The Great Fossil Enigma

by Simon J. Knell

  Sweet would never claim to be a palaeobiologist, even if his work on statistical assemblages paved the way for the science to recover its biology. However, no author of a book on conodonts could ignore the animal. It was the hot topic of the moment and he could not deny the exciting new evidence provided by the Scottish animals, but he said that on many key points – points on which Briggs and company had been only too happy to draw conclusions – these objects remained “mute.” Sweet willingly conceded to tentative conclusions that the fossils might preserve segmentation akin to the muscle blocks (myotomes) found in amphioxus and fish. What he found objectionable was the leap of interpretation that turned them into somites, with all that that might mean for possible vertebrate affinity. He thought it a conclusion “in no way required by the evidence.” This leap, he suggested, led into a “much more subjective discussion of anatomy.” The British workers had located a “head,” but where was the brain? There was no evidence for advanced head structures, and Sweet preferred a less suggestive terminology. He simply could not understand why Aldridge, Briggs, and others had seen a chordate animal.

  Sweet suggested that “chordates are not the only organisms that might yield such V-shaped impressions on compression.” Could these impressions not belong to a flatworm (nemertine)? The modern nemer-tine had many of the characteristics preserved in the Scottish specimens. Sweet was not, of course, the first to suggest this relationship. And why had the British workers interpreted the lines running down the center of the body as they had? There were other possibilities here too. He felt that “Aldridge et al. allowed their interpretation of the Scottish specimens as primitive chordates to restrict their survey of anatomic possibilities” and permitted them to preferentially see “a nerve chord, notochord or dorsal aorta.” Might the preserved “lines” at the front of the animal represent the nemertine's eversible proboscis, he wondered, while those preserved at the rear perhaps record the gut? If so, then that gut reached the tail of the animal. This was an invertebrate feature as it exits the body before the tail in vertebrates.

  Sweet also took issue with Jefferies’ mysterious “s”: “marine habit, eel-shaped body, head-trunk-tail, rasping teeth, ?lensless paired eyes, ?somites.” How many other animals offered this model? “I suggest that, as Jefferies intimates, the resemblance between the hypothetical animal ‘s’ and the Scottish conodonts maybe so striking because the latter were used as a general model for the former.”

  Sweet shared the view that the teeth possessed by hagfish merely resemble proper teeth; they bear no evolutionary relationship. They could not assist in giving the conodont chordate affinity, nor could the phosphate, the element shape, basal material, chevron impressions, or models of element function. All were unreliable. The proof, he said, was to find the notochord, gill slits, or dorsal nerve cord.

  Sweet's outlook was framed by Wonderful Life spectacles. He had had a chance to discuss conodonts with Stephen Jay Gould on two occasions when lecturing at Harvard. The thrust of Sweet's book, which promoted the idea that conodonts were a distinctive form of life, arose from these conversations. To him, the attempts of Briggs and friends to pigeonhole the conodont in the chordates was simply illogical. He believed paleontology was moving in the opposite direction, away from blinkered classifications and presentist attitudes: “It has several times been suggested that even the venerable Chordata is possibly no more than a shaky confederation of invertebrates and vertebrates.”

  The key message in Sweet's arguments was not, however, about the particulars of the animal's biology but the boundaries and frailties of interpretation. In this respect his arguments read like Richard Owen's chastisement of Pander. Owen simply could not comprehend, given all the evidence to the contrary, why the Russian had thought the animal a vertebrate.

  We don't know what Pander thought of Owen's opinion, but Aldridge and Briggs were certainly not the kind to take these criticisms lying down. They had been accused of producing mere “waffle.”27 When Sweet saw that Aldridge and Briggs had together written a six-page review of his book, he must have felt a little concerned. Book reviews are almost without exception written by individuals, not duos. Aldridge and Briggs were, however, gentle. Aldridge certainly knew, and had the greatest respect for, Sweet. And he was probably not that surprised that Sweet had produced a book that was as “provocative” as it was “controversial.” It was, they said, food for thought.28 But by the third page, their sleeves were rolled up: “Having been involved in presenting these candidates [the Scottish animals] for consideration and in using them to develop a hypothesis of chordate affinity…we have a particular interest in how they are treated in this milestone in the conodont literature.”

