore; Cabanis and Heine’s two specimens of C. lepidus from Java are, 
moreover, included in the synomymy of both the species recognised by Shelley. 
Authors, on the whole, are generally agreed in uniting specimens from neighbour- 
ing localities Avith which they are acquainted ; S c hi e g e 1 alone has grouped them 
together as one species, a decision which Bfittikofer (n 2) rightly declines to 
undo. Ihus Hume (e5) says: “It may be that the JaA'an bird is distinct, but 
certainly all those that Ave have seen from the Malayan Peninsula have been 
identical Avith those from A^arious parts of India and Burma”. Shelley crosses 
the bridge from the Malay Peninsula to Java, as S cl a ter had done t hir ty 
years earlier (c2), and the soundness of Avhich had been again indicated by 
A. Muller in 1882 
In another direction Capt. Shelley agrees with Mr. Oates (111) and the 
Abbe David & Dr. Oustalet (e 2) as to the identity of the birds from South 
China, with others of India; though he terms the race C. javanicus, and the 
other authors C. hengalensis. In the East Indies birds of the different Great 
Sunda Islands are very rightly, as it appears to us, united by Shelley as one 
species Avith those of the Moluccas, the identity of Avhich Avith C. javanensis had 
been already indicated by Salvador! (q 2) and Meyer (k 5). 
With a large series of 22 specimens from Celebes, and further specimens 
from iagulandang, Great Sangi, Siao , Talaut, Ceram, Halmahera, Ternate, 
4 labelled “Moluccas”, 4 or 5 Java, and 1 Sumatra before us, Ave are able to 
lend support to this aIoav. 
It is certain that, as a species, C. bengaleiisis ranges from the Himalayas 
and South China throughout the intermediate countries to the Moluccas. 
The question next suggests itself: does this species tend to develop any ex- 
treme forms of coloration or size in special points of the area over AA^ich it is 
spread, or is it to be regarded as “one harmonious Avhole”? Before a neAv 
species can arise, isolation of some description is necessary; in other words, a 
group mu.st be prevented by some cause or liking from interbreeding with the 
rest of the species before its complete separation by colour or structure can 
take place. An intergrading local race or subspecies aauII, hoAvever, ahvays 
be likely to come to a head, if w’e may use the expression, at s^jecial centres 
in the range of a species of uninterrupted distribution, especially if it be 
stationary. During migration, many birds undoubtedly get lost and do not 
alAvays find their Avay back to the place of their birth; such, settling in other 
localities, may interbreed Avith the individuals of the species found there and act 
as a check on the differentiation of local races. 
