The Great Fossil Enigma (7 page)

Hinde's identification of the first Scottish conodonts prompted the society's vice president, John Young, to reflect on the animal's growing mythology: “Although these curious tooth-like organisms have now been known to Palaeontologists for more than twenty years, great doubts still exist as to what group of animals they belong.” Smith would later find conodonts from many English localities and even noted, rather prophetically but without effect, “They are found chiefly in the powder of rotted limestone, and, but rarely, in shale.”
²⁸

Hinde reported the results of his Great Lakes investigations in two papers read to the Geological Society of London in March 1879. His search revealed that between their earliest occurrence and the Lower Carboniferous, the conodonts had no regular association with other fossils, and thus the tooth-like fossils really were the only surviving parts of the mysterious animal.
²⁹
He had found conodont species very similar to those described by Pander that provided the first indication that these new and tiny things were distributed over immense distances and might, as a result, be used to correlate rocks. In contrast to Charles Moore, who had been overwhelmed by their eccentric wonder, Hinde could show that simple teeth, which Pander had found in such profusion, were restricted to the older rocks while the compound forms, composed of multiple points or cusps, had a far more extensive range. The Ordovician (as we now know it) and Middle and Upper Devonian were rich in these fossils. By contrast, in the intervening Silurian and Lower Devonian he could find none.

He had found that particular strata had particular conodont distributions. One layer, less than three inches thick, exposed at Eighteenmile Creek near North Evans on the southern shore of Lake Erie, was so crammed full of these fossils that he named it the “Conodont-bed.” This rock belonged to the Middle Devonian, and in this and the overlying Upper Devonian he found the variety of forms unparalleled. And then, at North Evans in the Upper Devonian Genesee Shale, in one lucky hit he split open a slab to reveal a cluster of twenty-four conodont fossils and associated “plates,” all squashed into an area no larger than a quarter of an inch in diameter. Hinde could not believe these fossils had been swept together by water currents, as the fossils themselves were too delicate and would have been destroyed. An alternative explanation was to believe that they represented a natural association, reflecting their presence in the soft tissues of the animal. They had simply been “crushed together into a shapeless mass” during fossilization. His conclusion was striking, for it suggested that different kinds of the conodont teeth occurred within the same mouth. He called the animal that possessed them
Polygnathus dubius
, a name which signaled not doubts about the truth of the association but rather its zoological meaning, as the new find did nothing to resolve the affinity of the animal or the zoological structure and function of the animal's jaw apparatus. It was a revolutionary interpretation that Hinde saw as a step closer to the truth. Unfortunately, the rest of his specimens were isolated finds and he had no way of artificially constructing associations between them. This meant that each of these fossils – if they were unlike those making up
Polygnathus dubius
– had to be given a species name of its own. In other words, for these fossils, he was forced to continue that practice Pander had reluctantly adopted. Like Pander, Hinde knew that twenty of his species names had no biological meaning; he reckoned that in all he possessed only two or three true species. In time he thought these true species would be recognized, but for the moment he had to accept a mixed system that, while it sought to establish a biological truth, sacrificed clarity and simplicity. This sacrifice was small, however, if all one wanted to do was understand the animal.

Huxley told Hinde that his
Polygnathus dubius
closely resembled the teeth of the hagfish such “that it would be difficult to prove that they did not belong to fishes of this order.” But he added that no living fish held such an assemblage. Hinde had found annelid worm jaws in the same strata as those in which he had found his conodont fossils. Like Grinnell before him, he could show that the worm jaws were distinct and possessed the morphology and chemistry of modern forms (
figure 1.1
).
³⁰
Hinde was emphatic: The animal that possessed the conodont fossils was not a worm. All the evidence pointed to conodonts being relatives of the hagfish as Pander had proposed, and he told his audience not to prejudge the past diversity of these fish on the basis of the “pauperized descendents of the present day.” While Hinde admitted that his evidence permitted only tentative conclusions, in two complementary papers, he had masterfully broken the thread of thinking that linked Owen to Ulrich. Now the dissidents – Pander and Newberry – appeared prophetic. The worm was dead, the fish resurrected. And in a few years, as American geologists began to investigate the same rocks and localities, so Hinde's ideas began to be consolidated. For the first time a complex biological species of animal –
Polygnathus dubius
– began to inhabit the American scientific mind.
³¹

1.1.
Hinde's proof. Hinde claimed that these different kinds of conodont fossil (
left
) came from a single animal, which he called
Polygnathus dubius.
He showed that these were in every way different from worm jaws (
right
). From G. J. Hinde,
Quarterly Journal of the Geological Society
35 (1879).

