The Great Fossil Enigma Read online

  Their study, and the whole of the third issue of Conodont Studies, published in June 1934, concerned the Grassy Creek Shale of Missouri. Branson and Mehl's conodonts correlated it with the Chattanooga, the low Huron of Ohio, the Portage, the Hardin, and the Woodford; across them all there was a fairly consistent and distinctive fauna very much like that Bryant had described for the Genesee, which they took to mean that all these rocks were Upper Devonian. Ulrich and Bassler had claimed the Chattanooga and Hardin as undoubted Lower Mississippian precisely because they had not found Genesee or Portage conodonts. Branson and Mehl, however, drew upon their superior connoisseurship to claim that the faunas found in these different rocks were more closely related to each other than to those of undoubted Middle Devonian or Mississippian rocks. Indeed, they felt that distinctive faunas marked the Devonian-Mississippian boundary and these were not those claimed by Ulrich and Bassler. They also pointed out that Ulrich had foolishly dismissed the evidence of Devonian fish fossils. Ulrich's blinkers were now fully exposed, his godlike command of nature a matter of self-delusion.

  Branson and Mehl's new faunal criteria boiled down to just six seemingly diagnostic groups of conodonts: The “Polygnathus-like genera, Polylophodonta, Ancyrodella, Ancyrognathus, and Palmatolepis” were restricted to the Upper Devonian and were only ever found in the Mississippian as contamination. They were replaced in these latter rocks by Pseudopolygnathus and Siphonognathus. All of these kinds of conodont would become known as “platforms.” Ironically, Ulrich and Bassler had dismissed them as being “dermal plates,” not conodonts. Now they were the most important conodonts of them all.

  Branson and Mehl were in awe of the Grassy Creek conodonts, which were “in a highly plastic stage of their development.” The distinctive forms of major groups seemed to merge seamlessly one into another. Some genera, such as Polygnathus, were so variable that species could not be distinguished. Small forms could not be related to larger ones. The sheer variety and blending of form was perplexing and yet marvelous. And alongside these remarkable forms were other conodonts that seemed to belong to an earlier era. Branson and Mehl thought they must be contamination and they established a principle to guard against it: “It is the new elements that determine the age.”25 Old ones, they presumed, may result from erosion and redeposition.

  As the final issue of Conodont Studies was packaged up for the presses, Branson and Mehl must have realized the enormity of what they had achieved. They, rather than Ulrich and Bassler, had laid the bedrock for a new science, and across the American landscape they had left markers others could now use to locate the age of their own rocks using conodonts. Ulrich and Bassler may have asserted the utility of these new fossils, but in doing so they merely echoed Kindle's call. The proof of the method had not been theirs; they had merely stated what they thought they knew. The proof belonged to Branson and Mehl, who now said it was only necessary to identify a few kinds of conodonts in order to resolve the problem of the black shales. As a consequence of this work, in 1935, Stauffer felt it was possible to say that “Conodonts in Paleozoic sediments assume much of the importance of the Foraminifera in later sediments.”26

  But not everyone was enamored with the Missouri conodont factory. Ulrich and Bassler's disciple, Chalmer Cooper, must have felt his own work had been devalued by Branson and Mehl's proof that the Washington men had been wrong. In a review of Conodont Studies, Cooper found every possible excuse to complain: about their preference for mass processing, their re-identifications, their use of broken specimens so “valueless to other workers,” and their splitting of fossils into narrowly defined species and genera “to a degree that makes it difficult to recognize them.”27

  At Indiana University in Bloomington, the young John Huddle also grew concerned about the new Missouri orthodoxy. While Branson and Mehl were wrestling with the Grassy Creek, Huddle had been completing his doctoral thesis on the New Albany Shale of Indiana.28 Like the Grassy Creek, this was a disputed sequence. Most thought it entirely Devonian, but in 1911, Ulrich had suggested that it might be part Mississippian. Bassler had then supported this view.

