Bonehead mistakes: The background in scientific literature and illustrations for Edward Drinker Cope s first restoration of Elasmosaurus platyurus

Transcription

1 PROCEEDINGS OF THE ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA 152: OCTOBER 2002 HISTORICAL POINT OF VIEW Bonehead mistakes: The background in scientific literature and illustrations for Edward Drinker Cope s first restoration of Elasmosaurus platyurus JANE P. DAVIDSON Department of Art, University of Nevada Reno, Reno NV , USA ABSTRACT In 1868, Edward Drinker Cope incorrectly restored the type specimen of Elasmosaurus platyurus, by placing the skull at the end of the animal s tail. His error haunted him the rest of his career. In examining the scientific literature and popular science literature available to Cope prior to 1868, as well as taking note of the extent of knowledge which I suggest his professional colleagues shared with him concerning plesiosaurs, it seems impossible that he should have incorrectly restored the fossil. I further suggest that he made this error more than once, even after his mistake was pointed out to him. Mr. Conybeare has justly remarked how difficult it is to determine the number of the cervical vertebrae in a Plesiosaur, owing to the gradual transition in their lateral appendages from the condition of hatchet-shaped laminae to the ordinary elongated form of ribs. Richard Owen A Description of a Specimen of Plesiosaurus Macrocephalus, Conybeare, 1840 It is this apparent adaptation of the parts to the articulation of chevron bones which has led me to consider the vertebrae under consideration as caudals, otherwise from their resemblance to the cervical vertebrae of Plesiosaurus pachyomus, as represented by Prof. Owen, I should have viewed them as belonging to the cervical series. Joseph Leidy, writing on Discosaurus vestutus in Cretaceous Reptiles of the United States, 1865 It is the least able contribution to paleontology that we remember...itcontains no science. Thomas Huxley Review of Leidy s Cretaceous Reptiles of the United States Geological Magazine, 1868 Elasmosaurus platyurus... The anatomical characters of the different regions of the vertebral column...aredecidedly Plesiosaurian. Prof. Cope has described the skeleton in a reversed position to the true one and in that view has represented it in a restored condition... Joseph Leidy Communication to the Academy of Natural Sciences of Philadelphia March 8, 1870 ELASMOSAURUS PLATYURUS MAKES ITS APPEARANCE Certainly the most notorious act of paleontology in the entire career of Edward Drinker Cope ( ), a mistake to which his great rival Othniel C. Marsh generally referred when he frequently made fun of Cope s scientific blunders, was his incorrect reconstruction of the skeleton of the marine reptile, Elasmosaurus platyurus. 1 Cope never lived it down. The story is well known in the history of paleontology. He had reconstructed the skeleton by placing the skull on the wrong end of the vertebral column, thus rendering Elasmosaurus with a short neck and a long tail. In reality the anatomy of this relative of Plesiosaurus was just the opposite; the animal had a long neck, just like a plesiosaur. When O.C. Marsh made considerable light of this error, it helped to galvanize what became the lifelong antagonism between Marsh and Cope. Marsh returned to the Elasmosaurus incident repeatedly during the print controversy, which erupted between the men in the mid-1890s in the New York Herald. 2 Theirs was not only a professional rivalry but a genuine hatred of one another. Part of this hatred came from the embarrassment Cope felt over his reconstruction of E. platyurus. Edward Drinker Cope was one of the foremost paleontologists of the second half of the nineteenth century. He was also an expert herpetologist. Even as a young scientist, he was respected internationally for his 215