  Their assault began by countering Sweet's Wonderful Life view of the distant past. Now became clear the significance of Briggs's involvement in the review. It was Briggs, with Conway Morris, who had populated Gould's world with a myriad new life forms, which in turn shaped Sweet's outlook. But the tide was on the turn; Briggs admitted that some of these problematic animals were indeed finding a place in long-established phyla. This gave validity to his and Aldridge's attempts to understand the conodont animal's chordate affinity and suggested that one should not buy into Gould's vision unquestioningly.

  Now the five newest animals entered the debate for the first time, for they confirmed the “V-shaped” structures that the two men were convinced were unlikely to be found in preserved nemertines. They were very like chordate muscle blocks seen in amphioxus and hagfish. Sweet had, of course, made his arguments without possessing the animals – old or new – and thus it was rather easy for Aldridge and Briggs to undermine his interpretations; the fossils simply said otherwise, they said: “The characters of the conodont animals, such as phosphatised elements, a transversely operating feeding apparatus, V-shaped somites, a laterally flattened trunk, and ray-supported posterior fins extending further along one margin than another, all bear comparison to chordates.” The new specimens enhanced this view. The problem, they felt, was that Sweet had become wedded to the view that the conodonts represented a separate group and thus he looked for differences rather than similarities with other forms of life. Of course, only a few years earlier, confronted with the ambiguities of the first fossil, this had also been their refuge. In just a few years, however, the science had fundamentally changed and, they believed, it was no longer appropriate to think these old thoughts.

  In a parallel review, Paul Smith saw Sweet playing the role of devil's advocate: “One does not have to be an orthodox Popperian to conclude that Sweet is here being rather mischievous and is not advancing testable hypotheses.”29 He wished Sweet had maintained a “more objective viewpoint,” but that was Sweet's point: There was too little that was objective about the British animal and too much that exposed the frailties of interpretation.

  Aldridge and Briggs recognized that Sweet's forthright approach had produced a useful and stimulating book with which a new generation of scientist could argue. They had, in effect, begun those arguments in this review. Those arguments would, however, soon fade from view while the book would live on. It would remain on library shelves for decades, just as Lindström's had, there to suggest to outsiders that this was the way things stood.

  At the close of the 1980s, Aldridge was in Leicester and Briggs in Bristol. The decade ended with the final dispatching of the first and second contenders for conodont animal. Melton and Scott's animal was redescribed by Conway Morris in a paper that ironically, given the delayed publication of original paper, was lost in press for six years. Possessing considerably more material, Conway Morris was gentle on these earlier authors and celebrated the animal in its own right for its peculiar ecological interest and zoological strangeness. Now Conway Morris could reinvent it as the animal that fed on conodonts. The canard finally had scientific recognition. The first animal was dead.30 Some, though, looked at the rock in which these fossils had been found and wondered how a conodont-eating animal could exist where there are
, inexplicably, no signs of conodont life.

  Now it was Conway Morris's turn to suffer the ignominy of error. He had admitted his mistake not long after the Scottish animal had been published, but this did nothing to protect him from a little ridicule. As mistakes go, Gould remarked, Conway Morris had “made a beauty.”31 Through scientific eyes the path of progress is littered with mistakes, though Gould thought them “not badges of dishonor.” Conway Morris's interpretation of Odontogriphus had produced an animal of its time; had he come across it in 1989, he would have read it entirely differently. It was simply a price to be paid. But Conway Morris would, more than most, recognize how long-lived books like Gould's Wonderful Life could infect a scientific career. In the ephemeral press of scientific publishing, disproven interpretations are soon forgotten, but Wonderful Life kept Conway Morris's past alive, portraying him to readers as the author of outmoded ideas long after he had given them up. When Conway Morris later came to write his own popular book, The Crucible of Creation, its reviewer, Richard Fortey, detected the author's loathing of Gould.32 He accused Conway Morris of selective amnesia and the rewriting of his own history but reflected that Gould's book perpetuated old science and retained Conway Morris as its keeper. One could understand why so many scientists doubted the validity of books. As vehicles by which progress is made, they have the unfortunate effect of producing a freeze-frame image that is then left to drift into the future, giving the illusion of still being current. The scientific book is an impossibility; almost immediately it is a history book.

