Nanosaurus agilis: the smallest dinosaur you've never heard of (and for good reason)

Many millions of years ago, in a time that we would call the Jurassic and in a place we would call Colorado, small bipedal herbivorous dinosaurs frolicked and otherwise did things appropriate to small bipedal herbivorous dinosaurs. In their time, they died and a very, very few were selected by taphonomy to be fossilized. Of that tiny number, an even smaller subset have happened to be exposed at the surface at the right time and place to be found by a similarly tiny number of human beings who were specifically looking for such things. Having been found, their remains were sent off to be studied by another tiny number of people who had a lot of things on their minds, living in a world that has had little use for small bipedal herbivorous dinosaurs except as props to show off the (speculated) abilities of small bipedal carnivorous dinosaurs. It's not really that surprising that some of them have fallen through the cracks. Then there's Nanosaurus agilis.

When Othniel Charles Marsh began working on the great dinosaurs of the American West, he didn't lead off with the most impressive material. We already met Atlantosaurus (alias Titanosaurus), his first shot at a Morrison sauropod. It didn't take. Shortly thereafter, he snuck the subject of this post into a grab-bag of new fossil vertebrates from multiple states and formations (Marsh 1877a). This isn't to suggest that the contents of this grab-bag are unworthy; you've probably heard of Moropus, which is named in this publication, and the freshwater ray Heliobatis radians is also noteworthy. Nanosaurus agilis, regrettably, is dragging down the average, although not as badly as Apatodon mirus (Marsh's Mesozoic pig, which turned out to be part of an eroded dinosaur vertebra).

Nanosaurus agilis shows up at the end of the sixth page of the article. Having given the world a giant sauropod, Marsh now reported that he had the smallest dinosaur yet discovered, which he said came from "Mesozoic deposits of the Rocky Mountains." (Any competent editor would be honor-bound to murder you if you tried to get away with that today, but stay with me here; there's a payoff.) Nanosaurus agilis was "not larger than a cat" yet an adult, to which we can only grin indulgently at today, and its bones were hollow. Marsh mentioned lower jaw, tibia, and femur; the femur is stated as 63 mm (about 2.5 in) and the tibia 75 mm (just shy of 3 in). These seem to be kind of small to be comparing to a cat. Marsh provided zero speculation on what Nanosaurus agilis might have looked like, what it was related to, etc., which is one way to avoid making mistakes. Following N. agilis in the article, he introduced a second species, N. victor. This later got its own genus (Hallopus) and turned out to be a sort of small terrestrial crocodile relative thing.

Marsh wasn't quite through with Nanosaurus. He gave it a third species, N. rex, later that year (Marsh 1877b), apparently using the same principle to name it that people use when they give tiny dogs tough names, and gave the genus its own family (Nanosauridae, although he neglected to do any descriptive work) a year later (Marsh 1878). Finally, in 1896, nearly 20 years after naming it (a period of time surely greater than the actual lifespan of any Nanosaurus agilis), Marsh gave some useful details and even a couple of illustrations. Nanosaurus agilis was based on an associated partial skull and skeleton found on a single slab, split partially through some of the bones. Marsh firmly assigned the species to his Ornithopoda and regarded it as a herbivore. He considered it one of the most bird-like ornithopods known, and wrote that "nearly all, if not all, the bones preserved might have pertained to a bird, and the teeth are no evidence against this idea". This time around N. agilis was "about half as large as a domestic fowl". Marsh also finally gave a bit more specific information on the provenance, writing that it was found in Colorado in lower beds than the Morrison (his "Atlantosaurus beds"), along with Hallopus, former N. victor.

Go Nanosaurus agilis! (from Marsh 1896)

Marsh's initial coyness about the provenance may have been for a very good reason, beyond protecting productive sites and poor collection of stratigraphic information: he wasn't supposed to have Nanosaurus agilis at all. Oramel Lucas, who collected it, was working for Edward Drinker Cope. Marsh's assistant Benjamin Mudge visited Lucas in Cañon City, Colorado, and discovered Lucas was not entirely happy with his arrangements with Cope. Mudge managed to convince Lucas that the agreement only concerned large fossils, so Lucas was free to sell small fossils to anyone, such as Marsh. Thus did Marsh acquire N. agilis (Jaffe 2000). (This left a Nanosaurus-shaped hole in Cope's career, which he unsuccessfully tried to fill with two species of the immortal ?ornithopod Tichosteus. One of them he named T. lucasanus in honor of Lucas, although I can't help feeling he was being sarcastic with that ending.) It is entirely understandable if Marsh did not want to elaborate on who found the fossils, or where they came from.