2 216 J. P. DAVIDSON

3 ELASMOSAURUS 217 contributions to biology and paleontology. There was no question of his brilliance, Marsh s sniping notwithstanding. But Cope was also infamous for making rapid assessments of fossils and for rushing his sometimes sloppily-written observations into print. Was this sloppiness the cause of the Elasmosaurus incident? O.C. Marsh did not think so. It was plain that he thought his young colleague was incompetent, at least in this instance. The issue was, and remains, how could Cope have made such a seemingly obvious anatomical error in reconstructing E. platyurus? To this I would add the question of whether this was indeed such a seemingly obvious error? Can we elucidate the questions of what sorts of scientific knowledge Cope might have been able to access in 1868? EXPLANATIONS? As we look at the context of paleontology in which Cope was working in the late 1860s, we can suggest several possible explanations for Cope s mistake. This error could have been understandable in the face of the fact that no one had ever seen a specimen of Elasmosaurus platyurus. Cope s specimen turned out to be a unique fossil, not only the type specimen but the only known example of the genus and species. The skull was not found proximate to the vertebral column, so the articulation of skull to vertebral column might not have been immediately evident. Cope suggested this was the case. But, when Joseph Leidy commented in print on the fossil in March, 1870, he noted that the occipital condyle was still attached to the atlas/axis. 3 When Cope first examined the fossil, he believed the matrix actually contained two different specimens of marine reptiles. Perhaps this misunderstanding of the specimen helps to explain some of Cope s apparent confusion concerning the placement of the skull vis à vis the vertebral column. Whatever his interpretation of the osteological evidence, Cope was to place the skull at the wrong end of the vertebral column. Cope s incorrect restoration of E. platyurus might also have been understandable because it happened in 1868 when the professional science of paleontology was still relatively new and when Edward Cope was a very young man with limited experience in the field. It also seems likely that other scientists might have reconstructed the skeleton in the same way. There were after all still relatively few specimens of any fossil reptiles, including plesiosaurs or their relatives, known anywhere in the world to paleontologists. Scientists with greater academic credentials than Edward Cope s were puzzling over various specimens of Plesiosaurus. So, it does not seem impossible that a twenty-eight year old scientist who had only one course as his entire formal college training, could have easily got wrong his analysis of the vertebrae of the fossil because even his mentors were having problems with plesiosaur skeletons. Such are the explanations put forward by authors, myself included, who have discussed how Cope wound up with the animal s head put on backwards. I believe that the first difficulty with these explanations is that other scientists immediately realized Cope s error. It is my opinion that this mistake was pointed out to him immediately upon his restoration in 1868 by his mentor and friend Joseph Leidy ( ). I suggest here that Leidy would never have waited until March, 1870, the point of his published comments concerning Cope s mistake, to mention this to Cope. I believe that Cope simply did not change published examples of his erroneous restoration of the specimen until after Leidy s printed remarks. I think that Cope was simply not immediately convinced he had made a mistake. We are dealing here with an example of Cope s well known arrogance. The story of what happened in March, 1870, when Cope suppressed a preprint containing his incorrect restoration, seems to back this up. It is rather unfortunate that the publication customs and procedures from this period do not make matters clearer to the modern reader. Following the time lines of who said, wrote, [Facing page.] The suppressed version of Edward Cope s reconstruction of Elasmosaurus with a short neck. Cope included Plate II with privately distributed preprints of his Synopsis of the Extinct Batrachia and Reptilia of North America, Part I, an article that was to appear in the Transactions of the American Philosophical Society. The preprint, dated August, 1869, was recalled by Cope in March, The text and plate were corrected, and preprints were again distributed. Apparently, in addition to the plate, only those signatures containing the erroneous texts were replaced by corrected ones. The cover sheets of the two versions of the preprint are identical; each has the August, 1869 date. Cope s paper was again distributed with cover sheets dated 1870, this time with the indication that it formed part of Volume XIV (New Series) of the Society s Transactions. Customarily, many scholarly societies of the day published portions of their journals in installments, with or without cover sheets or separate wrappers, and it was up to the recipients to accumulate these mailings and bind them together when the volumes were completed. Volume XIV of the Transactions was fully collated in December, 1871, as noted by the title-page for the volume. (Plate courtesy of The Academy of Natural Sciences of Philadelphia, Ewell Sale Stewart Library, Collection 328.)

4 218 J. P. DAVIDSON and read what and at which date becomes almost as confusing as the issue of which end of E. platyurus s vertebral column was which. Cope had originally presented an erroneous reconstruction of the skeleton in 1868 when he spoke to the American Philosophical Society on September 18, His remarks before the Society were continued on April 2, The preprint of the paper accepted by the Society for publication showed an illustration with the restored skeleton of Elasmosaurus platyurus, its head at the wrong end of the vertebral column. That preprint was dated, August, Leidy spoke at a meeting of the Academy of Natural Sciences of Philadelphia on March 8, 1870, a year and a half after Cope s first presentations, and then made note of Cope s incorrect placement of the skull of E. platyurus. These remarks were subsequently published. Later in March, 1870, Cope hurriedly tried to recall all of the copies of his preprint article along with the incorrect illustrations of the specimen, in an effort to cover up his mistake. He had his corrected article reprinted, even using the same date of August, 1869, showing an illustration of the restored skeleton with the skull on the correct end of the vertebral column. In the text he replaced the word neck with the word tail, revised his descriptions of key skeletal elements, and substituted a corrected plate. 5 This second version of the preprint was issued by Cope sometime after March, Even though Cope had tried hard to get back and destroy the incorrectly written preprints, he was, alas, too late; a copy of the original preprint with its incorrect reconstruction of E. platyurus had come to Marsh s attention. And the rest is history. This rather confusing set of circumstances makes it seem as though Cope did not hear from Leidy or anyone else for that matter concerning his mistake until the spring of I disagree with this. It is my contention that Leidy certainly must have been aware that Cope was mistakenly restoring the skeleton in 1868 at the time of Cope s remarks to the American Philosophical Society. I further propose that it is unbelievable and illogical that Leidy, Cope s mentor and friend, would have waited so long to bring this to Cope s attention. Leidy was a better scientist and a more honorable man than that. I think that Cope had enough faith in his own original opinion of as a short-necked animal to keep this restoration Elasmosaurus platyurus until March, 1870, when he finally changed his mind. This essay also explores the question of how Cope could have made such a mistake in the context of the status of knowledge of fossil marine reptiles in the mid-nineteenth century. It turns out that the suggestions that Cope s limited educational background or the primitive status of contemporary paleontology were to blame for his error are not well founded. In reality, I believe that it can be demonstrated that Cope should have known better than to place the animal s skull on the wrong end of its vertebral column. A corollary question is what did Cope know about fossil marine reptiles prior to his examination of the fossil of Elasmosaurus platyurus in 1867 and 1868? Once we establish the paleontological literature and the scientists to which he had access, then the reasons why he made his mistake may become clearer. We need to explore the state of vertebrate paleontology at mid-century, and we need to explore Cope s own personality, motives and public statements concerning fossil marine reptiles and the Elasmosaurus platyurus. It is time to examine the case. Was Cope s error excusable because he was very young and had no formal college training in geology or paleontology? It is of course true that he was young and not college educated. He was twenty-seven when he received the specimen from Kansas. Two vertebrae were sent to him in 1867 by Joseph LeConte, and the remainder of the skeleton arrived in early March, He took a one year s course in comparative anatomy from Joseph Leidy at the University of Pennsylvania probably during the academic year This was the entirety of Cope s college education. Cope s papers contained a student s admission slip for this course for the year , so it is possible that he took the class during this academic year instead. In effect, he took the course once, and probably did not even make it to all the lectures because he was not in Philadelphia all the time during this period. We know that Cope was in Washington, DC quite frequently during because he had begun work as a research herpetologist in January of 1861 at the Smithsonian Institution. 8 But, in spite of his youth and lack of formal university training, Cope was still among the first scientists to discuss fossil marine reptiles and to posit theories about their ecology. How did Cope manage to establish himself as an important scientist without benefit of a college degree? First of all, he was just incredibly intelligent. But, his success as a scientist also had to do with early opportunities which he had to conduct research in two important scientific institutions. His early employment at the Academy of Natural Sciences of Philadelphia and the Smithsonian Institution provided him access not only to current scientific literature which described marine reptiles, but also to scientists who were working with these fossils. These resources in effect made up for what he might have learned in the fields of geology or paleontology at university. He became a member of the Academy of Natural Sciences of Philadelphia on July 30, 1861, two days after his twenty-first birthday. But his association with the Academy extended back to 1858 when he had been employed by the Academy to do research in and to recatalog its herpetological collections. Again, because