  Huddle had entered this disputed territory quite by accident, having been shown some Devonian brachiopods from near the top of the shale. This seemed to challenge Ulrich and Bassler's claim. It was while Huddle was investigating the source of these fossils that he came across some conodonts that he knew from Ulrich and Bassler's recent work might resolve matters. He began his fieldwork in the summer of 1932 and was soon corresponding with Bryant and Cooper. This “brownish to black in color, more or less massive, hard, brittle” shale becomes fissile when weathered and is easily split, giving it its local name of “black slate.” But Huddle found it resisted bulk processing. It was a technique Huddle distrusted anyway, having found that it destroyed the most delicate species. Instead, he painstakingly extracted or revealed his fossils using needles, in the process producing one of the most thorough investigations of those early years. Nevertheless, Huddle felt disappointed with his results. Of the 158 species he described, only 37 had been found previously. Only they would offer any chance of correlating these rocks with others, thus proving their age. He had no contact with Branson and Mehl and knew nothing of their Grassy Creek work. Huddle tentatively concluded that the lower part of the shale seemed to correlate with the Devonian Genesee. Those conodonts from the middle were most like those of the Hardin and Chattanooga but were also associated with a few Portage and Genesee species. These were found well below the Devonian brachiopods and proved that Ulrich and Bassler's supposedly distinctive Chattanooga species were not Mississippian. Conodonts from the upper part of the rock included the most new species, and while he found Chattanooga and Hardin species among them he could not say that the Mississippian had been reached. He did, however, find a shark's tooth, which Bryant suggested might indicate a possible Mississippian age. Huddle had independently and simultaneously reached the same conclusions as Branson and Mehl: Ulrich and Bassler were wrong. Cooper, who was now at the Walker Museum in Chicago and working on his own doctorate degree, remained oblivious to these new discoveries and continued to work on the Oklahoma rocks in the belief that Ulrich and Bassler were right.29

  Branson and Mehl soon consolidated the position of their “index genera.”30 The “index Devonian genera” were now Icriodus, Palmatolepis, Ancyrognathus, Polylophodonta and Ancyrodella. The “index Mississippian genera” were Pseudopolygnathus, Siphonognathus, Pinacodus and Solenognathus. Such clear presence-absence criteria were a stratigraphic ideal, and among them there was one particularly widespread Devonian form, Icriodus, the various species of which also permitted the division of the Middle and Upper Devonian. There was nothing like it in the Mississippian and it appeared to have no descendants. There was no ambiguity about this fossil, and to prove their point they revisited some disputed and imperfectly understood black shales, including those studied by Cooper and Huddle (they did not mention Cooper by name, perhaps wishing to avoid a personal dispute). They confirmed that the upper beds of the New Albany Shale were indeed Mississippian and moved all Cooper's Mississippian rocks into the Devonian. Now it seemed they were able to resolve the black shales problem on the basis of a single type of conodont.

  Remarkably, this advanced state of knowledge was achieved in 1938. But by then even Branson and Mehl knew that progress had been at the expense of a chaotic proliferation of names. Unrestrained by any biological paradigm, workers had attempted to locate as many distinctive species as possible. Now a new rationalism broke out, led by the soft-spoken and self-contained Will Hass. Hass had moved to the National Museum in 1935 to work with the USGS. Here Girty trained him in the art of micropaleontology and Roundy gently directed him toward the study of conodonts. In 1938, Hass coauthored an account of some Montana conodonts, calling on Cooper to help with his identifications. In all Hass identified forty-six new species. Here, a little naïvely, he took a swipe at Branson and Mehl's index genera – a division Stauffer had recently and un
knowingly confirmed. Hass claimed that Branson and Mehl's Mississippian Siphonognathus also occurred in the Devonian, while their Devonian Palmatolepis could be found with two of their Mississippian indicators. Branson and Mehl told Hass that he must have used contaminated samples, but Hass thought the associations a “curious coincidence.”31