  Eight years after the first true conodont animal had been found, the questions it posed had not been resolved. The animal remained locked in argument. Sweet seemed to imply that Aldridge, Briggs, Smith, and Clarkson had aspirations for their animal, that they had the vertebrates in their sights. Aldridge and company, however, felt that their ideas were simply developing as new data were revealed. Indeed, Clark's constant supply of animals had progressed many of those initially tentative arguments to a point of certainty. When, in 1989, Conway Morris came to review the progress that had been made, he felt sure that the animal was a chordate and that it held tremendous potential for understanding the origins of vertebrates.33 He also believed that more and better fossils would be found. How wonderfully ironic it would be if, after such a circuitous journey, the animal Pander had dared to imagine became a reality. Of course, there never really had been a circular journey through a hundred possible identities – that was just a myth. Those who knew the conodont – really knew the conodont – knew it had never really left the place where Pander had originally put it, that never-never land where vertebrate meets worm, and where hagfish, amphioxus, and a host of other chordates swim. The conodont animal's natural home was that most difficult and enigmatic of all palaeontological places, that place where vertebrates begin in time and in space. And this was where conodont science was now heading.

  “Over the Mountains

  Of the Moon,

  Down the Valley of the Shadow,

  Ride, boldly ride,”

  The shade replied –

  “If you seek for Eldorado!”

  EDGAR ALLAN POE,

  “Eldorado” (1849)

  FOURTEEN

  Over the Mountains of the Moon

  IN THE MID-1980S, OUT OF SIGHT OF THOSE DEBATING THE meaning of the first Scottish animals, the next big step was being taken in a part of the world that had thus far proven itself completely lacking in these extraordinary fossils: South Africa. Here, along a dirt road in the Cedarberg Mountains, some two hundred kilometers north of Cape Town, Geological Survey officers Danie Barnardo, Jan Bredell, and Hannes Theron came across a new borrow pit for road metal exposing the soft and rarely seen Upper Ordovician Soom Shale.1 They stopped to investigate and found their curiosity rewarded with some intriguing fossils reminiscent of graptolites. Graptolites are one of those classic groups of extinct animals all paleontologists study at some point in their training. Tiny, colonial – bearing a passing resemblance to corals and bryozoans – their fossil remains are most common in shales, where they look like minute flattened saw blades. Theron sent a specimen to Barrie Rickards at Cambridge University in the UK, an expert on this group, to see if they really were graptolites. Rickards said they were not.

  A number of scientists at the Survey headquarters near Pretoria, including palaeobotanist Eva Kovács-Endrödy, however, became intrigued by the similarity of these new fossils to strange spiny plants found in much younger Devonian rocks. If this was what they were, then these were clearly important finds. In an echo of Pander's conodont discovery, they would push back the origins of plants by forty million years. The oldest plants known at that time came from the late Silurian. Theron and Kovács-Endrödy prepared a paper naming these new plants Promissum pulchrum, meaning “beautiful promise.”2 They did not know that rather than being harbingers of a green and pleasant Eden, these new fossils held a beautiful promise of a rather different kind.

  As is the normal course with scientific publication, the paper was sent out for independent external expert opinion. The task of reviewers is to consider a paper's merits and to advise the editors on whether it should be published. On this occasion, one reviewer believed the authors had not demonstrated that these fossils really were the remains of plants but left the decision on publication to the editor. The paper was finally published alongside a reply by the critical referee and a response from the authors.