It also just goes to show what comes of deception, or at least carelessness. For decades following Marsh's passing, there have been two running threads about Nanosaurus agilis: its minute size, and trying to pin down Marsh's enigmatic "Hallopus beds". For a long time, people were convinced N. agilis was from the Late Triassic or Early Jurassic (e.g., Lull 1910). Charles Schuchert, decades removed from controversies with Minnesota brachiopods, seems to have been the first to place the discovery site of N. agilis and Hallopus in the upper Morrison Formation (Schuchert 1939). There's been an interesting disconnect in the past few decades between dinosaur researchers, who have generally adopted a Morrison Formation provenance for N. agilis, and crocodile researchers, who have been more cautious concerning Hallopus; see Leardi et al. (2017) for the most recent summary from the croc standpoint. It may have to do with the stakes involved: for dinosaur researchers, Nanosaurus agilis is a minor curiosity, while for croc researchers, Hallopus is a frustrating but important piece of a story of strange terrestrial crocs. Of course, when you get right down to it the Hallopus/Nanosaurus connection is probably overstated; they certainly had different collectors (Hallopus victor was purchased by David Baldwin from a curio shop in Colorado Springs; Ague et al. 1995) and may well have come from different horizons (Galton 1978; Ague et al. 1995). The Hallopus slab and Nanosaurus agilis slab have different mineralogies, indicating at least different environments (coarse channel for Hallopus, finer-grained channel or bank deposit for N. agilis; Galton 1978).

Scientifically speaking, Nanosaurus agilis has mostly been glossed over since Marsh's time. He had much better material of somewhat larger ornithopod "Laosaurus", some of which became Dryosaurus and some of which became Othnielosaurus, so he wrote more about it, but that's another story. About the only publications to take it seriously since the beginning of the Dinosaur Renaissance era are Galton 1978 and Galton 2007. The immediate problem is that the type specimen is lousy. It is actually part of an ornithischian hash slab, catalogued as YPM 1913, dotted with pieces and molds of small dinosaur bones. The most notable bones are that lower jaw and ilium Marsh depicted, and several limb bones. There are at least two individuals included, due to the presence of two right femora. Galton (1978) selected the jaw bone as the type specimen within this mess and felt that the ilium and most of the limb bones came from the same animal. He classified it as a fabrosaurid. Fabrosauridae is kind of hard to describe years later, but it was something like a club for unspecialized bipedal ornithischians that weren't known to be heterodontosaurs or hypsilophodonts. "Known to be" is important here, because one of the flagship members of Galton's Fabrosauridae was Echinodon, which has since been reclassified as a heterodontosaurid.

YPM 1913, as it looked more than a hundred years ago (Huene and Lull 1908). Name that bone!

Perhaps you'd enjoy a descriptive illustration with a German caption? (Huene and Lull 1908; Galton 1978 reidentified some of the bones, such as the "Hum" which is the femur of the second dinosaur.) In the photo there's a bone-looking light area near the upper ends of the "Tib dxt" and "Fem sin" but not here, so maybe it's a trick of the light on an uneven surface? (Note: I can't read German, so I don't know if the article addresses it.)

Galton (2007) revisited Nanosaurus agilis briefly, long after Fabrosauridae had evaporated with the coming of the cladistic era. By this time, he was firmly placing the species in the Morrison ornithischian fauna, which had expanded to include what would soon be named Fruitadens, and Drinker, which like Othnielosaurus and "Nanosaurus" rex is not quite an ornithopod and not quite anything else. Nanosaurus agilis remained just out of reach, not the same as anything else but not well-preserved enough to be of much use. Will the Morrison, or some other formation, one day spit out another?

References

Ague, J. J., K. Carpenter, and J. H. Ostrom. 1995. Solution to the Hallopus enigma? American Journal of Science 295:1-17.

Galton, P. M. 1978. Fabrosauridae, the basal family of ornithischian dinosaurs (Reptilia: Ornithischia). Palaeontologische Zeitschrift. 52(1/2):138–159.

Galton, P. M. 2007. Teeth of ornithischian dinosaurs (mostly Ornithopoda) from the Morrison Formation (Upper Jurassic) of the western United States. Pages 17–47 in K. Carpenter, editor. Horns and beaks: ceratopsian and ornithopod dinosaurs. Indiana University Press, Bloomington and Indianapolis, Indiana.

Huene, F. v., and R. S. Lull. 1908. Neubeschreibung des Originals von Nanosaurus agilis Marsh. Neues Jahrbuch für Mineralogie, Geologie und Paläontologie 1908:134–144.

Jaffe, M. 2000. The gilded dinosaur. Three Rivers Press, New York, New York.

Leardi, J. M., D. Pol, and J. M. Clark. 2017. Detailed anatomy of the braincase of Macelognathus vagans Marsh, 1884 (Archosauria, Crocodylomorpha) using high resolution tomography and new insights on basal crocodylomorph phylogeny. PeerJ 5:e2801 https://doi.org/10.7717/peerj.2801

Lull, R. S. 1910. Dinosaurian distribution. American Journal of Science, series 4, 29(169):1–39.

Marsh, O. C. 1877a. Notice of some new vertebrate fossils. American Journal of Science, series 3, 14:249–256.

Marsh, O. C. 1877b. Notice of new dinosaurian reptiles from the Jurassic formation. American Journal of Science, series 3, 14:514–516.

Marsh, O. C. 1878. Notice of new dinosaurian reptiles. American Journal of Science, series 3, 15:241–244.

Marsh, O. C. 1896. The dinosaurs of North America. U.S. Geological Survey Annual Report 16:133–414 (the text ends at 244; the rest is plates).

Schuchert, C. 1939. The geological horizon of the dinosaurs Hallopus and Nanosaurus agilis. American Journal of Science 237(1):19–26.