NMNH Home  |  What's New ?  |  Calendar of Events  |  Information Desk  |  Search

Smithsonian, National Museum of Natural History 

Natural History Highlight
 Building a Dinosaur     (Sep. 2000)
Past Highlights Past Highlights

Digital image of NMNH's Triceratops.

With the arrival of a prototype reproduction Triceratops skull in August, NMNH is nearing completion of a major project - the world's first anatomically accurate Digital Dinosaur, rendered from real fossils. The Museum's Triceratops has been the subject of intensive conservation, measurement, scientific discussion and interpretation, computer analysis, and animation.  Now, scientists can exhibit a newly-mounted Triceratops that is anatomically correct and, for the first time, model its movements to better understand the behavior of this three-horned, plant-eating animal from the Cretaceous Period, more than 65 million years ago.

(Click to see larger image)

as mounted during the 1960's

NMNH scientists from the Department of Paleobiology and the Applied Morphometrics Laboratory seized a rare opportunity when, early in 1999, it became necessary to disassemble the Museum's Triceratops for the purpose of cleaning and massive conservation measures.  This emergency permitted scanning of all the bones of the assembled skeleton prior to disassembly, in addition to doing so again while the disassembly was underway.

  The Smithsonian’s  Triceratops first went on exhibit in 1905, and had gone relatively unmodified since that time.  After nearly 100 years on display, conservator Cathy Hawks found in 1998 that the bones had weakened; probably due to "pyrite disease," temperature and humidity variation, accumulated effects of vibration, failed glue and bone hardeners, and uneven distribution of weight on the supporting metal armature - not surprising “wear and tear” considering  Triceratops’ 93-year stance against the ravages of time. This is alarming for a favorite dinosaur in the most popular exhibit in the Museum.

Conservation needs were not the only reason to embark on this ambitious project. Scientists can now correct some inaccuracies in the mounted fossil. The bones in the original 1905 Triceratops mount were not from just one individual, but rather the mount was a composite of the bones of at least ten different animals.  Unfortunately, these animals were of different sizes and from different places.  There still is no known complete skeleton of Triceratops anywhere.  In 1905, because no specimens of Triceratops had ever been found with feet, the preparators did the best they could by using foot bones of about the right size and shape, which were later identified as the hind feet of a hadrosaur or “duck-billed” dinosaur.  Despite being a composite, not only did the Smithsonian have the first-ever mount of Triceratops in the world, but it still is one of the best examples of this horned dinosaur ever assembled.  Advances in technology and in scientific interpretation now make it possible to correct shortcomings in the old mount while at the same time exploring new ideas about Triceratops, and acting to preserve the fossils themselves.

(Click to see larger image)

From left, Pete Kroehler, Steve Jabo and Fred Grady examining the skull as it is taken off of exhibit. Photo by Chip Clark.

As the skeleton was disassembled, two of the Department of Paleobiology’s fossil preparators, Steve Jabo and Pete Kroehler, treated each bone with cleansers, hardeners, and glue to halt the on-going damage.  Cleaning consisted mainly of delicate dusting, and where necessary, light vacuuming.  Hardeners and glues, on the other hand,  are state-of-the-art.  The hardeners, for example, are PVA (polyvinyl acetate) and butvar (polyvinyl butyral).  To fill cracks from stress and openings left in the bone after removing fasteners that held each bone to the armature, a mixture of butvar and papier-maché was applied.  Once these repairs were done, three different glues were used to bond broken pieces:  paleobond, or cyanoacrilate; thick butvar; and five-minute epoxy.  Yet another destroyer of fossil bone is the naturally ocuring mineral pyrite. After prolonged exposure to high humidity, pyrite disease will cause the bone to fall apart if the pyrite is not neutralized; paleontologists call this "pyrite disease."   The remaining pyrite pieces that couldn’t be extracted were coated with hardener so no oxygen could react with them, thereby halting the progression of this destructive chemical reaction.  After all these conservation steps had been taken, Jabo and Kroehler then began making molds and casts of the bones to use in a replacement cast mount of the skeleton. 