  In the following year, Cooper was bedazzled by 256 remarkably preserved conodont species in an eleven-inch-thick Mississippian (so he believed) shale, “the most striking assemblage of these forms encountered in my experience,” he noted. “Some species are so preserved that they are almost transparent and appear as though constructed of cellophane, giving the appearance of artificiality.” As we shall see in the next chapter, this extraordinary material permitted Hass to undertake internal investigations of conodont fossils and reveal the true extent of the name proliferation. It fundamentally altered Hass's outlook, and in 1943 he returned to his Montana specimens and decided that he had not found a single new species.32

  Two years later, Branson and Mehl began to talk defensively about the actions they had taken. Critics were all too ready to forget the wilderness out of which the two men had led the conodont. Branson and Mehl argued that without names there could be no progress, even if the full history of a form remained unknown at the time. Each named species then existed as a discrete entity in time and space. It possessed no future, no past, and no geography. It was from such flags in the ground that science built its knowledge, and inevitably it would then need to make corrections. It was to be expected that, as knowledge matured, a period of name inflation would be followed by deflation as they realized that the same fossil found in different rocks had been given different names.33

  The other source of error was contamination, which Branson and Mehl wrote about at length in 1941, using examples from Huddle, Cooper, and Stauffer. Detection, they said, in a rather circular argument, relied upon “knowledge of conodont successions rather than in field evidences.” There were, they said, such self-evident reports of mixing as the discovery of unworn conodonts in the Triassic of Germany that the problem could not be denied. Branson's son, Carl, had been one of the first to find Permian conodonts in the United States and fancied he saw the “senile specializations of a rapidly declining group.”34 Branson and Mehl thought the German fossils looked like those found in the American Pennsylvanian and Permian and must therefore result from contamination of the overlying Triassic rocks. To have persisted into the Triassic, these animals would have survived the great extinction that took place at the end of the Permian and crossed from the Paleozoic world into the Mesozoic virtually unchanged. The idea seemed impossible.

  At the heart of the contamination argument was a growing feeling, which had certainly taken hold by 1938, that the conodont fossils showed progressive advancement over time. One could guess the age of these fossils simply from how advanced they seemed. It was a powerful idea. One that would prove irresistible to later generations and that, in this era, caught the imagination of Bill Furnish, a young doctorate student at the University of Iowa. Here a hard-working cephalopod paleontologist, A. K. Miller, was pulling together the beginnings of one of the most influential conodont research schools of the mid-twentieth century. To achieve this, Miller, a former student of Ted Branson, drew upon the influence and encouragement of his Missouri and Minnesota neighbors. Furnish was the first product of this new school and was already examining some of the oldest faunas and giving serious consideration to classification. He concluded that the fossil record indicated “beyond reasonable doubt that compound teeth were derived by several distinct lines of evolution.”35 In other words, and contrary to what most believed, there was no single evolutionary progression; development had followed a number of parallel paths. This meant that Ulrich and Bassler's classification, which had grouped together “teeth” on the basis of superficial resemblance, was no more reliable than Pander's. These “lines of evolution” were, Furnish believed, open to interrogation and would permit a truer classification. Only by this means could the science move beyond the connoisseurship of abstract forms. But in order to achieve this, further action was needed to tackle the proliferation of names.

  Resolution of this problem required a curatorial mind – someone who could collate, distinguish, and arrange. Fortunately such a mind was possessed by Mehl's doctorate student and protégé, Sam Ellison. Ellison sought to bring order to the Pennsylvanian.36 Reviewing thousands of specimens, when his predecessors Gunnell, Stauffer, and Plummer possessed only a few hundred, he could confirm that minute differences of form were an irrelevance. With this he began a mass extermination of species. Whole lists of names were now vaporized as one name came to define what had previously required a dozen or more. This was in 1941.

