34 



P.C. RASMUSSEN AND N.J. COLLAR 



Table 2 Qualitative characters used in hybrid diagnosis between Psittacula cyanocephala, P. krameri, and S. R. Sane 's specimens and living bird. 



Character 



P. cyanocephala 





Sane's birds 



P. krameri 



Maxilla 



all yellowish 





all red 



all red 



Mandible 



black 





black 



black 



Cere shape 



moderately wide, rounded 



narrow, nearly straight 



narrow, straight 



Cere colour 



medium grey 





pale greyish-horn 



whitish 



Orbital skin 



dark grey 





fleshy whitish 



orange 





rather wide, not prominent 



rather narrow, very prominent 



rather narrow, very prominent 



Lores pattern 



no line 





slight line 



strong line 



Forehead 



reddish-purple 





greenish tinge 



green 



Auriculars and central face 



reddish-purple 





dull purplish-blue 



bluish-green 



Lower face 



purplish 





cerulean blue 



lime green 



Rear crown 



shining mauve 





cerulean blue 



powder blue 



Lower border neck collar 



as nape 





broken orange-chestnut 



nearly complete rose-orange 



Nape 



bright bluish-green 





viridian 



slightly bluish-green 



Upper wing coverts 



variably bluish-green 





slightly bluish 



lacking bluish 



Shoulder patch 



maroon 





slight tinge on one 



absent 



Underwing coverts 



pale turquoise-blue 





slightly bluish 



yellowish-green 



Rump 



variably bluish-green 





slightly bluish 



yellowish-green 



Uppertail 



rich dark blue 





pale blue-green 



lime-green 



Tail tip 



white, moderate width, 



spatulate 



concolorous with rest or 



narrow whitish; slightly spatulate 



concolorous with rest, not spatulate 



Foot 



dark pinkish-grey 





pale pinkish-grey 



whitish 



tip of next shortest feather; lengths of longest (Rl, central) and next 

 more lateral (R2) rectrices (both taken from insertion of central 

 rectrices); maximum width of yellow or white tip and approximate 

 maximum width at distal end of blue or green area of Rl (with 

 feathers flattened out); distances between tips of each rectrix (except 

 Rl) of one side and the next shortest (next more lateral) rectrix; 

 widths of each rectrix of one side at the tip of the next shortest one; 

 approximate distance from tips of Rl and R2 to definite blue or 

 green part of feather (= length of pale tip); tarsus length; minimum 

 width of tarsus; length of claw of middle toe (from distal edge of 

 scute); length of hindclaw (from distal edge of scute). Feathers in 

 sheath or in a damaged or heavily worn state were not measured, and 

 if there was a difference in length between rectrices of a pair, the 

 longer one was measured. Maximum skull width was measured over 

 skin and compressed feathers for specimens in which palpation and/ 

 or x-rays indicated that the widest portion of the skull was intact and 

 not padded with stuffing. 



Two specimens from the Rothschild series (BMNH 1980.3.1 and 

 AMNH 621545) showed very different plumage and mensural 

 characters from each other and from the remainder of the specimens 

 in this series, and so were treated as unknowns in the analyses. 

 AMNH 62 1 545, although thought a femaleintermedia by Rothschild 

 ( 1 907), was considered by Biswas ( 1 959, 1 990) and Forshaw (1973) 

 to be an immature himalayana, the latter opinion being shared by us 

 after examination. We therefore compared its plumage characters 

 with known immature himalayana and finschii, and its measure- 

 ments with nine juveniles (sexes combined) of the former. 



Univariate statistics and principal components analyses (PCAs) 

 using correlation matrices were run on external and skeletal meas- 

 urements using SYSTAT for Windows (Version 5.0) on an 

 IBM-compatible PC. Variables for PCAs were chosen to allow the 

 inclusion of selected individual intermedia specimens without esti- 

 mation of missing data, which would be inadvisable owing to the 

 small sample size of intermedia. 



Evaluation of origin of specimens 



To test the idea that the Rothschild Collection series of intermedia 

 had the same origin - an argument first put forward long ago by 

 Hartert (1924) - we compared preparation styles and materials used 

 among these specimens by external examination and study of x-rays. 



We also compared them with native-prepared (e.g., 'Bombay prepa- 

 ration', 'India', and 'Madras') skins of other Psittacula species 

 {cyanocephala, himalayana, roseata, finschii) at AMNH, MCZ. and 

 USNM. To permit analysis of certain aspects of preparation styles 

 and materials used, radiographs (x-rays) were taken of the Rothschild 

 intermedia specimens, and of native skin specimens of himalayana, 

 finschii, and cyanocephala for comparison. X-rays (ventral and 

 lateral views) were made of the AMNH and BMNH intermedia 

 (including the putative immature himalayana) by M. N. Feinberg, 

 Department of Ichthyology, AMNH (30 kV and 3 tnA for 2 min, 

 using Kodak Industrex-M Ready-pack film), and for the other 

 specimens by PCR in the Fish Division, National Museum of 

 Natural History (USNM; 25 kV and 5 mA for 30 sec, using Kodak 

 Industrex SR film). 



To evaluate whether the Rothschild Collection series originated 

 from wild, not captive birds - an argument used by Biswas (1959) to 

 support species status - we examined the Rothschild specimens for 

 presence of: overgrown bill and claws; broken remiges and rectrices; 

 overly worn feathers due to delayed moult; abrasion damage to 

 feathers of the type resulting from repeated contact with cage bars; 

 and dirt on plumage, bill, and feet consistent with a confined 

 environment. 



Examination of captive birds 



We examined five living adult hybrids belonging to Mr M. Sedgemore 

 that are the progeny of an experimental pairing of a male 

 cyanocephala and a female himalayana. The female parent, which 

 died in the nest shortly after producing the second of two hybrid 

 broods in successive years, was considered unsalvageable as a 

 specimen; the male parent died more recently and the skin is 

 preserved as BMNH 1998.33.2. We took hand-held photographs and 

 aviary videotape of all five hybrids, as well as several measurements 

 (taken by PCR while the birds were held by Sedgemore) of bill, 

 wing, and tail. All the hybrids were in some stage of moult, so certain 

 measurements could not be taken. The recently moulted central 

 rectrices of the single female hybrid are now at the BMNH, and 

 Sedgemore also gave us several photographs of the hybrids, both as 

 juveniles with their parents and as adults. PCR also examined and 

 videotaped the single live bird claimed to beintermedia remaining in 

 Sane's collection in December 1997. We also studied photos sent by 



