As a small child, this is the first picture that I ever saw of Ceratogaulus – in my trusty How and Why Wonder Book of Prehistoric Mammals, 1964 (© John Hull/Transworld)
Horned rodents, devil's corkscrews, and terrible snails may not seem to have a lot in common, but in reality these three ostensibly separate strands are intricately intertwined within a singularly unusual, interesting chapter in the history of zoological discovery, as now revealed.
It all began in 1891, when geologist Dr Erwin H. Barbour from the University of Nebraska was shown some extraordinary formations by local rancher Charles E. Holmes in the Badlands of northwestern Nebraska, USA. Holmes and Dr Barbour colloquially dubbed them 'devil's corkscrews', as they did indeed resemble gigantic subterranean screws, each one penetrating several metres below the earth's surface, and constituting an elongated spiral of hardened earth.
Daimonelix, illustration from 1892 (public domain)
Dr Barbour proposed that these were the fossilised remains of giant freshwater sponges, his theory having been influenced by the belief current at that time that the deposits in which they occurred, and which dated to the Miocene epoch approximately 20 million years ago, were the remains of a huge freshwater lake,
Moreover, recalling the informal 'devil's corkscrew' nickname that he and Holmes had coined for them, in a short paper published by the journal Science in 1892 Barbour gave to these perplexing structures the formal scientific name Daimonelix ('devil's screw'), sometimes spelled Daimonhelix or Daemonelix in later works. Not everyone, however, was convinced by his theory that they were prehistoric sponges.
A number of authorities favoured the possibility that they were artefacts, each one having been created by the intertwining of roots from some form of prehistoric plant that had subsequently rotted away (or even by pairs of prehistoric plants, one coiling tightly around the other), with the spiral-shaped space that they had left behind becoming filled with mud, ultimately yielding one of these remarkable giant underground 'screws'. And once subsequent research had shown that the deposits containing them were not the remains of a lake at all but were associated with semi-arid grassland instead, even Barbour quietly abandoned his freshwater sponge proposal in favour of the plant theory.
However, the name Daimonelix remained valid, because although scientific genera and species names are generally given only to organisms (modern-day or fossil), a notable exception to this nomenclatural rule concerns ichnofossils or trace fossils. These are fossils not of organisms themselves but of the traces left behind by them, such as footprints, burrows, coprolites, feeding marks, plant root cavities, etc, and they too receive scientific genera and (sometimes) species names.
Daimonelix, fossil rodent burrow, Sioux County, Nebraska, Early Miocene, close-up (public domain)
A third theory concerning the nature of the devil's corkscrews was put forward by Dr Theodor Fuchs and Edward Drinker Cope, who independently suggested in 1893 that they were the fossilised burrows of a Miocene rodent. This notion attracted appreciable interest – but if true, what kind of rodent could have been responsible? One candidate favoured in various popular-format publications for quite some time during the 20th Century was a creature no less extraordinary than the corkscrews themselves.
In 1902, Dr William D. Matthew published a paper in the Bulletin of the American Museum of Natural History in which he formally described a new species of fossil rodent hailing from Colorado and dating back to the Miocene, but which was so different from all previously recorded species that it also required the creation of a new genus. Based upon a skull found in 1898, he named this novel creature Ceratogaulus rhinocerus – a very apt name, because, unique among all rodents at that time, it bore a pair of short but very distinctive vertically-oriented horns, sited laterally upon the dorsal surface of its nasal bones' posterior section.
In later years, three additional horned species were discovered and named – Ceratogaulus anecdotus, C. hatcheri, and C. minor. Some of these were initially housed in a separate genus, Epigaulus (created in 1907), and C. minor has been reassigned by some workers to the related genus Mylagaulus, but the current consensus is that all four belong to Ceratogaulus. In addition, a fifth horned species, but which unequivocally belongs to the genus Mylagaulus rather than Ceratogaulus, was scientifically described as recently as 2012. Named Mylagaulus cornusaulax, it lived in western Oklahoma during the Miocene. Four other Mylagaulus species (not counting C. minor if classed as belonging to this genus) are also known, but none of these was horned.
Known technically and collectively as mylagaulids, the horned rodents and several closely-related genera of non-horned species constitute an entirely extinct taxonomic family, existing from the Miocene to the Pliocene and (in the case of the horned species) unique to North America, but belonging to the squirrel lineage of rodents (Sciuromorpha). Moreover, examination of complete and near-complete skeletal remains has revealed that they superficially resembled marmots and other ground squirrels too, both in size (measuring roughly 60 cm long) and in overall appearance – except of course for the five horned species' nasal horns, which make them the smallest horned mammals known to science. The horned species are sometimes colloquially referred to as horned gophers, but this is a misnomer, because gophers are only very distantly related to them. 'Horned marmot' would be a much more appropriate name.
Suggestions that the devil's corkscrews could be the fossilised remains of burrows excavated by these rodents, utilising their horns, attracted interest, and remained in contention as the solution to this longstanding mystery until as recently as the 1970s (my little How and Why Wonder Book of Prehistoric Mammals was still supporting it back in 1964). However, studies focusing upon the precise conformation of their horns and speculating upon what this conformation indicated in relation to their possible functions revealed that such an idea was inherently and fatally flawed. Both the position and the shape of the horns are inconsistent with their being efficient digging tools.
By being located on the posterior rather than the anterior section of the nasal bones, the horns could not be used for digging through earth without the animal's muzzle constantly getting in the way, severely impeding the efficiency of this activity. Moreover, in later species the horns were positioned even further back than in the earlier ones, so it is evident that these rodents' evolutionary development became increasingly contrary to their horns being used as digging tools. The horns' very broad, thick shape also argued persuasively against their effectiveness as digging tools (it is nowadays believed that they served as defensive weapons instead). And so too did the telling fact that no remains of horned rodents discovered in direct association with devil's corkscrews had ever been documented.