  This discussion mentioned Rickards's view that the fossils were not graptolites, but Rickards was concerned that he had been quoted when he had only seen one specimen. So more material was sent to him by diplomatic pouch and he confirmed his conclusion that they were not graptolites. He thought they looked rather jaw-like and discussed them with several of his colleagues. Together they alighted on the idea that these fossils just might be conodonts. The main problem with this identification was their size; they were more than ten times bigger than the conodonts they were used to seeing. Rather than being up to two millimeters long (though usually much smaller), these were up to two centimeters long!

  Rickards knew these fossils crossed the boundary of his expertise, so he contacted Dick Aldridge. Aldridge was intrigued and traveled to Cambridge the next day. It was to be another important day, for here he saw the first conodont specimens known from Africa south of the Sahara and the first complete apparatuses from the Ordovician. And if these were not in themselves major milestones, he could also confirm that they belonged to a giant animal. Preserved merely as molds, Aldridge empathized with the Survey officers who had struggled to identify them. They were, in every sense, conodonts like no others and totally unexpected.

  Seeking further confirmation, in August 1987 Theron took some specimens to the Devonian Symposium held in Calgary, where he showed them to a number delegates with expertise in fossil plants and conodonts. They were overwhelming of the view that these fossils were conodonts, albeit of amazing size. Nevertheless, back in Pretoria, Kovács-Endrödy and others remained wedded to the plant and insisted that they continue with the publication of a more extended paper advancing this idea. Theron was, however, now convinced the fossils were conodonts and withdrew as a coauthor, later joining Rickards and Aldridge in publishing a paper identifying Promissum as a conodont.

  African conodonts, apparatuses, giants – there were many reasons for the South African geologists to welcome the news. They may have lost their landmark discovery of plants, but they had gained something quite extraordinary. Communication lines now opened between Aldridge and the Survey workers, and in 1990 he found himself traveling to South Africa to study their collections of Soom Shale fossils. At the time, all he knew was that there was a wonderful opportunity to progress science's understanding of the animal in this part of the world. Ever since the animal had been found, Briggs and Aldridge, in particular, had been searching for new resources and new ways of looking. Briggs had searched Lagerstätte collections in the United States and now Aldridge did the same in t
he Soom Shale collections in Pretoria. He also collected some material himself and brought it back to Leicester. It was when he was examining these new finds under his binocular microscope – playing with the lighting to get the best possible illumination – that his eye came across two indistinct oval impressions. Were these those strange, dark-lobed structures he had seen in the Scottish animals? He felt “a buzz of excitement.” Had he chanced upon yet more animals and perhaps – given the relative ease with which apparatuses had been collected – the richest deposit yet? He needed to get back to South Africa – and urgently.

  In 1991, with a small research grant in his back pocket, Aldridge returned South Africa to search for the animal with Theron and others. Theron had arranged for the use of a mechanical excavator which had cut a trench into the exposure at the original find site. This produced a heap of rock ready to be split, and on the second day of collecting, August Pedro was holding a complete apparatus in his hands. It had associated with it two ring-like structures. It seems that Pedro had a Neil Clark–like talent for finding these fossils, and soon more followed. The Soom Shale was that day understood anew. It too was a Konservat-Lagerstätte, and of an age hardly represented by such deposits. A new window had been opened into the deep past. It had been an important day.

  With each new conodont fossil found, Aldridge gained an increasing sense of the three-dimensional structure of the hollow rings: They were deep and inwardly tapering. They did, indeed, seem to be the remains of the animal's eyes. Almost identical circular objects had been seen in the Silurian relative of the lamprey Jamoytius. An important animal, in the 1940s Jamoytius had been regarded as “undoubtedly the most primitive of the ‘vertebrate’ series of which we have knowledge.” In the 1960s, Alexander Ritchie from the University of Sheffield visited the Scottish site where the original fossils had been found and gathered more and better specimens.3 He interpreted the distinctive circular rings in these fossils as sclerotic cartilages surrounding the eyeball. These were so like those now found in these new conodont animals that Aldridge and Theron thought this a reasonable interpretation for them.