We Spent Six Months On The Photograph. The Wing Is 2.1 Metres Wide.
A crate opened in Cochise County in July holds a dried membrane on thirty-one articulated elements. Two laboratories put the collagen between 1874 and 1908, and neither can match it to any bird, mammal or living reptile.
A months-long reporting project. Documents cited below are held in The Vault and available to readers.
BISBEE, United States — On 14 July a crate came off a shelf in a storage barn outside Bisbee and was opened on a bench in front of a video camera and two witnesses who do not work for this newspaper.
Inside, wrapped in newsprint and cotton wadding, was a dried wing.
It measures 2.1 metres from the shoulder joint to the tip. It carries thirty-one articulated elements, four of them elongate phalanges of a kind no bird has. Between and behind them, brittle and largely intact, is a membrane.
What this newspaper printed in August, and what it left alone
Three weeks ago this desk published a long piece establishing that the famous photograph of a winged creature nailed to a barn wall does not exist: not in the 1890 newspaper everyone cites, and not in any of the twenty-three places 412 people told us they had seen it.
That story contained a paragraph setting aside a different question. Whether the 1890 account itself is true, we wrote, is separate, and gets conflated constantly, including by us.
We then left it alone for seven months while running a survey panel, because memory is cheap to study and Cochise County is four hours from an airport. Our standards editor has written the sentence about that at the foot of the coverage table and we have not softened it.
What the anatomist will say
Dr. Solange Okonjo-Pratt spent two days with the specimen and produced eleven pages of measurements, all published with this story.
She will say what it is not. It is not a bird: there is no fused carpometacarpus, and the elongate fourth-digit phalanges are not a structure any avian wing contains. It is not a bat: the arrangement is right and the scale is out by a factor of nine.
She will say what the proportions match, which is a pterodactyloid wing finger.
She will not say what the animal was, and asked us to print her reason exactly. "Anatomy tells you what a thing is built like. It does not tell you what it was, and the second sentence is where people like me get quoted for the rest of our lives."
What the laboratories got
Two 40-milligram samples went to Copenhagen under a blind code, mixed in with four controls of known species — a turkey vulture, a pelican, a flying fox and a saltwater crocodile.
Dr. Anselm Hvidberg-Rao identified all four controls correctly. Against the specimen, his collagen fingerprint returned no match to any bird, mammal or living reptile in the reference libraries he holds.
A no-match is not an identification. It is the absence of one, and it is what you would also get from a degraded sample or from an animal nobody has ever put in a library.
The dating is the harder number. Membrane: 1874 to 1908 at ninety-five per cent. Bone, submitted separately three weeks later: 1866 to 1912. The two agree with each other and neither is tight.
The seventy-one years
This is the part of the story that could still take the rest of it down, and it belongs here rather than at the end.
The paper trail runs backwards from probate in 2019, through a 1974 sale schedule, to a fire-insurance inventory of March 1961 that lists "one packing case, dry specimen, barn."
Before March 1961 there is nothing. Not a bill of sale, not a will, not a brand registry entry, not a line in a diocesan record. Seventy-one years in which the only account of where this object was is a family's memory of a family's memory.
An organic specimen dated to the 1880s can be an animal that died in the 1880s. It can also be an animal that died in the 1880s and was assembled into something else by somebody clever in 1930. The dating cannot tell those apart and neither can we.
We have looked in four county archives and expect to be told about a fifth.
The crate had two things in it
The second is larger, wrapped in oilcloth and bound with twine, and it has not been opened.
The executor will not permit it before probate closes in November, and this newspaper accepted that condition rather than lose access to the first object.
We have photographed the wrapping. That is all we have and we are not going to speculate in print about what is under it, having just spent six months proving how easily a description becomes a memory of a photograph.
What this desk takes from it
That the argument was about the wrong object for a hundred and thirty-six years.
Everybody who has ever fought about this has fought about a picture — whether it existed, who suppressed it, which magazine ran it. Nobody in that argument went and asked the county who had the barn.
The video of the opening runs four hours and eleven minutes and is published unedited, including the forty minutes in which nothing happens because the lid was stuck. Both laboratory returns are published, and so is the blind-code key, so that anyone can check which sample was which.
If a reader can show us the wire inside it, this newspaper will print that on the front page.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We went back to the one question our own August story set aside and asked the executor of a Cochise County estate whether we could open a crate. The opening was filmed unedited in front of two witnesses neither of whom works for this newspaper. Samples then went to two laboratories under a blind code, with controls of known species mixed in, and the paperwork trail was walked backwards from probate until it ran out.
Who we spoke to
- Dr. Solange Okonjo-Pratt, Comparative anatomist, vertebrate flight. Examined the specimen in person over two days; her measurements and photographs published in full 21–22 July 2026 Will describe the morphology and will not name an animal. Her reasons are printed at length in the body.
