( 19« ) 



not very scientific) or separated iuto a number of subspecies, each with a different 

 geographical distribntion (a very difficult and tedions treatment, but surely more 

 scientific, though at present not approaching full exactness and finality). What 

 I can least understand is why " Halcyon davisoni" the Andaman race, should be 

 a subspecies of " kumii " and not of chloris, and why these two are separated from 

 chloris by sordidus, the more than doubtful fomfeni, solomoiiis, and surensis. 

 Surely all these, as far as separable at all (also sordidm) are subspecies of chloris. 

 There are frequent intermediate colorations between "//iim/i" and chloris in the 

 same countries ! 



At present I can recognise the following facta : — 



1. Specimens from the Moluccan Islands : Burn, Batjan, Tukan-Besi Islands, 

 Key Islands, South-East Islands, Banda, Tenimber (Timorlaat) Islands, Dammer, 

 South-West Islands (Koma, Wetter, Letti, Kisser), Alor, the Timor grnu]) to 

 Lombok, are generally (though not always) larger, have the ear-coverts more or 

 less distinctly blackish, and connected with a wide blackish nuchal collar, the 

 npperside somewhat duller and often rather darker greenish, the wings more blue, 

 and thus more in contrast with the back, the crown very frequently tinged with 

 brownish. As a rule the females are more brownish on head and back and have 

 the ear-coverts purer black ; but there are exceptions, unless some specimens are 

 wrongly sexed by the collector. This form is the Ilalojon chloris chloris (Bodd.). 



2. Specimens from the Greater Sunda Islands (Java, Borneo, Sumatra), the 

 Malay Peninsula, and the Philijipines are generally (though not always) smaller, 

 have the ear-coverts greenish like the crown, or a little darker, but hardly ever 

 blackish, are above more uniform and somewhat lighter greenish, the wings less 

 in contrast with the back ; only a narrow blackish nuchal collar, or sometimes none 

 at all. The females seem to be also duller, more brownish above, with generally 

 darker ear-coverts. This form must be called Halcijoti chloris collaris (Scop.).* 



3. Some specimens from the Pelew Islands, collected by Kubary, are, as far 

 as I can see, perfectly similar to Philippine //. chloris collaris (Scop.). 



4. Specimens from the Abyssinian coast (Arafali, Dankali) are very close to 

 //. chloris collaris, but are a shade more uniform above and a shade duller ; the 

 tips of the wings, especially the tips of the inner webs of the primaries, are less 

 blackish, but more pale greenish, the outside of the wings not bluish. There is no 

 black nuchal collar, the ear-coverts are green ; but while the differences in the males 

 are very slight, the/emales differ more, being strikingly different from the males ; 

 the upper surface dull olive, the tail of a very pale, faint green. The name of this 

 form is of course Halcyon chloris abi/ssinica (Licht. ). 



5. In one jdace only that we know oi—i.e., in the Talaut Islands, north 

 of Celebes — occur two forms, rather large ones and small ones, both alike in 

 colour. AVhat is the meaning of this ? The difference is very great, and there 

 are evidently no intermediate examples. Messrs. Meyer & Wiglesworth {Birds 

 of Celebes I, pp. 293—295) came to the conclusion that the small sjjecimens from 

 Talaut were "young and not full-grown." This conclusion is erroneous. The 

 young H. chloris is not " similar to the adult in cciloration, recognisable as young 



• Alci-ilo collaris Scopoli, Bel. Flor. et Faun. Inmhr. ii. p. 90,1788, ex Sonnerat, Mart. pech. i 

 collier blanc des Philippines— tvp. hab. Philippinos ! No doubt all the hinls from the Philippines and 

 Randa Islands as far east of Java belong to this form, and. I think, also those fiom Celebes, though the 

 Celebes birds are oscillating, sometimes more like typical chluriH. Those from the Malay Peninsula m.iy 

 have to be separated, as among them occur tlie curious blue so-called Inimii. Andaman birds appear to be 

 quite similar, and also tliose from the Sunderbunds. 



