As the nineteenth century drew to a close, the reality of evolution became firmly ensconced within the scientific community. Darwin deserves credit for the overused idea of a paradigm shift—evolution was a scientific “fact,” but not so for Darwin’s theory of natural selection. This intellectual retreat happened for a number of reasons. A common perception suggests that when Gregor Mendel’s (1822–1884) work “Versuche über Pflanzen-Hybriden” (Experiments in Plant Hybridization, 1866) on particulate inheritance was rediscovered at the turn of the twentieth century, its sole importance came as a mechanism for inheritance. In the hands of scientists such as Hugo de Vries (1848–1935), ideas of genetic mutations provided not only the source of new genetic material but also the cause of proposed rapid evolutionary change, which for some biologists largely replaced Darwin’s natural selection. This perception, while in part true, does not provide a complete answer.
Retreat from Darwin’s natural selection began even before the rediscovery of Mendel and the birth of modern genetics. For example, one of the coauthors of natural selection, Alfred Russel Wallace, increasingly viewed some aspects of human faculties as mostly independent of natural selection, an exclusion likely caused in part by his later misguided acceptance of spiritualism. More important, Darwin could never provide an adequate hypothesis of inheritance or the source of variation on which his natural selection was to act. Darwin’s self-described “much-abused” theory of Pangenesis was roundly dismissed.
Cope’s Neo-Lamarckian Trees
These shifts of fortune had some effect on how evolution was represented in trees, but it was not because of work in genetics; rather, it came from the still emerging field of vertebrate paleontology, notably in the United States, and the ideas of evolutionary mechanisms that accompanied them. In 1868, only nine years after the publication of On the Origin of Species, an American vertebrate paleontologist was advocating a different mechanism for evolution in his paper “On the Origin of Genera.” The author, Edward Drinker Cope (1840–1897), became well known outside of science because of one of the nastiest, most protracted confrontations in all of the history of science, the so-called bone wars that raged between Cope at the Academy of Natural Sciences in Philadelphia and Othniel Charles Marsh (1831–1899) at the Peabody Museum of Natural History at Yale University in New Haven. To make matters worse, Marsh was by and large a supporter of Darwinism, whereas Cope was not.
In his paper, the twenty-eight-year-old Cope (1868), another young Turk but never in Darwin’s sphere, writes:
That a descent, with modifications, has progressed from the beginning of the creation, is exceedingly probable. The best enumerations of facts and arguments in its favor are those of Darwin, as given in his various important works, The Origin of Species, etc. There are, however, some views respecting the laws of development on which he does not dwell, and which it is proposed here to point out. In the first place, it is an undoubted fact that the origin of genera is a more distinct subject from the origin of species than has been supposed. (243)
This provides a prime example of the acceptance of evolution, but not Darwin’s version. Cope argued that genera, families, and other groups arose more through the retardation and acceleration in the process of development of an individual and that natural selection was subservient to this process. Whereas Cope’s intuition about developmental timing would fit in today’s understanding of its importance in the evolutionary process, his views on how this would affect higher categorical levels—genera, families, and so on—were completely misguided. Although not universal, the consensus is that evolution operates at and below the species. In and of itself, Cope’s thesis was not Lamarckian, but he argued that the environment acted on the individual to cause these developmental changes, which were then passed to the next generation. Much of Cope’s argumentation came from what he as well as colleagues, notably in the United States, believed they saw in the fossil record of vertebrates. This was termed neo-Lamarckism in 1884 by a contemporary and sometime colleague of Cope, Alpheus Hyatt (1838–1902) (Regal 2002 and references therein). The record for Cope, Hyatt, and others showed an orderly, almost rectilinear pattern of progression that did not accord well with Darwin’s natural selection.
Cope produced nearly 1,500 scientific publications in his relatively short life, some a few lines long and others massive tomes, the most impressive of which is 1,009 pages of text plus almost as many additional pages, or at least an equal thickness of plates, many of them foldouts, published in 1884. This is a single, hardbound 12- by 9-inch (30 by 24 cm) quarto volume measuring an astounding 4¾ inches (12 cm) thick; no wonder it has earned the epithet “Cope’s Bible.” Many of the illustrations of specimens produced for Cope’s publications were quite detailed and elaborate, such as in this ponderous 1884 tome, but the same cannot be said of his phylogenies, which were as simple as Haeckel’s were grandiose. In this large volume, Cope provides only six tree-like diagrams, one for turtles and the remainder for mammals. The diagrams do not show consistency in orientation; some open upward, whereas others open downward. He identifies them by various names—tables, phylogenies, lines of descent. Their appearances, however, clearly conform to a general pattern, as seen by the three shown in figure 6.1A–C. The trees use Cope’s terminology, and although there are many incorrect systematic aspects to them, they are not of concern here. Of greater interest are the regular spacing of names and the relative position of the names based on his perception of their grade of evolution. For example, in the downwardly opening tree (see figure 6.1A), among living members of his “Taxeopoda,” the Perissodactyla (odd-toed ungulates) and Artiodactyla (even-toed ungulates) are placed farther along or more terminally, implying more evolutionary advancement compared with Proboscidea (elephants) and Hyracoidea (hyraxes). Similarly, the upwardly opening tree (see figure 6.1B) shows Equidae (horses) farther advanced than Rhinoceratidae (rhinos), which in turn occur farther along than Tapiridae (tapirs). Figure 6.1C shows successive steps in the lines leading to Felidae (cats) and Canidae (dogs). The repeated names of these two families indicate advanced living members along with more primitive extinct members of each family.
