The Great Fossil Enigma Read online

  William Bryant, Devonian fish specialist and director of the Buffalo Museum of Science, did not think as Ulrich did, but he too valued the conodont's utilitarian possibilities. A supporter of Kindle, Bryant washed and boiled lumps of George Jennings Hinde's “Conodont Bed” from Eighteen-mile Creek and discovered that conodonts made up 50 percent of the mass left behind.14 He found the fossils easy to identify once freed from the surrounding rock and could show that Hinde, who had examined his fossils on the surfaces of the rock, had made a number of misidentifications. He also thought Hinde's anatomically complex Polygnathus dubius another mistake. Both he and Roundy believed the association was merely coprolitic (fossilized feces) or “the ejecta of some fish.” More than anything they wanted the conodont fossils to be effective “horizon markers,” which meant that each kind needed to be narrowly defined. In pursuit of this goal, Bryant disassembled Hinde's Polygnathus, restricting that name to the distinctive plate-like forms and renaming all the other fossils.

  Ulrich and Bassler were also pushing ahead with these “difficult toothlike organisms,” hoping to locate a definitive answer to the black shales problem that supported their view. By the early 1920s, Bassler had began to nurture an interest in conodonts in Ray Hibbard, an optician, World War I veteran, and fossil enthusiast living in Buffalo. Over the coming decades, Hibbard would maintain regular correspondence with Bassler and contribute significant collections to the National Museum long after Bassler's own interest in these fossils had waned.15 In time, Hibbard would develop his own private museum and library, swap collections with distinguished foreign paleontologists, experiment with acids and photography, and even install a microfossil washer machine. Hibbard was that vital collecting cog that had long powered the geological engine. And for those who drew upon his services, he had the added advantage of possessing no great ambition to publish. He was simply a servant to others.

  Ulrich and Bassler now drew upon collections belonging to Hibbard and to the National Museum – probably the best collections of conodonts in the country – to work on “a reasonable and, it is hoped, natural classification.” Bassler presented the preliminary results of this work at a meeting of the Paleontological Society in Ithaca, New York, in December 1924. Two years later, their jointly authored paper appeared.16 In it, and with relatively little fanfare, the two men claimed that “conodonts” – the fossils – were the remains of fish and that each type of conodont belonged to a distinctive species. The simple pointed ones belonged to fish species related to hagfish and their kind, the more complex arched and pick-shaped ones possibly belonged to relatives of the sharks and rays, while those they called “dermal plates” may have belonged to a distinct group but were not to be considered true conodonts. The two men also rejected Hinde's complex Polygnathus, claiming it was contradicted by fish biology. But in that subject they had no real expertise. The arguments they made for their simple fishes were superficial. Ulrich now admitted that his earlier preference for a worm had been a mistake – he simply had not then seen “true conodonts,” and when he did, he soon changed his mind.

  Armed with a comprehensive history and bibliography of the subject, which had been prepared by the museum's Grace Holmes, Ulrich and Bassler grouped the conodont fossils into four families according to their shape: simple conical teeth, pick-shaped teeth, arched bars, and plates.17 In doing so, they increased the fossils’ usability by making them easily recognizable. They then sought to demonstrate that conodonts could resolve the black shales dispute. They said it was possible to prove that none of the conodonts recorded in the Devonian Genesee and Portage rocks of New York state could be found in the Ohio black shale, the upper New Albany in Kentucky, or the Chattanooga in Tennessee and Alabama. These Genesee and Portage fossils could, however, be found in the rocks below these horizons. They concluded: “In so far then as the evidence of conodonts is concerned, the post-Devonian age of the Chattanooga and Ohio shales, as long advocated by the senior author, seems conclusively established.” But how could they have concluded anything else?

  It was this apparently unequivocal solution to the black shales problem that thrust Ulrich and Bassler's paper into the limelight. Its readers now took the conodont seriously. And by simplifying the animal so that each kind of fossil could be said to represent a single species, the conodont became as easy to use in stratigraphy as single-celled foraminifera.