5 ELASMOSAURUS 219 of his association with the Academy, he would have had access to a wide community of American scientists and should certainly have been able to read publications of European paleontologists while he was still a teenager. Cope s first biographer, Henry F. Osborn, wrote about what a tremendous resource the Academy of Natural Sciences was to Leidy and then to Cope in his turn, it is difficult to realize how much the privileges of the academy library, apparatus and collections as well as the interest and cooperation of the specialists in its circle of members could mean to the young scientist. 9 By the winter of 1861, Cope had extended his connections to include the Smithsonian Institution where again, as a very young man, he was working as a research scientist. Here too he would have had access to current paleontological literature and to other scientists. Since we can easily prove that Cope was in a position to learn what there was to know about the current status of paleontology at the Academy of Natural Sciences and at the Smithsonian, it seems that his mistake with E. platyurus is all the more surprising. But before we decide that this is the case, we next need to examine what exactly Cope could have learned from reading contemporary scientific literature or by meeting other paleontologists at these institutions. PLESIOSAURS ARE PROBLEMATIC Elasmosaurus platyurus is today believed to be a genus related to Plesiosaurus. 10 Plesiosaurs were among the first fossil reptiles discovered and described in the nineteenth century. Some of the earliest specimens were found by the English fossil collector and seller Mary Anning. The earliest scientific descriptions were made by William Daniel Conybeare ( ) and Henry Thomas de la Beche ( ) in The genus, Plesiosaurus, was erected in 1821, and a description of the type species P. dolichodeirus was made by Conybeare and De la Beche in This specimen was a skeleton which had been discovered by Mary Anning in Shortly after this, many additional specimens were discovered in Dorset and some partial plesiosaur specimens were also found in New Jersey. Plesiosaurs were relatively big and rather fantastic-looking creatures. Because plesiosaurs had jaws filled with menacing teeth, most early paleontologists immediately presumed that these animals had been aggressive carnivores. It was recognized that they were reptilian on the basis of osteology, but their anatomy seemed confusing, and questions arose as to how to properly classify plesiosaurs with respect to living reptiles. The problems and perhaps the incredulity presented by the uniqueness of plesiosaur anatomy was described by William Buckland ( ) in 1837 in The Bridgewater Treatise: Geology and Mineralogy Considered with Reference to Natural Theology. Buckland wrote: To the head of a lizard it united the teeth of a crocodile, a neck of enormous length, resembling the body of a serpent; a trunk and tail having the proportions of an ordinary quadruped, the ribs of a chameleon and the paddles of a whale. Such are the strange combinations of form and structure in the Plesiosaurus. 12 Note that Buckland immediately commented upon a most salient feature of plesiosaurs as they were known to him, its neck of enormous length, even in this very preliminary remark on the animal. Further, Buckland published engravings of the reconstruction made of P. dolichodeirus by Conybeare. Confusion about how to articulate plesiosaur skeletons and how to classify individual specimens into particular genera or species continued through much of the nineteenth century. And evidently such classification questions still arise. Writing in 1997 about the early status of research on plesiosaurs, Glenn Storrs commented that after Conybeare in 1824 and Owen in 1840 and 1865, there have been few attempts to clarify the anatomy and the status of the type species of Plesiosaurus or, indeed, of the genus itself. As a result a historical lack of understanding of the range of variation in plesiosaurs and a poor knowledge of their generic characteristics has reduced Plesiosaurus to the status of a wastebasket taxon A statement like this might certainly lend credence to the notion that Edward Cope could have made his osteological errors, but is this really the case? The plesiosaur became immediately famous in the middle of the nineteenth century as an example of fossil life. Its inclusion was de rigeur in any geology or paleontology texts. It was discussed in texts written by authors whose field was natural theology, that attempt to square new science with the Bible, and the plesiosaur even appeared in books designed to bring geology and earth history to school children or to the general public. It was as though everyone needed to be told about Plesiosaurus. Much of this notoriety came from the fact that there were so few vertebrate fossils known in the early nineteenth century. Along with the ichthyosaur and the pterodactyl, this wondrous marine reptile came to represent to the popular mind which prehistoric or perhaps better said, pre-adamite life was like. Scientists would not really begin to grasp the enormous vastness of the fossil vertebrate record until the 1870s, when scientists like Cope and Marsh were shipping home literal tons of specimens from the American West. Thus, one can easily imagine

