614 SYSTEMATIC SYNOPSIS. — PSITTACL 



ambiens, no furculum ; (7) Chrysotinrc, no ainbiens, a furculum, no oil-gland. There are thus 

 7 subfamilies of 2 families of Psittaci, as based solely upon the particular set of anatomical 

 characters utilized by Garrod for his purposes. But this surprising result is " far from being 

 satisfactory ; " it violates some obvious evidences of affinity in other respects, and in particular 

 makes the geographical distribution of the order unintelligible ; it has therefore been rejected 

 by common consent of the later monographers, as Finsch, Reichenow^, and Salvadori. 

 Reichenow's arrangement (1881) presents 9 families, as follows : (1) Pioxid.e, American (ex- 

 cept the African genus Paeocephalus) , short-tailed ; (2) Conurid^, all American, long-tailed ; 

 (3) PsiTTACiD.E, restricted to the African genera Psittacus and Coracopsis, or the Jak()S and 

 Vasas ; (4) Pal^ornithid.e, Old World ; (5) Trichoglossio.e (or LoriidcB), the Lories and 

 Brush-tongued Paroquets, of Australia, Polynesia,- etc. ; (6) Micropsittid^, a small family 

 of diminutive Paroquets of the 3 genera Psittacella, Cydojisittacus, and Nasiterna, Austro- 

 Malayan and Papuan ; (7) Platycercid^, Old World ; (8) Plictolophid^ (or Cacatuidce),. 

 Austro-Malayan and Papuan, with the New Zealand genus Nestor forced into this connection ; 

 (9) Stringopid.e, with the New Zealand genus Strwgops and the Australian Geopsittacus and. 

 Pezoporiis. In the British Museum Catalogue of 1891, Salvadori arranges the Psittaci as fol- 

 lows : (1) Nestorid^e, one genus. New Zealand; (2) Loriid^, 14 genera, Austro-Malayan 

 and Polynesian ; (3) Cyclopsittacid^, 2 genera, Cyclopsittacus and Neopsittacus, both 

 Papuan ; the group thus not coincident with Micropsittid(B as just given ; (4) Cacatuidce : 

 (a) Cacatuince, 5 genera, Austro-Malayan and Papuan, and {h) Calopsittacince , one Australian 

 genus; (5) Psittacid^e, with 6 subfamilies: (a) Nasiterninae, one genus, Papuan; {b) Co- 

 nurince, 15 genera, all American; (c) Pionince, 10 genera, all American except the African 

 Poeocephalus ; (d) PsittacincB, 2 African genera, Psittacus and Coracojisis, with the Papuan 

 genus Dasyptilus ; (e) PalccornithincB, 15 genera, widely distributed in the Old World; 

 (f) Platycercina, 11 genera, Australian and Polynesian; and (6) Stringopid^j, one New 

 Zealand genus. This arrangement is an improvement upon all earlier ones, although my dis- 

 tinguished friend confesses that it " does not bring us nearer to an understanding of the mutual 

 or phylogenetic relations of the f;unilies.'' 



It seems to me probable that certain genera of Psittaci will prove refractory to any scheme 

 which may be devised. Thus, the African Poeocephalus and the Papuan Dasyptilus are inex- 

 plicable in geographical distribution, if they be really members of the respective subfamilies to 

 which they are assigned by Professor Salvadori. A number of genera of chietly pygmy Paroquets 

 are particularly troublesome ; such are Psittacella, Psittinus, Psittacula, Agapornis, Cyclo- 

 psittacus, Bolbopsittacus, Nasiterna, Nymphieus, and Nanodes. Nevertheless, if we duly cor- 

 relate anatomical with external characters, and both of these with certain faunal considerations, 

 we may arrive at a classification of the Parrots which probably requires a minimum of excep- 

 tions to be taken or anomalies to be further provided for. Without prejudice to a few questions 

 of fact which remain open, from lack of information on some structural points as correlated with 

 geographical distribution, one way out of our present difficulties may be found by recognizing a 

 greater number of families and subfamilies, primarily divided into two subordinal or superfamily 

 series. Thus, if we make the genus Stringops a family Stringopid^ of a suborder Strin- 

 gopes, characterized as ecarinate and efurculate, with an ambiens, complete bony orbits, two 

 f'arotids, tufted oil-gland, soft plumage with a facial disc as in Owls, wings unfit for flight, etc., 

 the remaining Parrots may form the 



Suborder EUPSITTACI, Carinate Parrots, 



with the furculum normally complete, exceptionally defective. With a single possible excep- 

 tion, that of the genus Poeocephalus, the Eupsittaci may conveniently and not unnaturally be 

 considered in two series, Pal^eogtean and Neogaean. 1. Pal^og^i. 1. The Platycer- 