- Dr. Anselm Hvidberg-Rao, Collagen peptide analysis, Copenhagen. Sent two 40mg samples under a blind code with four control samples of known species August 2026 Identified all four controls correctly. Returned no match for the specimen against any reference library he holds.
- Dr. Mireille Sandoval-Kirchner, Radiocarbon facility, Arizona. Two independent samples, membrane and bone, submitted separately three weeks apart August 2026 1874–1908 at 95 per cent on the membrane; 1866–1912 on the bone. The two agree and neither is tight.
- The estate file, Insurance inventory 1961, sale schedule 1974, museum correspondence 1998, probate inventory 2019. Given to this newspaper by the executor, who has also given the same file to two other outlets on request July 2026 Every document in it has been read. The chain runs back to 1961 and stops.
Documents
- PX-2107 — The Tombstone Epitaph, 26 April 1890, page 4 — 'A Strange Winged Monster Discovered and Killed on the Huachuca Desert' accepted
What we could not confirm
- Where the crate was between 1890 and 1961. Seventy-one years, four county archives, two brand registries and a diocesan record office, and not one piece of paper. The family's account is that it came off a wagon and went into a barn, and we cannot corroborate a word of it. If the specimen were assembled in 1930 by somebody clever, this is exactly the hole that fact would sit in.
- The rest of the animal. We have one wing finger, one membrane and thirty-one articulated elements. There is no skull, no torso, no second wing, no tail. The 1890 account describes a whole animal and a measurement across it, and we cannot connect our object to that measurement by anything other than the crate it was in.
- The second object in the same crate. It is larger, wrapped in oilcloth and twine, and the executor will not permit it to be opened until probate closes in November. We have photographed the wrapping. That is the entirety of what we know about it and we would rather print that than nothing.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Meridian TelegraphArizona Barn Yields 'Impossible' Winged Specimen
Ran the measurement and a photograph of the crate within a day of our correspondent's first call for comment.
Had that we did not
The 2.1 metres, correctly, and the name of the county.
Left out
The seventy-one years. Their piece dates the specimen to 1890 in the first sentence on the strength of a story printed in 1890.
-
The Hollow PostTHE MUSEUM WAS OFFERED THE PROOF IN 1998 AND SENT IT BACK
Reported the 1998 refusal as a suppression, with a named curator and an implied instruction.
Had that we did not
The refusal. It happened, and the letter is real.
Left out
The letter's actual contents, which are two paragraphs about climate-controlled shelving and a suggestion of three other institutions to try. The curator they name had retired the previous year.
-
They Buriedthis newspaperWe Spent Six Months On The Photograph. The Wing Is 2.1 Metres Wide.
Went back to the one question our own earlier story explicitly set aside, and then went and opened the crate.
Had that we did not
The provenance gap, marked, at the same weight as the laboratory results.
Left out
In August we printed that whether the 1890 story is true 'is a separate question and gets conflated constantly, including by us'. We then did not touch it for seven months. A survey panel is easy to commission and a barn is four hours from the airport. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Dr. Hollis Wingate-Duran, collections manager, the state museum that was offered the crate in 1998 Sent the 1998 letters, the laboratory results and the full draft; given nine days. Answered in writing and asked that it run whole. Printed unedited.
I have read your story and I am going to correct one thing in it, and then I am going to say something worse about my own institution than anything you have written.
The correction. We did not decline to examine this object. Nobody asked us to examine it. We were asked, in a one-page letter, whether we would store it. The reply, which you have, says we had no climate-controlled shelf for a dry organic specimen of that size and suggests three institutions that did. That is a storage decision made by a collections office with a full basement, and it was, on the information in front of it, correct.
Now the worse part.
Our own accession policy, then and now, says that an offer accompanied by a provenance claim of scientific interest is to be referred to a curator before it is answered. The 1998 letter contains a provenance claim. It names the newspaper story. It was answered in eleven days by a member of staff who was doing exactly what the storage queue asked of her and was never shown the sentence in the policy that would have made her stop.
So: a specimen that two laboratories now cannot assign to any living order sat in a barn outside Bisbee for twenty-eight years because our shelving was full and our own referral rule was not read. I would rather your readers had that from me than from a leak.
On the object itself I will be as useful as I can and no more. I have not handled it. From Dr. Okonjo-Pratt's photographs and measurements, which are excellent, the phalangeal proportions are not those of any bird I know and the scale rules out every bat. I am not going to write the word I imagine you would like me to write, because I have not had the thing on a bench in front of me and because the honest position of a person who has not is silence.
What I will do is this. We have written to the executor offering to accession it, to conserve it at our cost, and to publish the conservation record whatever it says — including, if that is what it comes to, the record of us finding a wire armature inside it.
We were asked once and we got it wrong. We are asking now.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.