6 220 J. P. DAVIDSON Fossil remains of Plesiosaurus macrocephalus, accompanying Richard Owen, A Description of a Specimen of Plesiosaurus Macrocephalus, Conybeare, in the Collection of Viscount Cole (Transactions of the Geological Society of London, 2nd Series, Vol. V, 1838, Pl. 43 [detail]). Plesiosaurs and their allied taxa have long been understood to have extended necks. the excitement of Edward Cope when a great specimen of a large unknown new marine reptile similar to Plesiosaurus arrived on his own doorstep in The fossil which Cope would name E. platyurus was found in a small exposure of shale in the middle of the prairie near Ft. Wallace, in western Kansas, by the Army surgeon, Dr. Theophilus H. Turner. 14 An early discussion of the fossil based on Cope s preliminary analysis of the specimen is to be found in John L. LeConte s Notes on the geology of the survey for the extension of the Union Pacific Railway, E.D. [Eastern Division] from the Smoky Hill River Kansas, to the Rio Grande, published in February, 1868 by Review Printing House of Philadelphia. LeConte had transported part of this fossil to Cope in 1867, and the remainder of the skeleton arrived there in early In a letter written to his father, Alfred Cope, on April 3, 1868, Edward mentioned the fossil and his initial assessment of its vertebrae by noting, I have recently received a very interesting fossil from 500 miles west of Leavenworth, Kansas. It consists of 100 vertebrae, large parts of the scapular and pelvic arches, parts of limbs and head of a great Sea Saurian, like the Plesiosaurus. In proportions it is like the latter reversed. He went on to comment that this was an animal representing not only a new genus, but a new family.... I call the Enaliosaur Elasmosaurus platyurus. 16 While Cope would shortly afterward tell John LeConte that he thought there were two distinct specimens in the matrix he had received from Kansas, he did not say as much in this letter to his father. Apparently that assessment came during the summer of 1868 while he worked with the fossil material. Cope then

7 ELASMOSAURUS 221 Typical reconstruction of a plesiosaur skeleton, depicting a creature with a long neck. Note also the attention drawn to the orientation of the cervical vertebra (inset). (Charles Lyell, A Manual of Elementary Geology, Little, Brown and Co., Boston, 1855, fig. 416.) proceeded to study the specimen and present his findings at the American Philosophical Society on September 18, LeConte s description of the fossil material was taken directly from comments made to him by Cope prior to his revision of the position of the skull. This description is useful on several levels. First, it tells us something about the relative positions of the bones to matrix as they were found by Turner. And secondly, it contains a most interesting description of the fossil quoted by LeConte from Cope himself. LeConte wrote, Mr. E. D. Cope has kindly furnished me with the following communication upon the collection of fossil bones, made by Dr. Thos. H. Turner U.S.A.... which was received at the Academy of Natural Sciences of Philadelphia while the last pages of this report were passing through the press. 17 LeConte went on to quote Cope at length concerning the specimen of E. platyurus as well as a second specimen which had been collected by Turner. These comments elucidate some of the status of the specimen as Cope received it, and thus might have indicated some reasons why he would have initially confused the nature of the vertebral column. And, just as he had told his father in April, Cope told LeConte that the animal which would be later identified as E. platyurus had a very long tail. It is also significant that Cope believed his new specimens related to plesiosaurs. The collection recently received from Dr. Turner contains remains of two species of Marine Saurian the larger of some forty feet in length, the other not more than half as large. Both are allied to the plesiosaurus the smaller being provisionally referred to the immediate American representative of that genus. The larger approaches genera already known from cretaceous rocks of New Jersey and Southern State, called by Leidy Discosaurus and Brimosaurus...(an amount of fossil from the larger) sufficient to indicate its distinction from them and that it was one of the largest and most formidable of the group of Saurian to which it belongs... A dorsal view of a fleshed-out skeleton of a plesiosaur, illustrating the relationship of the creature s long neck to the rest of its body. (G. F. Richardson, Geology for Beginners, 2nd ed., Longman, Brown, Green and Longmans, London, and Wiley and Putnam, New York, 1843, fig. 245.)

8 222 J. P. DAVIDSON Edward Cope s Laelaps [Dryptosaurus] grins over the last gasps of a short-necked elasmosaur flailing in the Cretaceous New Jersey surf, while the text describes elasmosaurs as whale-like, the neck long and flexible. (Cope, The Fossil Reptiles of New Jersey, The American Naturalist, Vol. III, April 1869, Pl. 2.) Cope continued, concerning what he believed to be the larger specimen: It possesses a tail of great length which was elevated compressed and adapted for sculling the ponderous body through the water. The limbs appear to have been disproportionately small... of the head unfortunately little is known and the form of the neck has not yet been determined as the details of the structure are as yet concealed in the matrix... Elasmosaurus platyurus Cope. The smaller saurian was represented by the end of the muzzle with the teeth and perhaps a few vertebrae. It has received the name of Discosaurus carinatus Cope. 18 This statement to LeConte obviously indicated that Cope originally thought that the elasmosaur should have a short neck and long tail. It seems clear from this statement that Cope did not even realize he might have made a mistake until Joseph Leidy pointed it out to him. This I believe happened later in Cope s misunderstanding of Elasmosaurus is also revealed in an early article which Cope published in the American Naturalist beginning in 1867 and continuing in The Fossil Reptiles of New Jersey reflected the results of his own explorations in New Jersey and those of other scientists, like Leidy and O.C. Marsh. 19 The publication timetable of the article on New Jersey s fossil reptiles overlaps the period during which Cope received the fossil of E. platyurus. Material in the article which deals with plesiosaurs and their relatives is in the second section of the article from 1869, but while this article was published after the correction of E. platyurus, the article s contents still show confusion on Cope s part as to the anatomy of marine reptiles. This article was not a scientific report as such, but a more general descriptive presentation of paleontology. It was, as such, an early attempt at paleoecology, a theme which always intrigued Edward Cope. From the beginning of his career in paleontology, he tried to theorize about how animals had lived and behaved. Cope described reptilian life in the ancient seas of eastern North America by writing: While grim and monstrous Dinosaurs ranged the forests and flats of the coast of the cretaceous sea, and myriad s of Gavials basked on the bars and hugged the shores, other races people the waters. The gigantic Mosasaurus, the longest of known reptiles, had few rivals in the ocean. These Pythonomorphs were the sea-serpents of that age; and their snaky forms and gaping jaws rest on better evidence than he of Nahant can yet produce. 20 He went on to describe a rather misunderstood Mosasaurus by commenting,