Scanning the Bones 

Ralph E. Chapman, head of NMNH’s Applied Morphometrics Laboratory, initiated the scanning work.  He teamed up with Arthur Andersen, head of Virtual Surfaces Inc., Mount Prospect, Illinois and Lisa Federici of Scansite 3-D Services, San Rafael, California, to develop the project and digitize the bones.  The first step in the process was to digitize key points in the original dinosaur mount before it was disassembled so that if need be, it could be restored to its original position.  Next, each bone was scanned by Henry Wede of Steinbichler-Dimension Data by means of a surface scanner. Then individual computer files were built for each bone, leading to more than 200 files, equaling at least 20 gigabytes  of data.

The History of the 
Original Specimen

(Click to see larger image)

Triceratops with Norman Boss, upon 
completion of move to NMNH

How did Triceratops come to be on exhibit?  The story goes back over 100 years.  In 1888, John B. Hatcher was working in Wyoming for the famous paleontologist, O. C. Marsh, professor at Yale and Vertebrate Paleontologist for the United States.  While in the town of Douglas, Hatcher was introduced to Charles A. Guernsey, owner and manger of the 999 cattle ranch.  Mr. Guernsey showed Hatcher some examples from his considerable collection of fossils.  Among them was a fragment of a very large horn core that had been taken from a skull several feet in length which had been found by the ranch foreman, Edmund B. Wilson, at the bottom of deep canyon on the ranch in Converse (now Niobrara) County.  Owing to the press of business, Guernsey offered to conduct Hatcher to the locality at some future time.  During the following field season, on May 7, 1889, Mr. Wilson escorted Hatcher to the canyon and showed him the remainder of the skull belonging to the dinosaur that Marsh was to name Triceratops that very same year.  Between 1889 and 1892, Hatcher collected bones from at least 50 individual horned dinosaurs in Wyoming, more than 30 of which included parts of the skull.  Much of that material became part of the United States Geological Survey collections, which were then transferred to the Smithsonian Institution where they are currently part of the collections of the NMNH Department of Paleobiology.

With assistance from Hatcher, who had trained them on their first job at the Carnegie Museum of Natural History, two new arrivals at NMNH, Charles W. Gilmore (later to become a curator in the Museum) and Norman H. Boss (later to become Chief Preparator) used some of these bones to construct the world’s first mounted skeleton of Triceratops.  Their mount was just under 20 feet long and just over eight feet tall at its highest point.  It went on display in what is now the Arts and Industries Building in 1905, and eventually was moved to the Dinosaur Hall of the Natural History building after the building opened in 1910.  This mount has served as a model around which the current concept of Triceratops, the living animal, was interpreted, although the ideas of what Triceratops looked like and how it lived have varied considerably through the years. 

The extensive data files started as large clouds of location points in three-dimensions, representing the external surface and characteristics of each bone.  Arthur Andersen used the scan points first to built a virtual model of each bone; and then the final, composite skeleton, composed of three-dimensional images in the computer.  

New Bones from Old

Knowledge accumulated about Triceratops’ anatomy since 1905, and compromises due to the composite nature of the original mount were accommodated prior to prototyping new bones.  The Museum’s staff knew the dinosaur’s left humerus, or upper forelimb, was a smaller mismatch to the right humerus, and that the back feet were from a different dinosaur group altogether, the skull was too small for the body, and other bones were also mismatched in size or so fragile that they should not be on exhibit.  Arthur Andersen created a new, full-sized version of some of the problematic bones: the left humerus, left ilium, and left scapula, by making mirror images of the complementary bone from the opposite side.  Satellite Models of Mountain View, California, constructed full-size replicas of the mirrored bones by using a material called Ren Shape, which was then reproduced by making molds, and casting the final versions to be used in the mount.  The prototyping process involved using the computer data files to direct the milling of a block of Ren Shape with progressively smaller and more accurate drill heads until the final product was a very accurate representation of the original.  Also, a miniature prototype was made by Jason Dickman, of Hasbro, Cincinnati.

