1872.] ON THE BIEDS OF CELEBES. 159 



An examination of a considerable series of this group has led me to conclusions widely differing 

 from those contained in the Catalogue of the Leyden Museum. 



The difficulties which meet a student of the genus Centrococcyx arise from the general 

 resemblance in the plumage of its members, the blue, the green, or the purple hue of the black 

 portion, and the deeper or less intense shades of the rufous not being sufficiently striking and 

 well marked, except in perfect plumage, to be relied on as distinguishing characters. We also 

 find in the Coucals, as in other natural groujjs the members of which are numerous, the colouring 

 of the adult in one species representing, more or less, the transition colouring of the young of 

 another species. Thus the dingy greenish-brown hue of the rectrices in an immature C. rectunguis 

 changes to glossy dark green in the next stage, and is again converted into deep blue in the adult 

 bird. But in the common Indian Crow-Pheasant the colour of the rectrices is arrested at the 

 green stage, and green remains the hue of that part of the plumage in the fully adult bird. A 

 complete series of fully adult examples from all parts is consequently essential before characters Tr.Z. S.viii. 

 founded on the colouring can be relied on. Another source of difficulty is the extreme variability P" '• 

 of the plumage in the first and second years, the young wearing a livery greatly resembling in 

 general characters the adult garb of Australian C. ])liasianiis and its allies. It is likely that this 

 variability is more apparent than real, and that each species, as in Eudynamis for instance, has 

 special phases of immature plumage peculiar to itself. Happily certain characters are always 

 present whereby typical specific groups can be separated from one another. The most important 

 are the dimensions and the form of the bill. Three distinct groups of Asiatic Coucals can thus 

 be separated ; and as Java is the only known locality where all three are represented, we may 

 use the Javan species as standards : — 



First, C. huhutus, Horsf , as the type of the large species, such as : — C. sinensis, Stephens, ex 

 China; C. nifipennis, Illiger, India and Ceylon; C. eurycercus, A. Hay, Malacca; C. lorneensis, 

 Bp., Borneo ; and the very distinct C. cMororhynchus, Blyth, ex Ceylon. This group appears to 

 be unrepresented in Celebes ; nor has it been discovered in the Philippines or in Formosa. 



The second group comprises species smaller in size, with short, but proportionally very high 

 bills, the diminished length of the bill making it appear disproportionately broad at the base. 

 The Javan representative is C. affinis, Horsf. An identical form inhabits Flores ; and a some- 

 what larger species is found in Ceram, which, if the same as the Ambo}Tia Centrococcyx, must 

 stand as C. medius, Bp., ex Miiller. A Macassar individual, collected by Mr. Wallace, although 

 in young plumage, bill pale flesh-colour, and plumage streaked and mottled, so closely resembles 

 Javan C. affinis in dimensions and form of the bill, that I cannot separate it. 



The third group consists of the smallest species, represented in Java by C. javanensis, Dumont, 

 = C. lejndus, Horsf. In these the bill is a miniature resemblance of that of the second group. 

 The upper tail-coverts are highly developed, or, in other words, they are the tail-coverts of the 

 C. affinis group. C. hengalensis, Gm., of India ; C. viridis, Scop., of the Philippines ; C. moluc- 

 censis. Cab., ex Bernst., of Ternate, very near to C. javanensis, but with a proportionately longer 

 tail; and C. dimidiatus, Blyth, =C lignator, Swinh., of Formosa and China, come within these 

 limits. Malaccan and Banjermassing individuals scarcely differ from the Javan species ; and, 

 judging from the measurements given by Professor Schlegel, representatives occur in Ceram, 

 Sumbawa, Bangka, Sumatra, and Amboyna. 



t2 