9 ELASMOSAURUS 223 A recurring theme of 19th century paleontological reconstructions: dullard ichthyosaurs, writhing long-necked plesiosaurs, and hapless flying pterodactyls. These two examples are among numerous and fancied portrayals. Top, Worthington Hooker, Science for the School and Family. Part III. Mineralogy and Geology (Harper and Brothers, New York, fig. 152, 1881). Bottom, Sanborn Tenney, Geology; For Teachers, Classes, and Private Students, revised ed. (J. H. Butler, Philadelphia, 1876, fig. 211). The Mosasaurus was a long slender reptile, with a pair of powerful paddles in front, a moderately long neck and flat pointed head. The very long tail was flat and deep, like that of a great eel, forming a powerful propeller. The arches of the vertebral column interlocked more extensively than in other reptiles except the snakes, presenting in a prolongation of the front of one, which enters beneath that immediately in advance of it, a rudiment of that extra articulation called the zygosphenal. In the related genus Clidastes, this structure is as fully developed as in the serpents, so that

10 224 J. P. DAVIDSON A megathere and a flying pterodactyl eye a floundering plesiosaur in this frontispiece to David Thomas Ansted s The Great Stone Book of Nature (George W. Childs, Philadelphia, 1863). we can picture to ourselves its well known consequences: their rapid progress through the water by lateral undulations; their little motions on land; the rapid stroke; the ready coil; or the elevation of the head and vertebral column, literally a living pillar towering above waves or brush of the shore swamps. 21 At this point it seems that Cope had the mosasaurs confused with the plesiosaurs. This article also discussed elasmosaurs with some similar evidence of confusion as to anatomy and possible classification as plesiosaurs. His next remark concerning the characteristics of the vertebrae and the tail of the elasmosaur indicates that he still had the vertebral column running the wrong way. The reptilian whales of those troubled times were the Cimoliasaurus and Elasmosaurs. These were the plesiosaurs of Cretaceous life and probably had a great range over the earth. Portions of them have been found in England and North America to our far western regions. Cimoliasaurus appears to have resembled Plesiosaurus in general which Elasmosaurus added to its type an enormous and flattened tail, relatively as that of the Mosasaur, or the modern Iguana but not so flat as in the former....butboth of these types present one strange feature. The processes which connect the arches of the vertebrae are related to each other in directions the reverse of that which prevails among vertebrata generally, being perhaps the same as the zygosphen of the serpent and Clidastes without the usual accompaniment. Cope continued his discussion by describing Elasmosaurus orientalis, presumably from New Jersey, which he differentiates from E. platyurus, which of course comes from Kansas. At this point, he describes E. orientalis as an animal with a long neck. The bulk was whale-like, the neck long and flexible, while short paddles and the serpent-like tail sped this most colossal of our sea-saurians on his destructive career.... Cimoliasaurus magnus is more abundant in New Jersey. In bulk it was little inferior to the last, but it was apparently abbreviated and depressed behind and so must have presented a very peculiar form 22 Cope then continued, Elasmosaurus platyurus was forty-five feet in length. As if these descriptions did not provide enough contradictory or at least unclear comments concerning E. platyurus and its putative relative, E. orientalis, Cope then proceeded to outline in detail the illustration which was published along with this article. In the plate accompanying this article the artist has attempted an ideal representation of a few of the subjects which haunted the shores of our country when our prairies were the ocean bottom and our southern and eastern borders were far beneath the Atlantic. Laelaps aquilungius occupies the foreground on a promontory where his progress is interrupted by the earnest protest of an Elasmosaurus. Mosasaurus watches at a distance with much curiosity and little good will, while Osteopygis views at a safe distance the unwonted spectacle. On the distant shore a pair of the huge