During disassembly of the skeleton, Jabo and Kroehler determined that the skull of NMNH's Triceratops was too small for the rest of the body by about 15 percent, based upon studies of the proportions of the various bones of the body from other articulated specimens.  

In April, 2000, a consortium of paleontologists convened at NMNH to determine the animal's proper posture and articulation when alive, and examine the miniature prototype of Triceratops.  Kent Stevens of the University of Oregon, Rolf Johnson of the Milwaukee Public Museum, Cathy Forster of SUNY Stony Brook, Brenda Chinnery of The Johns Hopkins School of Medicine, and NMNH staffers Michael Brett-Surman, Ralph Chapman, Steven Jabo and Peter Kroehler met at NMNH for a day of discussion, comparisons and the best kind of intellectual "play."   

Using the 1/6 scale prototype, they could easily manipulate the model and test different ideas about Triceratops’ range of motion.  NMNH staff were exploring the various postures Triceratops could take as they planned for the new mounting. The paleontologists compared the computerized scans of Triceratops’ bones with the different positions using the miniature prototype, and shared a sense of what a skeleton of a complete single Triceratops should look like.

To remedy the disparity of bone size, posture, and articulation for the new mount, data were supplied to Shared Replicators and Tulsa Technology Center to create a prototype of an enlarged version of the original skull, using a form of stereolithography that produces high-resolution versions of the original in lightweight plastic material.  This process uses a laser to shape and cure a light-sensitive epoxy-resin

(Click to see larger image)

Dinosaur specialist Mike Brett-Surman among the collections.  Photo by Chip Clark

Conserving Dinosaurs

When they are not on exhibit, virtually none of the Museum’s ninety or so genera of  dinosaur fossils is kept as an assembled skeleton.  Instead, their components have individual “jackets” or storage supports.  These jackets are made by specially trained museum technicians and conservators who build a removable shell of plaster and fiberglass, padded with an ethafoam lining, custom-made for each bone. Each jacket protects one specimen, providing some buffering against vibrations and variations in temperature and humidity.  The jackets also make handling these large, heavy fossils much safer and easier to do. 

How do vibrations and swings in temperature and humidity hurt a fossil?  Fossilized bone can still absorb moisture.  Temperature and humidity changes affect bone and yield results much like the “heave and thaw” effect seen in the ground during the freeze-thaw cycles in winter time.  Vibrations, even if they seem unnoticeable or minor to us can literally shake a museum specimen apart over time.  When these two types of damage are combined, the results can be disastrous for the long-term preservation of a museum specimen.
Building the New Mount 

Casts of the newly made humerus, ilium and scapula; the new computer-modeled skull, seven feet in length; and new hind feet will be remounted together with casts of the rest of the bones to yield an up-to-date model of NMNH's Triceratops.  Using casts has a lot of advantages: the new “bones” are lighter and easier for mount-makers to handle. The casts are sturdier than the fossil bones, and will withstand the conditions of an exhibit hall. If casts are damaged, they can be replaced using new casts from the existing molds.  Casts can also be drilled to hold the supporting hardware so that the finished mount looks more like a real skeleton, not one visibly held together by metal rods and screws, all of which means that model's final pose can be more realistic and dramatic. 

What is Triceratops and 
when did it

(Click to see larger image)

Artist Charles R, Knight’s painting of 1901, which served as the basis for Gilmore and Boss’s reconstruction of Triceratops.