  By the time of Ulrich's retirement, in 1932, the paper had been widely publicized in the United States and abroad, and their conodont classification was said to be in “extensive use.” By then, microfossils were so fashionable that Ulrich's brief excursion into conodonts was considered a career high point. It had been a groundbreaking piece of work that Bassler thought had performed a little revolution.18 No one saw it as an attempt to take a resource Kindle had identified and turn it against him. But if Ulrich really had manipulated the fossils to support his outspoken views, or simply been blind to objective reason, surely the conodont would in time expose him, for those who now examined the black shales were as attentive to the conodont as they were to the words of Ulrich.

  By 1932, Clinton Stauffer was a seasoned professor at the University of Minnesota. A specialist in Devonian stratigraphy, with experience of the contentious black shales in Ohio and Ontario, he had worked for the Canadian Geological Survey and was another of Kindle's associates, agreeing with him on the matter of the black shales. By happy accident, Stauffer was now to become a national expert in conodonts as a result of taking an interest in building work on the campus. To his delight he discovered a thin lens of sediment in the Ordovician Decorah Shale exposed in the excavations that contained finely preserved conodonts. Recognizing the novelty of this find, he described and published the conodont fauna knowing that all his species were entirely new to science. No one had previously looked for conodonts in this rock. He imagined Ulrich and Bassler's fishes “may have migrated over a much wider area than that covered by the Decorah shale” and were likely to be found in other rocks of similar age that might then be correlated with the Decorah. Sure enough, a short while later, Fanny C. Edson of the Gypsy Oil Company in Tulsa, Oklahoma, got in contact with him having found conodonts in well cuttings that penetrated the Decorah Shale in distant Reno County, Kansas. Edson, who was one of many women involved in oilfield geology at the time, needed to be sure that her local Decorah really was the same as that named after the town in northern Iowa. Stauffer told her that the conodonts proved it. News of Stauffer's newfound expertise spread and soon he found himself collaborating with foraminifera worker Helen Plummer, one of the instigators of the Society of Economic Paleontologists and Mineralogists and half of a noted paleontological marriage. Together, Stauffer and Plummer pioneered an investigation of the conodonts of a wholly different group of rocks belonging to the Pennsylvanian. By chance, then, Stauffer had stumbled into a paleontological lacuna and at once found himself the expert. As a result, and seemingly with no particular plan in mind, he then began to populate the American landscape with these mysterious fishes. At each location, he put in place a time marker to which others could correlate their own local rocks using conodonts.19

  Another in this advance guard taking up the conodont was Frank Gunnell, of the University of Missouri. He began by undertaking reconnaissance collecting throughout Missouri, Kansas, and Oklahoma, finding abundant conodonts in the Devonian, Mississippian, Pennsylvanian, and Permian.20 He discovered that they occurred mainly in shales and decreased in abundance in sandstones, conglomerates, and limestones. This idea – that conodonts mainly occurred in shales – soon became embedded, and it was on these rocks that conodont-seeking geologists would concentrate their efforts.

  Gunnell realized that the conodonts’ size and occurrence in otherwise “unfossiliferous” strata had made them invisible. It now made them peculiarly useful. Indeed, it seemed quite magical that these fossils should come to light in rocks that so desperately needed them. Like Stauffer, he too thought Ulrich and Bassler's notion o
f free-swimming fish added greatly to the fossils’ utilitarian potential. That potential was heightened still further by Gunnell's belief that many species had a very limited stratigraphic range. It meant that a single species might signify a narrow period of geological time yet be found across wide geographical areas. Buoyed up by a new optimism, which came from finding these fossils in unexpected places, he felt that lithology and depositional environment had little impact on their distribution. If this was the case, then the conodont was the ideal utilitarian fossil, provided it could be found in sufficient numbers. Gunnell now turned his attention to the conodonts of the Pennsylvanian, wishing to document the stratigraphic range of every species he found. Noting and valuing every ridge and node as he identified them, he soon possessed more than one hundred new forms.21