11 ELASMOSAURUS 225 Hadrosauri browse on the vegetation squatting on their haunches and limbs as on a tripod. Thoracosaurus crawls up the banks with a fish and is ready to disappear in the thicket. 23 All that description seems harmless enough until one actually looks at the illustration and discovers that Cope has had the artist depict an Elasmosaurus with a short neck! and a Mosasaurus which looks precisely like a Plesiosaurus. The mind reels. What has happened in this article finished, presumably, long after Cope knew that an elasmosaur had a long neck? And more to the point, how can he have placed this illustration directly after his remarks about the long neck of E. orientalis? Did Edward Cope make two or three boneheaded mistakes? Which elasmosaur species does this illustration depict? If it is E. orientalis, then Cope has the animal s head on the wrong end yet again and the illustration contradicts his own text. If it is E. platyurus, then he was still showing the reader an animal with a short neck after he knew that this was incorrect. Something is frightfully confused in this article. Edward Cope was always a man in a hurry to publish his scientific findings, but at the same time he was particularly careful with illustrations in his scientific publications. He personally drew the sketches which he provided to the engravers and saw to their correctness and quality before the plates were printed. We know that when he finally came to agree with Leidy in 1870, he then took immediate steps to change his article in the Transactions of the American Philosophical Society, which had the mistaken anatomy of E. platyurus, so that the illustration of the skeleton would appear correctly. I think we can safely presume that he drew a sketch of this scene of prehistoric New Jersey. And its clear that this article was published after his corrections of the skeleton of E. platyurus. If, however, as I believe, Leidy discussed the restoration and Cope s mistakes with him in 1868, how can this illustration with its short-necked elasmosaur have appeared in print? Apparently, Cope continued with his mistake in the elasmosaur. Given the considerable embarrassment that his first mistake should have caused him, it is astonishing that he did it again. How could he possibly have let this ambiguous and incorrect article get into print in 1869? The answer must be that he thought he was right in the first place and that he had not accepted Leidy s opinions. And it is more astonishing that the mistake in this article in The American Naturalist went unnoticed. I have not been able to find any contemporary comment indicating that anyone had noticed that Cope had published another mistaken reconstruction. Perhaps this mistake was not noticed because this was an article more directed to general readers and was not a scientific report as such. But nonetheless, it seems almost unbelievable that someone did not comment upon it. After all, at this point, Cope had something literally in print concerning Elasmosaurus, which contained errors. The illustration used by Cope in The American Naturalist for Fossil Reptiles of New Jersey was scarcely kept secret. It appeared again in 1869 in a popular magazine, Harper s New Monthly Magazine, with the caption, restored fossil reptiles of New Jersey, from The American Naturalist. Where was Leidy and where was Marsh? Cope s illustration of prehistoric life in New Jersey was derived from several sources. None of these sources was his corrected reconstruction of E. platyurus. For sake of clarity in this discussion, I will review the positioning of the animals. Cope places the animal he calls Elasmosaurus in the foreground of the illustration. Next to it is Laelaps and in the background at the right is a Mosasaurus. 24 One of the sources for this illustration was a published book. This clearly demonstrates that Cope was guided by the works of other authors when he prepared this illustration. Another source was Cope s own sketch of Laelaps. Cope s discovery from New Jersey appears in the scene. 25 But, Cope took inspiration for the animal he calls the Mosasaurus from an earlier illustration in Thomas Hawkins ( ), The Book of the Great Sea Dragons, published in The frontispiece engraved by John Martin depicted two plesiosaurs in combat with an ichthyosaur. These plesiosaurs are the direct source of Cope s Mosasaurus, down to and including a forked tongue shown on one animal. 26 The head of the New Jersey Reptiles elasmosaur is also shown similar to that of Hawkins plesiosaurs with placement of eyes, nostrils and jaws being very much alike in both illustrations. One difference however is that Cope shortened the neck of his elasmosaur from that of Hawkins plesiosaurs. In addition, Cope s mosasaur in New Jersey Reptiles has some anatomical characteristics in common with Hawkins ichthyosaur in that both animals are shown with eyes on top of their skulls and with prominent nostrils placed directly forward, between their eye sockets. It would seem that while Cope was obviously looking at and utilizing Hawkins illustrations, he had decided that his opinions of anatomical reconstruction for Elasmosaurus and Mosasaurus were more accurate than those of Hawkins. This illustration shows the workings of Cope s mind. Here he presented prehistoric reptiles. He took pains to show them in an ecological setting and to describe the animals in activity in that setting. Furthermore, he reasoned that plesiosaurs and elasmosaurs and mosasaurs all looked similar in that they were all marine reptiles. In addition,