Triceratops was a four-footed, plant-eating dinosaur that lived during the Cretaceous Period of geologic time, more than 65 million years ago, in the region that is now the Rocky Mountain area of the United States.  Triceratops was one of the last surviving dinosaurs of the Age of Reptiles.  These animals are characterized by three horns on the front of the skull: two massive brow horns located just above the eyes and a smaller nasal horn right above the snout.  Triceratops also had a large frill on the back of the skull.

These dinosaurs lived in herds.  Evidence for herding behavior comes from several sources.  The large number of fossil skeletons from closely related species that were found in “bone beds” or large bone deposits that suggest that a large number of these animals lived together and were all killed at the same time.  Large numbers of Triceratops skeletons found also suggests that Triceratops was very numerous and by implication, quite successful for its time.   Similarly, preserved footprint track ways show several dinosaurs moving about in the same direction at the same time.  In herds, the young are in the center, flanked by adults as a means of protection against predators.  Scientists also believe that  the size of the brow horns combined with scars on the frills, indicates that Triceratops used its horns for defense.  The horns may have been used for fighting others of their kind for position in the herd, for territorial defense, or for defending themselves against predators such as Tyrannosaurus, which lived at the same time as Triceratops.

The new Triceratops should be on display by mid-winter.  It will be located directly across the exhibit gallery from Tyrannosaurus. With the new potential for posing Triceratops in a more life-like way, the new mount will better show what the scientists believe its behavior was like.  Some real fossil material will again be put on display, but will be in a more environmentally controlled setting.

What have Researchers learned so far about Triceratops during this 
process ?

The computerized scans of Triceratops hold a wealth of information, which researchers are only beginning to tap.  To date, some of the more interesting findings related to the Digital Dinosaur project are:

Ralph Chapman and visiting scientist Kent Stevens (University of Oregon, Eugene) noticed that something new about the elbow of Triceratops.  They cut up a mouse pad to simulate cartilage between the humerus and ulna, and rotated the joint.  They observed that Triceratops’ elbow could have locked in place, much as horses and cows can lock their elbow joints to sleep while standing. Researchers have also determined that the hind legs could lock.

Conservators found "pyrite disease" in the right humerus bone as they disassembled the mount - a highly destructive situation that could not have been detected without disassembly.  This has led the researchers to consider even more carefully the relative merits and detractions of placing original fossil material on exhibit. 

Using the 1/6 scale model, as researchers examined the condyle at the back of the skull and first vertebra they were struck by how well balanced the skull was on the vertebral column.  The bones’ shape would allow Triceratops great flexibility in moving its head.  Given the very large size of Triceratops’ head among terrestrial animals, the balance and fluidity of movement possible are especially remarkable.

Researchers can now calculate the volume of  Triceratops’ skull - all 183,000 cubic centimeters worth, using the three-dimensional measurements maintained in computer files.  Armed with that information, they can revise earlier estimates of the skull’s weight. Triceratops’ head is now believed to have weighed about 400 pounds, about 2/3 the weight that was previously thought.

QUESTION:   How do you move a 7-foot-long Model dinosaur skull?
 Very Carefully, especially when that skull must travel from Tulsa, Oklahoma to Washington, DC.  The material from which the new Triceratops' skull is made is light and heat sensitive.  The Smithsonian appreciates the special flight that Evergreen Aviation International, Inc, of McMinnville, Oregon, made to bring the skull to Washington.

visit these links:
All the organizations and individuals mentioned here donated their time, talent, and material to this project.  But none of it would have been possible without the imagination and leadership of Richard H. Benson, Chair of the Department of Paleobiology - Many Thanks to One and All !


"Natural History Highlight" features interesting and exciting activities and objects from the Museum.  We will frequently introduce new highlights that come from our research, collections, exhibits, and projects.      
Past Highlights

NMNH Home  |  Information Desk  |  What's New?  |  Calendar of Events  |  Research & Collections  
Discovery Center (IMAX)
  |  Exhibits  Educational Resources  |  Natural History Highlight | Search

March 2001 -  NMNH Webmaster