Materials and methods.—

The male specimen we chose for this treatment was collected by Frank M. Chapman on 24 March 1890 in a cypress swamp along the banks of the Suwanee River, Florida, 20 miles north of the Gulf of Mexico (Brewster and Chapman 1891; see also Chapman 1933, Tanner 1942, Hahn 1963, Austin 1967). Chapman (1933) later stated that he never encountered this species again. To the consternation of his field companion, William Brewster (Austin 1967), Chapman removed the skeleton (AMNH 4708) from the specimen and prepared only a folded pelt (AMNH 49569), not a traditional study skin. It is the only skeleton of this taxon in the AMNH collection. Jackson (2004) explained why Chapman might have prepared the specimen this way. According to Jackson, when the photo of Brewster holding the freshly collected specimen (Fig. 1) is enlarged, it appears as if part of the mandible has been damaged by shotgun blast. We examined the skull of this specimen carefully and found that both mandibles are intact and undamaged. Interestingly, Shufeldt (1890), noticing the absence of Ivory-billed Woodpecker skeletons in collections, had suggested to Chapman that he might prepare a skeleton should Chapman obtain a specimen during his Florida expedition.

We photographed the skeleton and pelt for documentation (Fig. 2), then removed the original label from the pelt and placed it in a Mylar envelope for storage. To relax and soften the integument, a humidity chamber with an airtight lid was prepared from a plastic storage container measuring 51 × 35 × 40 cm. The chamber was filled with 2.5 cm of clean, dry play sand. To prevent growth of microorganisms, a 1% solution of phenol (carbolic acid) was used to dampen the sand. We placed a three-sided, plastic-coated wire rack 8 cm above the sand. The entire pelt was arranged dorsal- side-up on the rack, and the lid was secured. The relax- ation process was monitored twice daily. The pelt was checked for mold, gently hand-stretched as it softened, and flipped over after each inspection. A daily log was kept of observations and progress. Within 24 h, integument and feathers were moist, and a slight softening of the integument was noticed. In addition to monitoring for mold, we also checked that feathers had not slipped because of overexposure to humidity. After seven days, when the integument reached a point when it was malleable and feathers were still securely embedded, we judged that the specimen was sufficiently relaxed and we removed it from the humidity chamber for the next phase of preparation.

The left wing was severed from the pelt at the shoulder, then extended and pinned on Styrofoam board, with primary slots opened (see Winker 2000). The pelt was then spread and pinned on Styrofoam board. Lastly, the wing and pelt were dried under a small fan at low speed for 10 days. We note that at some point during previous handling, before our preparation, the tail had become detached from the body at its base but had remained associated with the specimen. Thus, before pinning the pelt, we stitched the tail to it with waxed thread. The original label was encapsulated in Mylar and fastened onto the head of the pelt. A new label, bearing the identical number (AMNH 49569) and data as the pelt, was attached to the extended wing. A cross-reference to the skeleton was noted on both labels, and the original entry regarding this specimen in the museum's catalogues was updated. Two Mylar envelopes measuring 41 × 22.5 cm and 31 × 23.5 cm were constructed to store the wing and pelt separately (Fig. 3).

Description of specimen.—

Unlike most modern spread-wing specimen preparations, Chapman's removal of wing bones created an unstable region between the primaries and secondaries along the wing's leading edge. The normal complement of 10 primaries, including the foreshortened (Short 1982) primary (P10), and 10 secondaries, however, remained intact. The wing measures 260 mm in length from point of origin to longest primary (P6) and 170 mm in width from the alular feathers to the tip of the first secondary (S1). We caution that mensural characters in such an unstable wing preparation may be subject to distortion as a result of original and subsequent manipulations. The wing is somewhat rounded. Dorsally, P5 to P10 are uniformly dark in color, whereas P1 to P4 have varying degrees of white distally (Figs. 3 and 4). All 10 secondaries are dark to the midpoint, from which they become white distally (Figs. 3 and 4). Two white tertials are present. All white portions of the feathers show slight discoloration.

We believe that this specimen is an adult on the basis of its acutely pointed mandibles (Jackson 2004). The fresh, velvety, bluish-black lesser secondary coverts, in contrast to brownish, faded, worn primaries, secondaries, and remaining coverts, suggest partial molt. Faint horizontal striations, or growth bars (Michener and Michener 1938, Grubb 1989), are sparsely distributed among multiple feathers over several feather tracts. The alular feathers are fresh but otherwise unremarkable. Ventrally, P5 to P10 are dark overall, whereas P1 to P4 are dark proximally, with varying degrees of white distally (Fig. 4). All 10 secondaries are dark to the midpoint, from which they become white distally; underwing coverts are entirely white (Fig. 4).

Conclusion.—

The preparation of a skeleton, pelt, and extended-wing specimen gleaned from an Ivory-billed Woodpecker collected 117 years ago increases its already high scientific value. The availability of such an extended-wing specimen, showing the characteristic Ivory-billed Woodpecker wing pattern, also allows for comparative study among members of Picidae, such as the sympatric Pileated Woodpecker (Dryocopus pileatus; Fig. 4).