12 226 J. P. DAVIDSON some of their anatomical characteristics must have been similar to those of the ichthyosaur. But Cope has in effect disassembled and reassembled the fossils into how he thought the animals should look. And that reconstruction included an Elasmosaurus with a short neck. He copied rather liberally from Hawkins illustrations. I think that Edward Cope owned a copy of The Book of the Great Sea Dragons. Cope left some of his scientific books to the University of Pennsylvania, and today the University s Rare Book Collection holds a copy of Hawkins book as well as his other early treatise entitled Memoirs of Ichthyosauri and Plesiosauri (1834). While the Rare Books Collection does not have a record of accession which lists a previous owner of these books, they could have been among those that were left by Cope to the University. 27 Cope might also have had access to Hawkins two books at the Academy of Natural Sciences. Today the Ewell Sale Stewart Library of the Academy owns both books. There does not appear to be an accession record for The Book of the Great Sea Dragons, but the other Hawkins volume was given by S.G. Morton. Samuel G. Morton ( ), was an ethnologist and paleontologist. He may well have had some tangential connection to the Cope family, as he had attended the Westtown School, where the Cope children, including Edward, were given their elementary and secondary educations. I have not been able to determine when Morton s donation of Hawkins was made. That Cope utilized Hawkins illustrations for his own purposes does not explain why Cope would have continued with the short-necked elasmosaur as late as The fact that he was sometimes sloppy with his work cannot provide an excuse. He must have thought he was right. Cope was a very busy person when he wrote the New Jersey Reptiles article. He was teaching at Haverford College, he was working hard at his own research, and he had a new young family. 28 Could these factors have caused him make the same mistake again? It seems unlikely. And even if this were the case, one cannot help asking again how could others have missed this? By 1869 we have two or perhaps even three incorrect elasmosaur reconstructions made by Edward Cope. Was it possible that Cope had not trusted the opinions of Joseph Leidy, O.C. Marsh and other scientists? Did he somehow really think that elasmosaurs had short necks and long tails? In looking at this question, it is useful to examine further examples of scientific literature and illustrations of plesiosaur fossils to which Cope could have had access prior to We have already noted that there was confusion among paleontologists about how to identify and to articulate plesiosaur vertebrae. There was also confusion about how to represent the appearance of a living plesiosaur. The evidence which Cope could have derived from contemporary scientific literature, even from popular science texts, might have fueled his own confusion about the E. platyurus. It might possibly account for some recalcitrance on his part in fully accepting others opinions. The operative term here is might account. The fact is that the available published data concerning plesiosaur should have given him ample information with which to make a correct restoration of the skeleton of the Elasmosaurus platyurus. As we noted the original publication of Plesiosaurus material was made by Conybeare and De la Beche in 1821, followed by a detailed description of P. dolichodeirus by Conybeare in Their Notice of the Discovery of a New Fossil Animal, Forming a Link Between the Ichthyosaurus and Crocodile, Together With General Remarks on the Osteology of the Ichthyosaurs, was read before the Geological Society of London, on April 6, 1821 and published in Volume 5 of the Transactions of the Geological Society of London in that same year. 30 This report created the name, Plesiosaurus, meaning almost a lizard. De la Beche and Conybeare wrote, That [name] of Plesiosaurus has been chosen, as expressing its near approach to the order Lacerta. 31 They continued their discussion of nomenclature for the new fossil by discussing how they perceived it to be a link between the crocodile and the ichthyosaur, writing of the plesiosaur that it appeared to be forming in its whole structure, a link between [the crocodile] and the Ichthyosaurus, hence it acquires a high and peculiar interest as affording a middle term of comparison This comment was followed by a lengthy description of the skeleton of Ichthyosaurus. Conybeare and De la Beche described the plesiosaur as follows: General character. A marine animal, intermediate in its structure between the Ichthyosaurus and Crocodile... Jaw and Head bones. Not as yet discovered. 33 Following this, they proceeded to a rather detailed description of the vertebrae as they felt that plesiosaur vertebrae compared with either those of a crocodile or an ichthyosaur. This description of vertebrae goes on for almost six pages. The specimen which they were describing in this paper was collected by a Colonel Birch and they believed that the specimen had at least 63 vertebrae. But as they were dug loosely out of the lias marle in which they lay, several appear to be missing and it comprises only the first 12 of the tail. 34 The vertebrae were described in descending order from cervical down the vertebral column towards the tail. Important for Cope s understanding of the articulation of vertebral column of a plesiosaur

13 ELASMOSAURUS 227 was that the authors note that the vertebrae change shape as one follows the vertebral column towards the tail end of the animal. At length the lower tubercle also disappears, and is swallowed up in a still longer prolongation of the margin of the annular suture; at the same time the stems by which the annular part was attached to the body expand their bases laterally, so as to form incipient transverse processes: these at first point downwards, and are close to the middle of the side, but as they become longer they rapidly ascend and point upwards; this character extends through the middle of the dorsal series; thence they again begin in incline downwards in the posterior dorsals, and in the last which bears a very short rib become very much reduced. 35 Various vertebrae were illustrated so as to show that their shapes were distinct. We read that Fig. 9A and fig. 9B exhibit the middle dorsal vertebrae with an without its annular part, to shew the form of the suture and its different countour from that in the cervical and first dorsal; the posterior dorsals differ only in having the transverse process less elevated, and did not therefore require representation. 36 Conybeare and De la Beche go on to further describe the lumbar and caudal vertebrae, again to demonstrate their distinctness from one another. The lumbar and caudal vertebrae appear to differ in form one another only in a less inflection of the lower margin, for the purpose shortly to be mentioned, they are however very distinct from all the others; they have no regular transverse processes, but, instead of them, two separate bones flattened at the extremities, and articulated into a socket near the upper part of the sides of the bodies of the vertebrae; their position is exactly horizontal.... The lower portion of these vertebrae on the posterior extremity is inflected so as to form two regular indentations, exactly as in the Crocodile, for the reception of the chevron-shaped bone beneath its tail 37 Put simply, Edward Cope should have known to place the skull of E. platyurus correctly because plesiosaur vertebral columns and the backbone of the elasmosaur have vertebrae which point in the direction of the tail, not of the skull. That is what Leidy told him. Had he thought that his colleague was mistaken, he would not have eventually rewritten his earlier remarks. But, even if he had thought his colleagues wrong, the above detailed descriptions by Conybeare and De la Beche of various vertebrae of Plesiosaurus would have provided him with even a more detailed and scientifically accurate way of looking at the vertebrae of E. platyurus and placing them into a correct articulation vis à vis the skull. This description was written in 1821 and published in an important journal the Transactions of the Geological Society of London, a journal which should have been available to him, possibly at the Academy of Natural Sciences and or at the Smithsonian Institution, or through one of his scientific colleagues in Philadelphia and Washington. We can demonstrate that members of the Academy of Natural Sciences had access to the Transactions of the Geological Society of London and that they were aware of the classification of Conybeare and De la Beche as early as An article in the Proceedings of the Academy of Natural Sciences of Philadelphia for February, 1825, cites a paper read by Richard Harlan on September 7, 1824, in which Harlan discusses some vertebrae found in New Jersey which were similar to Plesiosaurus vertebrae. He commented, The specimen under examination, however, differs from any species of the Plesiosaurus heretofore described... as is demonstrated in the following measurements of the vertebrae of that animal, described by Messrs. Conybeare and De la Beche. 38 Cope certainly would have seen the 1837 or 1841 editions of William Buckland s Geology and Mineralogy Considered With Reference to Natural Theology. Buckland s famous treatise on paleontology, which was published by William Pickering in 1837 in London, was also published in the United States in that year. It was indeed published in Philadelphia in 1837 by Carry, Lea and Blanchard. Lea and Blanchard of Philadelphia reissued the work in The considerable importance of this book s availability in Cope s home city lies in the illustrations which Buckland included of P. dolichodeirus. Buckland s engravings depicted several complete views of the fossil, based on Conybeare and De la Beche. Of this fossil Buckland wrote, The most anomalous of all the characters of P. dolichodeirus is the extraordinary extension of the neck, to a length almost equaling that of the body and the tail together. 39 In discussing the back and tail of the fossil Buckland noted, The tail, being comparatively short, could not have been used like the tail of fishes, as an instrument of rapid impulsion in a forward direction, but was probably employed more as a rudder Again, here was a source of information not only on the length of the neck of a plesiosaur, but quite detailed illustrations which showed the fossil s backbone quite clearly. Buckland s illustrations taken from Conybeare s restoration of P. dolichodeirus really helped to popularize Conybeare s fossil and the illustrations were reproduced in a number of subsequent scientific and popular texts.