In his book The Primary Factors of Organic Evolution (1896), published the year before his death, Cope presents seven simple trees, much like those just discussed, but two of them differ in that they include geologic time. One offers a schematic phylogeny of plants (see figure 6.1E), and the second gives his views on mammalian evolution (see figure 6.1F). In the plant phylogeny, groups such as Cope’s Protophyta (an obsolete term for a mix of groups sometimes including, for example, bacteria and blue-green algae) appear in the Ordovician period with successive appearances of eight other major plant groups, possibly coming from a common ancestor but then paralleling one another during their evolutionary and geological history. He argues in the pages preceding the plant phylogeny that this parallel evolution in various plant groups demonstrates what he terms “successional development,” with changes in reproductive mode being particularly important. He sees two components for this, a trend going “from the simple to the complex” and a trend leading “from the generalized to the specialized.” Cope used quotation marks for the three phrases, as shown here. He quickly notes that these concepts are not identical. Further, in the latter trend, he notes that both progressive and retrogressive processes occur, which relates to his earlier noted ideas of the importance of embryological development in evolution. For the mammalian phylogeny, separate from greatly out-of-date systematic interpretations, his placement of various extant groups of mammals such as Chiroptera (bats) and Glires (rabbits and rodents) well back into the geological past is somewhat odd, until the pages preceding and following the phylogeny are read. Although not explicitly within the text, he clearly attempts to indicate when he believes these groups first appeared.
FIGURE 6.1 Edward Cope’s evolutionary trees for (A) “Taxeopoda,” including Perissodactyla (odd-toed ungulates), Artiodactyla (even-toed ungulates), Proboscidea (elephants), and Hyracoidea (hyraxes); (B) Equidae (horses), Rhinoceratidae (rhinos), and Tapiridae (tapirs); and (C) successive steps in the lines leading to Felidae (cats) and Canidae (dogs), from The Vertebrata of the Tertiary Formation of the West (1884). Cope summarized his ideas on development and evolution in (D) a very simple diagram, from “The Method of Creation of Organic Forms” (1871), and included geologic time in his schematic phylogenies of (E) plants and (F) mammals, from The Primary Factors of Organic Evolution (1896).
Examining Cope’s theorizing on evolution provides some basis as to why he produced these as well as other trees as he did. In his 1868 paper, Cope did have some diagrams, but except for possibly one, none comes close to evoking a tree. In his slightly later paper “The Method of Creation of Organic Forms” (1871), he encapsulates views on development and evolution in a simple diagram at the end of the text (see figure 6.1D) that was repeated twice in his book-length treatment of his theory of evolution, The Origin of the Fittest (1887).
In the paragraph before the description of this small figure, Cope (1871) praises Darwin for demonstrating
the origin of varieties in animals and plants, either in the domesticated or wild states…[and that] species have been derived from other species among domesticated animals…[and that inferences] that other species, whose origin has not been observed, have also descended from common parents…[is] to be justified; but when from this basis evolution of divisions defined by important structural characters, as genera, orders, classes, etc., is inferred, I believe that we do not know enough of the uniformity of nature’s processes in the premises to enable us to regard this kind of proof as conclusive. (231)
He then goes on to explain the diagram on the same page:
In A we have four species whose growth attains a given point, a certain number of stages having been passed prior to its termination or maturity. In B we have another series of four (the number a matter of no importance), which, during the period of growth, cannot be distinguished by any common, i.e., generic character, from the individuals of group A, but whose growth has only attained to a point short of that reached by those of group A at maturity. Here we have a parallelism, but no true evidence of descent. But if we now find a set of individuals belonging to one species, or still better, the individuals of a single brood, and therefore held to have had a common origin or parentage, which present differences among themselves of the character in question, we have gained a point. We know in this case that the individuals, a, have attained to the completeness of character presented by group A, while others, b, of the same parentage have only attained to the structure of those of group B. It is perfectly obvious that the individuals of the first part of the family have grown further, and, therefore, in one sense faster, than those of group b. If the parents were like the individuals of the more completely grown, then the offspring which did not attain that completeness may be said to have been retarded in their development. If, on the other hand, the parents were like those less fully grown, then the offspring which have added something, have been accelerated in their development.
