46 



P.C. RASMUSSEN AND N.J. COLLAR 



intermedia between one or both members of the himalayana/finschii 

 and roseata/cyanocephala species pairs in every character set exam- 

 ined. Because roseata and cyanocephala are similar in size and 

 proportions, roseata is not mensurally ruled out as a parental spe- 

 cies, but it is ruled out on plumage (see above). However, finschii is 

 smaller than himalayana, particularly in wing characters, and its tail 

 proportions are different from any other species and AMNH 

 intermedia, so it is unlikely to have been parent to the latter. We 

 know of no cases in which a good species, which is intermediate 

 between two patently different congeners in numerous phenotypic 

 characters, totally lacks distinctive features of its own. Also, it is 

 scarcely conceivable that a wild species would duplicate exactly the 

 character states found in known hybrids between two quite distinct 

 taxa Thus both plumage and mensural analyses very strongly 

 support the hypothesis that AMNH intermedia are of hybrid origin, 

 and this is further validated by their identity with known hybrids 

 between himalayana and cyanocephala. 



Arguments previously used in favour of specific status 

 Hartert (1924) stated that if intermedia were a hybrid, 'so many 

 specimens would not very likely have come at the same time, 2 and 

 one would expect them to vary, but they are all alike', with foot- 

 note 2 disclosing that 'Our six males were selected by Mr. 

 Dunstall, a dealer in feathers, from a greater number of these 

 birds, he told us'. These statements have often been repeated as 

 evidence of specific status (Biswas 1959, Walters 1985, Inskipp 

 and Inskipp 1995), but both are flawed. First, the holotype did not 

 originate with the other intermedia. Second, one of the six re- 

 maining specimens is a typical immature himalayana (see above), 

 leaving only five intermedia supposedly of similar origins and 

 identity. Hartert's assertion that these came from a greater number 

 (though not 'a much greater number', contra Walters 1985) in the 

 possession of and selected by Dunstall implies that Hartert him- 

 self did not see additional intermedia but had taken the London 

 plumassier's word for it. There is nothing to indicate that Dunstall 

 would have recognized the difference between intermedia and 

 cyanocephala, and (our third objection) there seems no compel- 

 ling evidence that he actually had more specimens of intermedia: 

 the 'greater number of these birds' may have referred to the rest of 

 a shipment of other Psittacula parakeets, a possibility supported 

 by one of the six 'intermedia' being a juvenile himalayana. There 

 was a considerable millinery trade in Psittacula skins around this 

 time (Hartley 1907). and only a very small percentage would have 

 ended up in reference collections. 



Fourth, preparation styles and materials used in the five Dunstall 

 intermedia plus the immature himalayana differ strikingly among 

 the skins. These differences strongly suggest that, although all are 

 native skins, they were not all prepared at the same time and place. 

 They may have come to Rothschild's museum at the same time, but 

 not necessarily so to Dunstall or his supplier. Additional support for 

 staggered acquisition of the material lies in the fact that native skins 

 of cyanocephala and himalayana strongly resemble, in style and 

 materials, not only those of intermedia but also those of finschii, 

 which (given their distribution) must have originated farther east. 

 Conversely, one adult intermedia (AMNH 621541) and the imma- 

 ture himalayana are so similar in preparation materials as to make it 

 highly probable that they were prepared together. Many of the 

 'Bombay preparation' parakeet skins very likely came from bird 

 markets to which captive birds had been brought from afar. Indeed, 

 the incidence of at least three partial lutino specimens of this same 

 preparation suggests selective breeding for this trait, which has long 

 been highly desired by Indian aviculturists (Greene 1 884). 



After restating Hartert's (1924) contentions that intermedia is a 



valid species, Biswas (1959) indicated he had concluded the same 

 independently. However, his only further evidence was as follows: 

 'Besides, if they were man-made hybrids, they would necessarily 

 have been cage birds. But the character of their toes does not indicate 

 this. Psittacula intermedia may, therefore, be regarded as a genuine 

 wild species'. Besides the obvious fact that bird hybrids are not 

 necessarily 'man-made', Biswas (1959) gave no indication of which 

 features of the toes were found inconsistent with captive origin. 



Our examination showed that the intermedia specimens exhibit to 

 varying degrees several conditions consistent with their having been 

 in captivity under suboptimal conditions (Harrison and Harrison 

 1986). It thus seems likely that most or all of the intermedia 

 specimens in existence were captives for some period immediately 

 prior to their death. In India, Psittacula parakeets have long been 

 extremely popular cagebirds (Finn 1906, Ali 1927, Dharma- 

 kumarsinjhi 1954, Sinha 1959), and although most are taken from 

 nests (Hume 1890), others are captive-bred commercially (H. S. A. 

 Yahya, pers. comm. 1997), and they have long been bred for the 

 Indian aristocracy (Greene 1 884). Mutations in particular are bred in 

 captivity in India (S. R. Sane, pers. comm. 1997), garnering up to Rs. 

 20,000 (Ahmed 1 997). From the prices of cagebirds considered to be 

 intermedia (Rs. 2,000 vs. less than Rs. 25 for ordinary cyanocephala: 

 A. Rahmani in Inskipp and Inskipp 1995), it is self-evident that 

 captive-breeding of such hybrids would be well worth the trouble. 



Walters (1985) drew attention to previously overlooked descrip- 

 tions of captive-reared cyanocephala x himalayana (Tavistock 

 1932-1938) which do not match Forshaw's (1973) description of 

 intermedia. Based on this, as well as Hartert's (1924) arguments, 

 Walters (1985) concluded that intermedia could not be a hybrid 

 between those species and must therefore be a valid species, and in 

 this he has been followed by most recent authors. Presumably also 

 on the basis of this captive-breeding event, Arndt (1996) stated 'it 

 has been discovered that hybrids between [himalayana and 

 cyanocephala] differ considerably from Psittacula intermedia. . . .' 

 However, a reevaluation of the aviculturist Tavistock's writings 

 shows some relevant discrepancies. 



First, although Tavistock (1932-1938) did repeatedly pair a male 

 himalayana with a female cyanocephala, rearing at least seven 

 young over a period of five years, he published only a very brief 

 description of just two of those young, which were nestmates 

 (Tavistock 1933), and did not describe their adult plumage. Thus 

 there is no assurance that the other hybrids resembled these two, nor 

 is there information enabling comparison between his hybrids and 

 AMNH intermedia. 



Second, while it is true that Tavistock's (1933) description of the 

 two young cyanocephala x himalayana as having white tail tips 

 does not match the specimens of intermedia (the discrepancy noted 

 by Walters), other statements Tavistock made throw doubt upon his 

 entire account. His remark that 'they resemble young Plumheads, 

 but their central tail feathers are brighter blue with white tips and 

 their heads have adusky tinge' is nonsensical, as youngcyanocephala 

 do have white tips to their tails, and their heads may have a dusky 

 tinge much like juvenile himalayana. Also, it is counterintuitive that 

 hybrids would have brighter blue central rectrices than those of 

 young cyanocephala, since juvenile cyanocephala have these feath- 

 ers considerably bluer than do juvenile himalayana, which are 

 green-tailed. Sedgemore stated that his juvenile cyanocephala x 

 himalayana had 'blue green' upper tail surfaces (Sedgemore 1995), 

 and this is confirmed by photos of them as juveniles. The central 

 rectrices of the adult plumage of both parental species are bluer than 

 in the juvenile plumage, while those of adult cyanocephala are bluer 

 and less purple than for adult himalayana. Since the two young 

 hybrids described by Tavistock (1933) had only hatched that year, 