14 228 J. P. DAVIDSON PLESIOSAUR VERTEBRAE UNDER STUDY Another early source of illustrations, data and scientific discussions of various species of Plesiosaurus was the work of Richard Owen ( ). Owen s earliest studies of plesiosaurs appeared in 1840 in his Report on British Fossil Reptiles, Part I. Report of the British Association for the Advancement of Science, 1839, pp This he followed with A Description of a Specimen of the Plesiosaurus macrocephalus, Conybeare, in the Transactions of the Geological Society of London, 1840, Vol. 5, pp This paper had been read before the Geological Society on April 4, 1838, so its contents actually represented work done prior to that in the previous paper. The specimen in question was owned by Viscount Cole. The paper was accompanied by beautiful engravings showing the specimen and some individual bones. Notably the spinal column of the animal as well as the skull are shown in the engravings. Concerning his study, Owen made a most telling comment, writing, I have endeavoured to devote to it an examination equivalent in care and detail to the admirable completeness of many of its parts, and have reason to hope that it will throw additional light on the Plesiosaurian modifications of the vertebrate skeleton Owen proceeded to describe his specimen in comparison with P. dolichodeirus and with a specimen studied by Thomas Hawkins and published in his Memoirs of Ichthyosauri and Plesiosauri. Owen noted that the vertebrae change shape as one progresses down the animal s vertebral column: For the costal appendages, which are generally anchylosed to the other vertebral elements in the cervical, sacral and caudal regions of the vertebral column of the warm-blooded vertebrate classes, retain their original separate condition throughout the vertebral column in the Plesiosaur and pass by such imperceptible gradations from one condition of physiological subserviency to another, that their nature cannot be mistaken when the entire series is viewed in a complete skeleton This comment was followed by analyses of the vertebrae with careful attention paid to the changes which might be noticed. Again and again, Owen pointed the way, as it were, down the vertebral column from the head to the tail. As the cervical vertebrae in the genus Plesiosaurus approach the dorsal, the inferior part of the costal articulation becomes smaller, and a corresponding increase of surface is afforded by the superior facet, which also gradually rises from the centrum to the neurapophyses and in the dorsal vertebrae stands boldly out as a true transverse process from the upper side of the base of each neurapophysis. At the sacral vertebrae, however, the transverse processes subside to the level of the neurapophyses; and, as the caudal vertebrae recede from the trunk, the articular surface, which, as in the neck, represents, or is in the situation of the transverse process, gradually descends and passes from the neurapophysis of the side of the centrum; but it is not divided by the longitudinal groove which characterizes the costal surface in the neck. 43 Dorsal vertebrae. These are characterized, agreeably with the previous definition, by the absence of articular surfaces for the ribs at the sides of the centrum, and by the development of a superior transverse process (exclusively supporting the ribs) from the base of each neurapophysis... The transverse processes progressive increase in length towards the middle of the trunk, and again diminish as they approach the tail: the bases of the neurapophyses from which they rise diminish in vertical extent in the same ration and leave a greater proportion of the centrum free from their embrace. 44 Owen s 1851 Monograph on the Fossil Reptilia of the Cretaceous Formations continued his contributions to the state of scientific knowledge of the vertebral characteristics of the plesiosaurs. This treatise provided the reader with carefully drawn, very detailed illustrations of various types of plesiosaur vertebrae from several specimens. The vertebrae constitute most of the discussion and illustrations. Owen acknowledged Buckland by reproducing his famous quote of the rather griffon-like qualities of the plesiosaur, which we noted earlier in this paper and called plesiosaurs a remarkable extinct genus (with its) length and flexibility of the neck Owen s illustrations and his very complete descriptions of individual bones in the vertebral column would have provided any reader with much valuable and definitive data as to what a specific plesiosaur vertebra should have looked like. In other words, here were careful information on how to arrange the vertebral column of a plesiosaur should one find one in a disarticulated state. I will cite a brief example of his thoroughness. Writing about a specimen which he named Plesiosaurus pachyomus, Owen gives the following information concerning a cervical vertebra and then a caudal vertebra: In the vertebra which I take to be the penultimate or antipenultimate cervical, the upper half of the costal surface has passed upon the base of the neurapophysis, and, for what remains upon the centrum, as at pl. fig, 5 T.XX, we