The many lives of Triceratops skull No. 21

Spanning from Wyoming to the Peabody and then the Netherlands by way of stormy seas, the journey of this unique fossil has been one for the history books.
Triceratops skull No. 21

A 3D model of the restoration of Triceratops skull No. 21, which the Peabody traded to the Delft University Geological Museum in the 1950s. The restoration is underway in the Netherlands. (Photo credit: Javid Jooshesh, Naturalis)

In November 1956, a Yale-owned Triceratops skull made a perilous transatlantic journey from the Peabody Museum of Natural History to the Delft University Geological Museum in the Netherlands.

The ship carrying the tri-horned herbivore’s head, which the Peabody had agreed to trade for a collection of marine invertebrate fossils, encountered rough seas, and the 1,000-pound skull, though carefully packed, jostled violently in the cargo hold. More fatefully, its crate was apparently dropped upon arrival in Delft: The skull, which had been painstakingly reconstructed from fossil fragments at Yale, smashed into hundreds of pieces.

Curators in the Netherlands again reconstructed the shattered treasure, using plaster and wire frames to fill in gaps caused by the damage and shortcomings in the original specimen. Then they put it on display, and there the matter rested until January 2015.

That was when Daniel Brinkman, a museum assistant in the Peabody’s Division of Vertebrate Paleontology, fielded an inquiry from a researcher at the Naturalis Biodiversity Center, the Netherlands’ national natural history museum. The Dutch were planning a third, more scientifically accurate, reconstruction of the skull, and the Dutch researcher, Dylan Bastiaans, sought information about “Triceratops skull 21.”

It came in out of the blue,” said Brinkman, who had been unaware that a Triceratops skull had left the Peabody for Delft in 1956. “Dylan was hoping we had some of the missing pieces that for one reason or another weren’t included in the original reconstruction.”

The Dutch team works on reconstructing Triceratops skull 21.
The Dutch team works on the skull reconstruction. They are using data from skulls that were on view at the Peabody to design and create missing pieces with a 3D printer. (Photo credit: Hanneke Jacobs, Naturalis)

Bastiaans provided a photograph showing that the skull bore the Peabody’s standard late-19th-century field-number designations, a promising start. At first, Brinkman couldn’t find any records of the fossil exchange. But the field number strongly suggested Bastiaans was referring to Triceratops specimen YPM VP 001832 — a skull collected during an 1891 Yale expedition in the Lance Formation in Wyoming led by John Bell Hatcher 1884 B.Phil., a protégé of famed Yale paleontologist O.C. Marsh.

Digging into the division’s library, Brinkman made the big find he needed — a 1986 paper co-authored by the renowned Yale vertebrate paleontologist John Ostrom. It confirmed that “skull 21” was, in fact, the specimen Hatcher excavated in 1891. Brinkman forged ahead, scouring the vertebrate paleontology collections in the Peabody’s basement, and ultimately came up with four drawers of fossilized bones belonging to the triceratops, including pieces of cranium.

But he still had no record of the swap.

Then Bastiaans found one: A Dutch publication had cited the deal, placing it in time. This prompted Brinkman to ask Yale colleague Jessica Utrup, a museum assistant in the Division of Invertebrate Paleontology, to search the archives of Yale invertebrate paleontologist Carl Owen Dunbar, who had directed the Peabody Museum from 1942 to 1959.

Utrup discovered a sheaf of related correspondence.

In a letter to a Dr. Kruizinga dated Oct. 19, 1953, Dunbar expressed admiration for the “fine collection of Triassic and Permian of fossils from Timor” that had arrived that summer. He promised to send a “good skull of a Triceratops.”

It was Dunbar who selected “skull 21” for the exchange. He promised that Yale would reconstruct the skull, then in pieces and still partially embedded in rock, before shipping it to Delft. The work took years due to preparators’ workload, according to the correspondence. Eventually, in the fall of 1956, the skull made its fateful journey across the Atlantic.

A Triceratops skeleton on display in the Netherlands
“Dirk,” a Triceratops skeleton on view at Naturalis, the Netherlands' national natural history museum. Data from Peabody specimens contributed to the reconstruction of Dirk’s skull. (Photo credit: Mike Bink Fotografie)

While the Dutch geologists’ original reconstruction, finished in 1961, was precise in places, it left the specimen with scientifically inaccurate features, including the orientation of its horns and frill — the bony shield at the back of the skull — Brinkman and the Dutch researchers said. The Dutch had decided a complete overhaul was necessary to make the specimen truly suitable for study and exhibition.

In the summer of 2018, Bastiaans and colleague Valentin Vanhecke came to New Haven to make 3D scans of the Peabody’s three best Triceratops skulls, then exhibited in the Great Hall. (One is now displayed outside the new O.C. Marsh Auditorium in the new Yale Science Building. Marsh used this skull specimen in 1889 to first describe and name the Triceratops species.)

Additionally, the Peabody shipped to the Netherlands about 80 pounds worth of skull fragments Brinkman had rediscovered.

Today, the Dutch conservators are using information from the scans made in New Haven to prepare the completer and more accurate specimen. They’ll use 3D printers to make pieces for filling the final gaps in the skull.

It’s like putting together a jigsaw puzzle that’s missing some pieces,” said Brinkman.

The accuracy of the 3D scans and printing eliminates much of the guesswork and artistic license that can accompany fossil reconstructions using plaster and clay, said Anne Schulp, a Dutch paleontologist and researcher at Naturalis.

It’s real data from complementary specimens, which makes it much easier to accurately connect the dots,” he said.

The Dutch conservators also used information from the Peabody skulls to help reconstruct the frill on a separate, nearly complete Triceratops skeleton, nicknamed “Dirk,” that went on view in September at Naturalis’ new museum in Leiden.

We had scans from a variety of skulls, but the data from the Yale collection was by far the best match,” Schulp said. “It was extremely helpful.”

Schulp and his colleagues had excavated Dirk in the Lance Formation — near the place where John Bell Hatcher collected the Peabody’s specimens more than a century ago.

Once completed, the now thrice-reconstructed Triceratops Skull No. 21 will be exhibited at Delft University of Technology.

It’s a famous local personality there,” Schulp said.

Share this with Facebook Share this with X Share this with LinkedIn Share this with Email Print this

Media Contact

Bess Connolly :,