  A third pioneer was Chalmer Cooper, a West Virginian who had returned from fighting in World War I to complete a degree in geology and engineering at the University of Oklahoma. It was in this state that he found his first job in the science, as chief geologist at the Oklahoma Geological Survey. Inspired by Ulrich and Bassler's assurances of a distinction between the conodont fossils of the Devonian and Mississippian, he began his studies by gathering reference material from Ohio in the late 1920s, then used this to demonstrate the Mississippian age of several local rocks. As a disciple of Ulrich and Bassler, though unconvinced by their fish, Cooper soon found himself embattled with Roundy, Girty, and others at the National Museum.22

  Fired up by Gunnell's discovery of conodonts in the neighborhood of Columbia, his colleagues, Ted Branson and “Doc” Mehl, hatched a plan to reveal the stratigraphic utility of the conodont completely. They were to do what Gunnell had began to do but on a grand scale. Both were former students of Samuel Wendell Williston, who himself was a former assistant of dinosaur addict Marsh. Williston had risen to a position of distinction as a vertebrate paleontologist, and Branson had followed him from Kansas to Chicago in order to continue his studies. The first to arrive in Missouri, Branson had orchestrated a renaissance in the Geology Department's fortunes. Ambitious, and still known as a fish paleontologist, the conodont fish must have come to him as something of a revelation. No other fish fossils held this kind of stratigraphic potential. Branson's collaborator, Maurice Mehl, had as a young man teaching in Oklahoma been active in the founding of the AAPG. They made a good team. Branson was a man of ambition and strategy, while Mehl was regarded as an exemplary scientist.23

  Their extraordinary conodont odyssey started in 1930, when the two men, with the help of their students, began an assault on the Missouri strata. They published as they went, mapping a geological landscape strewn with the teeth of thousands of corpses and fundamentally altering the animal and its utility. “We discover new conodont localities and horizons nearly every week,” they reported in 1933.24 Their data began to accumulate so rapidly that they, rather remarkably, acquired permission to fill all four issues of the university's research journal, University of Missouri Studies, with conodonts. At 349 pages in length and published between June 1933 and October 1934 under the title Conodont Studies, this was the first book-length treatment of these fossils.

  Like Ulrich and Bassler, Branson and Mehl believed conodonts were the teeth of more than one group of primitive fish. It was the finding of bony material attached to some conodonts around this time that convinced them that this was the case, not the weak arguments that had come from the National Museum. They also differed from the Washington men on one other important point: They believed “the teeth in some species were arranged shark-like with two or more kinds in the same mouth.” But, ever the utilitarians, they recognized that the “likelihood of finding the teeth in their original associations is remote.” They used this to justify giving each kind of tooth its own species name, and in doing so they included the leaf-like forms Ulrich and Bassler had considered dermal plates. These were now thought to be “crushing teeth” and were to become particularly important. While praising the “epoch-making contributions” of Ulrich and Bassler, they thought the Washington classification premature and refrained from using it. Branson and Mehl were their own men and were determined to forge their own way. They saw conodont studies as a blank slate and made it their mission to fill it. They wished to create a new research field and place themselves at the head of it.

  They realized that this would involve both evangelism and education and so began Conodont Studies with a “how to do it” guide. It gave assurances to those who doubted they would be able to find what were still largely unknown fossils. Branson and Mehl encouraged these novices to exploit plastic and sandy clays, suggesting that they were often more productive than many shales. They also implied that limestones were often devoid of conodonts, so making their investigation unattractive. Instead, readers were advised to adopt mass-processing techniques that involved boiling samples of sediment, sieving the results, and then separating fossils from the remaining sediment on the basis of their differing densities using heavy liquids, before finally picking over the resulting residue to extract the fossils. This, they claimed, was far more reliable than picking individual conodonts off bedding planes by hand. “Starting with a meagre knowledge of general micropaleontological technique” and unable to draw upon the technical expertise of “industrial organizations” (oil companies), who saw such information as commercially sensitive, they pulled together an arsenal of methods from their own innovations and borrowings. Among these was the innovative use of stereoscopic photography to study the fossils. They even went so far as to publish stereoscopic images, thus enabling their readers to see the fossils in 3D and better understand their complexity.