While the French, British, and Americans had each in turn gotten a little excited, even perturbed, by the arrival of the conodont, the Germans had remained surprisingly quiet on a subject that had first been published in their language. It was the evangelizing Darwinian Friedrich Rolle who first broke this silence. Reviewing opinion, he concluded that if conodonts were not relatives of the hagfish, they might be related to the small eel-like amphioxus (or lancelet)
³²
or to the sea-squirts (which have remarkable free swimming larvae). Both these latter animals possess a notochord, a structure in place of the spine, seen in the embryos of vertebrates. It suggested that these invertebrate animals were related to vertebrates and that they might be brought together with the vertebrates in a group known as the chordates. It was a suggestion with connections to the pioneering embryological work of Baer and Pander and seemed to place the conodont in the realm of those animals from which vertebrate life first evolved.

German-speaking scientists could now no longer ignore the conodont. For Germany's Owen-like encyclopedist, Karl von Zittel, who had risen to fame during the 1870s, the conodont became a problem requiring resolution. Zittel was now working on his five-volume
Handbuch der Palaeontologie
, which was to become the definitive late-nineteenth-century guide to the study of fossils. Zittel's interest in the conodont may have arisen from the necessities of the book or have been triggered by the arrival of Hinde in Munich. Hinde had traveled there to begin a doctoral dissertation on fossil sponges under Zittel – an expert in the group – not long after his conodont paper had been read in London. To tackle this new conodont problem, Zittel teamed up with established Baltic fish zoologist Josef Victor Rohon, who over the previous decade had very usefully published on the anatomy of all those vertebrate and near-vertebrate animals considered to be relatives of the conodont: sharks and rays, lampreys and amphioxus. Rohon was also no stranger to the microscope.

From a review of the literature on the enigma, Zittel and Rohon noted that the conodont's “most faithful companion” was the annelid jaw. Perhaps this was important given that the conodont fossil itself was rather poor evidence for such anomalously early vertebrate life. Believing that they had the advantage of superior technologies and new knowledge resulting from recent discoveries in other fields, the two men felt a solution to the “conodont question” was in sight. What they lacked, however, were conodonts – a deficiency soon rectified with material from St. Petersburg and from Hinde's American collections.

Their investigation revealed the presence of organic material in the conodont fossils, which tacitly signaled a link to the chitinous teeth of worms. The sheer variety of form was to them even more conclusive: “Against an interpretation as the toothed jaws of vertebrates; it points rather to seizing, catching or supporting organs.” They could confirm the presence of interior structures described by Pander, but then denied most of them any significance. The lamellae survived their assault and they introduced fine radial canals, but the rest of Pander's carefully detected structures were now to be considered mere artifacts of fossilization. Their final and most effective blow, however, relied upon the kind of visual argument Hinde had used to crush the worm, only here they were to destroy the fish. These pictures conclusively showed that the horny teeth of hagfish and their relatives “do not have anything in common with the conodonts.” With the fish destroyed, Pander's figures were compared with Ernst Ehler's published illustrations of bristle worms and with another kind of worm, the gephyrean,
Halicryptus spinulosus.
³³
The similarity of external form was remarkable, and stripped of the materiality of the fossil and a history of interpretation that had long doubted that simple morphological comparisons were useful, Zittel and Rohon's argument was made complete. One did not need to be able to read German; one could simply look at these illustrations and believe.

The two men were in no doubt that these animals were worms of some kind and that “from the large diversification of form it can be concluded that the Conodonts represent numerous kinds and that thus in the Palaeozoic age the coasts of the seas were populated by a substantial number of the most different worms.” Only a few years earlier, Newberry had imagined something similar, only for him the seas were filled with small fish. Now that fish was dead, its lifeblood apparently an illusion brought on by erroneous interpretations. However, a more open-minded reader might have seen in Zittel and Rohon's arguments a conspiracy to murder an unsettling – if tiny – monster.

Rohon confirmed the truth of their new worm four years later when searching through hundreds of conodont fossils collected from the same rocks in which Pander had found his specimens. Rohon was looking for evidence of fish teeth and found ten examples. These specimens were confusingly similar to conodont fossils but were nevertheless different. It made him even more certain that conodonts were worms, not fish. This discovery was picked up by Britain's leading geologist, Sir Archibald Geikie, who wrote in his popular textbook, “According to Dr. Rohon, however, all ‘Conodonts' are not annelidian, but include undoubted teeth of fishes with recognizable dentine, enamel, and pulpcavity.”
³⁴
Something had been lost in translation, but fortunately for Rohon, no conodont worker ever noticed Geikie's interpretation.