  Delighted with his results, Ellison now turned his attention to the whole of the Paleozoic. In 1944, his former mentors, Branson and Mehl, had summarized the state of knowledge in Shimer and Shrock's definitive Index Fossils of North America.37 Frequently cited, the book nevertheless lacked the clarity and security of argument that Ellison, then at Stanolind Oil and Gas in Midland, Texas, would find just two years later. Extracting the stratigraphic ranges of some fifteen hundred species mentioned in more than 180 papers, Ellison mapped the ranges of those genera he considered valid on charts ranging from the Lower Ordovician to the Upper Permian. The result had a clarity and logic never before seen in conodont studies. In this respect it was much like earlier pictorial arguments. It spoke directly of the wonderful utility of these fossils and their seemingly distinctive forms. It seemed to offer an almost perfect portrayal of their progressive evolution: “The fibrous and non-fibrous conodonts develop in form from simple cones through the bladed and bar stages to the platform types. The simple cones give rise to blades and bars by addition of denticles to the basic cone. The basic cone becomes the apical or largest denticle. Bladed and bar forms give rise to platforms by a process of lateral thickening of the blade or bar…. Such excellent examples of evolution impress the idea that the conodonts are among the best fossils for family tree studies.”38 Here was the key to a new utilitarian future: abstract things on an evolutionary journey that left behind the footprints of their development.

  2.1. The solution to the black shales dispute. Ellison's mapping of the stratigraphic ranges of conodont fossils demonstrated the fossils’ utility at a time when most paleontologists knew little about them. This part of the chart, illustrating Branson and Mehl's index genera, showed that conodonts permitted the detection of boundary between the Devonian and the Mississippian. From S. P. Ellison, Bulletin of the American Association of Petroleum Geologists 30 (1946). AAPG ©1946, reprinted by permission of the AAP G, whose permission is required for further use.

  This convincing summary of knowledge had implications for the black shales, for it embedded Branson and Mehl's index genera and made concrete the error of Ulrich and Bassler's interpretation (figure 2.1). Kindle's vision for the fossil and the black shales had been realized. This was certainly Hass's view as he attempted to zone rocks across the disputed boundary in Ohio in the mid-1940s. The results demonstrated the truth of Branson and Mehl's index genera – evidence he, Cooper, and Huddle had previously doubted.39 In the late 1940s, Hass began to sample against a tape measure at “at the tenth of a foot scale.”40 This was a level of resolution above that achieved by other workers, and on Hass's early death in the following decade, Carl Branson wrote that Hass had made “the first real progress in solving the black-shale problem.”41 This was, of course, untrue; that honor surely belonged to Branson and Mehl.

  In little more than a decade of intense activity, helped on by Ulrich's outspoken views, the conodont fossil had become a sophisticated utilitarian tool. But it remained in the hands of a cottage industry. By 1941, some five hundred people were engaged in the oil industry studying microfossils at an annual cost of $2.5 million. The pages of the ever-expanding Journal of Paleontology indicate that six to ten times more attention was being paid to foraminife
ra, and three to five times more to ostracodes, than was given to conodonts.42 At that moment of peak conodont activity in 1933, Brooks Ellis, working on a definitive Catalogue of Foraminifera, had eigthy-two artists, photographers, geologists, typists, and librarians working for him. They were funded through the Civil Works Administration, which had been set up to relieve unemployment during the Great Depression. By 1935, his catalogue had twelve to fifteen thousand species descriptions, twenty-five thousand illustrations, and three hundred thousand references.43

  By 1953, four out of five geology graduates became oil geologists, earning salaries of three hundred dollars a month or more. By then the oil and gas industry employed eight to ten thousand geologists, and many, such as Chalmer Cooper (who lived beyond his one-hundredth birthday), would be involved in encouraging the next generation: “You may have to be a nomad instead of settling down for life in one spot. You may have to ‘sit on’ a well all night and then drive a hundred miles to report on it. You may have to burn in India, freeze in Alaska, or do both in the Texas Panhandle.”44 Some were deeply alarmed by the transformation that had taken place in this science: “In North America, particularly, we have tended to be concerned with the uses of invertebrate fossils in the solution of geologic problems rather than with the meaning, in the broadest sense, of the fossils.”45 Oil had made paleontology the handmaiden of industry. The dour Arthur Cooper, at the National Museum, thought the well-paid micropaleontologists were splitting the science in two.46 Nowhere could this split be better seen than in the science's treatment of the conodont. The utilitarians now felt they possessed it. But they did not. Elsewhere, others were finding evidence of the animal.