  Branson and Mehl's optimism for these new fossils was, however, overshadowed by one great fear: contamination. They worried that their samples might be corrupted by splashes from boiling pots and residues left on sieves. But with care they knew they could avoid these problems. They were rather more concerned about natural contamination resulting from older rocks and fossils being eroded and redeposited as younger rocks, and so mixing together conodonts belonging to different periods. They also imagined younger deposits containing conodonts penetrating cavities in older rocks, much as Moore's Triassic mammals had fallen into fissures in the older Carboniferous Limestone. What they did not consider was that their minds could also become contaminated as a result of this obsession. It affected their ability to see. An unexpected conodont was, for them, always an error and never simply a rarity. But perhaps they did not have the luxury to think otherwise in a science so new.

  Branson and Mehl's “how to do it” guide was followed by papers discussing rocks ranging in age from the Lower Ordovician to the Lower Mississippian. In these, the two men demonstrated how conodonts could be used to solve existing stratigraphic problems. The Ordovician Harding Sandstone, for example, was problematic because it held Ordovician invertebrate fossils and fish fossils that looked rather Devonian. The conodonts alone proved that this sandstone could be correlated with the Middle Ordovician Joachim of Missouri. The fish were merely deceptive. The conodonts were, in contrast, quite distinctive: dull and amber colored, “a very primitive group” that possessed what is “best described as ‘fibrous’ structure” that splits lengthwise. They seemed to be fused to parts of the animal's jaw.

  In the Lower Ordovician Jefferson Formation of Missouri, they found their oldest conodonts. They were simple, diverse, and distinctive, and they offered the possibility of correlating this rock for the first time. An examination of the Silurian Bainbridge Formation of Missouri produced rather different results. Here they were pushed to find any conodonts at all, but when they did they were a predictable mix of “typical Ordovician genera” and “obvious forerunners of typical Devonian and Mississippian genera.” Although these rocks were positioned between the Ordovician and Devonian, they could not connect these two groups of conodont fossils. Indeed, their difficulties in finding conodonts suggested to them that the Silurian was a low point in their history.


  As they worked their way through the stratigraphic column, a more comprehensive picture of changing conodont faunas began to emerge in their minds and it became possible to guess the relative age of the strata on this basis. Locating rocks of equivalent age was not, however, as easy as might be imagined. The geological map of America, which can so easily be painted in large blocks of uniform color, conceals a complex, three-dimensional patchwork of rocks. It was from the pieces of this jigsaw that Branson and Mehl extracted their fossils and thus began to solve the puzzle of their age relationships. It was to their advantage that they treated their subject abstractly, as by removing their samples to the laboratory, they also removed themselves from the problems and complexities of the field. It simply became a matter of matching fossils at different sites. This also helped them to imagine the conodont fossils as an evolving continuum of forms and to guess the form of those conodonts that occupied gaps in their knowledge. They became connoisseurs of their fossils, and this gave them an ability to predict.

  Before long, Branson and Mehl's climb up the stratigraphic ladder led them to consider those conodonts that marked the division between the Devonian and Mississippian – those contentious rocks Ulrich and Bassler had claimed to have tamed using conodonts. To aid their interpretations, Branson visited Hinde's type specimens at the British Museum in London. With a lack of empathy for a science still in its early days, Branson was rather dismissive of Hinde's achievement: “He spent twenty days in making the study and it was far from thorough in many respects.” Hinde's material was embedded in matrix, and for this reason Branson considered it imperfect, but it had been supplemented by specimens from Bryant. Branson also had some five hundred specimens of his own from Eighteen-mile Creek, which his son Carl had collected. Armed with this material, Branson and Mehl now entered this contentious terrain.