But if the horned rodents were not responsible for these structures, then what was? As far back as 1905, Dr Olaf A Peterson from the Carnegie Museum had revealed that some of them contained fossilised bones from Palaeocastor fossor and P. magnus - two prehistoric species of small terrestrial beaver. They had existed in Nebraska and elsewhere in North America's Great Plains region during the late Oligocene and Miocene epochs. However, it was not until 1977 that their responsibility for creating the devil's corkscrews was confirmed, via a scientific paper published in the journal Palaeogeography, Palaeoclimatology, Palaeoecology, and authored by Drs Larry D. Martin and D.K. Bennett.
In it, the authors disclosed that these enigmatic underground spirals were in fact the helical shaft sections of Palaeocastor burrows, each complete burrow consisting of a single entrance mound, a long spiralled shaft, and a lower living chamber. These burrows also possessed interconnecting side-passages, and the authors' paper revealed that very extensive subterranean Palaeocastor colonies had existed (Dr Martin had discovered one that contained over 200 separate burrows), which were comparable in size and network complexity to the underground labyrinthine 'towns' or 'cities' produced by those modern-day North American ground squirrels known as prairie dogs.
In addition, Martin's research at the University of Kansas had uncovered that the beavers excavated these screw-shaped burrow shafts with their incisor teeth, not with their claws (as various previous proponents of a rodent origin for such structures had wrongly assumed). For instead of finding narrow claw marks on the burrow walls, which is what he had expected, Martin instead discovered numerous broad grooves – which he was able to duplicate exactly by scraping the incisors of fossil Palaeocastor skulls into wet sand. The very regular spirals of their burrows' shafts (i.e. the devil's corkscrews) had been constructed by the beavers via a continuous series of either left-handed or right-handed incisor strokes.
And as final proof that Palaeocastor was indeed the engineer of the devil's corkscrews, the wider chambers immediately below these spiralled shafts were sometimes found to contain perfectly-preserved fossil skeletons of adult beavers and beaver cubs, thereby verifying that they were indeed the burrows' living quarters for these beavers.
After almost a century, the mystery of North America's devil's corkscrews was a mystery no more; but across the Atlantic in England, an equally spectacular edifice of spiralled structure has continued to baffle the scientific world. Its name? Dinocochlea – 'the terrible snail'.
In 1921, during the construction of a new arterial road near Hastings in the Wealden area of Sussex, an enormous spiral-shaped object was uncovered and excavated from early Cretaceous clay after having been spotted by site engineer H.L. Tucker. Outwardly it resembled the spiralled shell of certain marine gastropod molluscs, in particular those of the genus Turritella, which is represented by numerous living and fossil species.
Accordingly, when it was formally described in 1922 by London's Natural History Museum molluscan specialist Dr Bernard B. Woodward within the Geological Magazine, he named it Dinocochlea ingens, and did indeed categorise it as a fossil gastropod, albeit one of immense proportions.
Measuring more than 2 m in length, it was far bigger than any other gastropod species known then, or now. However, this identification incited much controversy.
For whereas spiralled gastropod shells normally bear ridges and possess coils that taper to a point, Dinocochlea did not, and there were no shell traces preserved with it either. Its freakishly large size was also difficult to reconcile with a gastropod identity.
Recalling the devil's corkscrews of North America, was it possible, therefore, that Dinocochlea was actually the fossilised burrow of some still-undiscovered species of prehistoric rodent? Alternatively, bearing in mind that it was uncovered near to a quarry famous for the quantity of Iguanodon and other giant reptilian fossils discovered there, could it be a dinosaur coprolite (fossilised faecal deposit)? Once again, however, its gargantuan size (even for a coprolite of dinosaur origin!) and also its spiralled shape's very precise, regular form argued against this, as did the fact that there was no partially-digested organic material associated with it, which is normally the case with preserved coprolites. So what could this very curious, anomalous object be?
In June 2011, palaeontologist Dr Paul Taylor from London's Natural History Museum (where Dinocochlea had been deposited following its discovery) officially presented a new and very plausible explanation.
In a paper published by the Proceedings of the Geologists' Association, he proposed that it had indeed originated as a corkscrew-shaped burrow, but a horizontal one rather than the vertically-oriented devil's corkscrews, and had not been created by any rodent but instead by a fossil species of capitellid polychaete worm known as a threadworm. Yet as these were only a few millimetres in diameter, how could so tiny a creature have produced such a monstrously huge trace fossil as Dinocochlea?
Having examined cross-section specimens of it, which revealed that they were filled with concentric bands of sediment resembling the growth rings of tree trunks, Dr Taylor suggested that although initially very small, this worm burrow had acted as a nucleus for concretion growth (which is characterised by the presence of such rings or bands internally).
That is, the space originally created by the burrow would induce the movement into it of surrounding mineral cements, which would themselves then leave behind a space that would in turn induce the movement into it of more surrounding cements, and so on, until eventually, if conditions for its preservation were just right, what began as a tiny thin worm burrow would ultimately become enormously enlarged, yielding the very dramatic pseudo-gastropod, mega-burrow trace fossil that we know today as Dinocochlea.
From horned marmots and burrow-digging beavers to devil's corkscrews and terrible snails-that-weren't, it is evident that however distant our planet's past may be, it still possesses the power to perplex, surprise, inform, and fascinate us in a myriad of different ways.
The very attractive front cover of the How and Why Wonder Book of Prehistoric Mammals (© John Hull/Transworld)