Osborn’s Aristogenetic Trees and the Rise of Neocreationism
Cope possessed a grasp of the importance of development in evolution, but his neo-Lamarckian views on how variation in development might arise never found traction among other biologists and paleontologists, with a few notable exceptions—in particular, one aspiring young vertebrate paleontologist, Henry Fairfield Osborn (1857–1935), who rose to considerable prominence in early-twentieth-century American science. Osborn would begin with Cope’s neo-Lamarckian ideas but carry them even further and in the process produce some of the most elaborate tree-like diagrams of the early twentieth century.
Osborn studied anatomy at a New York hospital and embryology with Thomas Henry Huxley in England and received his doctorate in paleontology from Princeton. He also received his undergraduate degree at Princeton under the tutelage of Cope; there, he became fast friends with William Berryman Scott (1858–1947) and Francis Speir Jr. (1856–1925). With newly minted undergraduate degrees, the three undertook an expedition to fossil beds in the American West in 1877, which was written up and published (Osborn, Scott, and Speir 1878). Scott, like Osborn, went on to become a vertebrate paleontologist, whereas Speir earned a law degree in New York; but all three remained friends, and certainly Osborn and Scott early on influenced each other in their respective careers (Regal 2002). Scott even dedicated a later edition of A History of Land Mammals in the Western Hemisphere, first published in 1913, “to the memory of his friends” Speir and Osborn; Speir named one of his children Henry Fairfield Osborn Speir (Chamberlain 1900).
Cope clearly influenced Osborn and Scott. This can be seen in some of the phylogenies in their earlier works. In 1899, Scott published a paper, “The Selenodont Artiodactyls of the Unita Eocene,” on the even-toed ungulates such as cows and deer that have crescent moon–shaped ridges (selens) on their teeth, with which they grind plant material much as tools used to file wood or metal. He provides a number of figures of the fossils and near the end one very simple but understandable phylogeny of the mammals discussed in his paper (figure 6.2A). Scott notes that this “subjoined table exhibits the relationships of the various genera as conjectured in the preceding pages.” Other than the names being placed vertically, the simplicity in form matches well the trees of Cope (see figure 6.1). Even though similar to Cope’s trees, Scott’s style appears clearer in intent, not requiring explanation for clarity.
Fourteen years later, Scott (1913) presents a figure he titles “Evolution of the Proboscidea” (see figure 6.2B). He indicates modification after an earlier figure (Lull 1908) (see figure 6.2C). Scott’s version adds lower molars on the left and outlined heads on the right. These are earlier renditions of this sort of representation. Scott shows a progression in the evolution of elephant heads and teeth. Although it is not strictly a scala naturae, it certainly approaches one. The tree shown in figure 6.2D comes from Osborn’s published version of “The Rise of the Mammalia in North America” (1893), an address he presented to the American Association for the Advancement of Science the same year. He calls this very broad-brush phylogeny a “Hypothetical Phylogeny of the North American Mammalia.” It is placed within the context of geologic time. Osborn makes some attempt to show the quality of the record and taxonomic richness by the thickness of the black columns. Recall that Edward Hitchcock (1840) and, even more similarly, Louis Agassiz (1844) had attempted a similar theme of using thicknesses of branches, albeit in a creationist context, to show relative taxonomic richness (see figures 3.10 and 3.11).
FIGURE 6.2 William Scott’s (A) stratigraphically arranged phylogeny, from “The Selenodont Artiodactyls of the Unita Eocene” (1899), and (B) “Evolution of the Proboscidea,” including lower molars to the left of the skulls and outlined heads to the right, from A History of Land Mammals in the Western Hemisphere (1913), based on (C) a similar figure by Richard Lull, from “The Evolution of the Elephant” (1908); (D) Henry Osborn’s “Hypothetical Phylogeny of the North American Mammalia,” placed within the context of geologic time, with some attempt to show the quality of the record and taxonomic richness by the thickness of the black columns, from “The Rise of the Mammalia in North America” (1893).
Later Osborn used other designs, such as phylogenies radiating within concentric circles, that echo in later figures. Based on his developing ideas of evolution, he often incorporated the idea of a concentric radiation with a series of radiating lines showing small outline figures of the animals in question. The popularity of this sort of representation has waxed and waned throughout the twentieth and early twenty-first centuries, notably in popular literature, where the public always remains fascinated with knowing what the animals looked like. A tree resplendent with many small vignettes of animals, especially those of long-extinct species, still holds great interest for the public. Two classic examples of Osborn’s (1936) trees come from the first volume of his two-volume monograph on elephants and their relatives. Figure 6.3A is the “Phylogeny of Moeritherioidea, Deinotherioidea, and Mastodontoidea,” all of which are extinct relatives of elephants, and the flame-like tree in figure 6.3B is titled “Final Diagram (1935) Showing the Adaptive Radiation of the Forty-three Generic Phyla of the Proboscidea.” Although the phylogeny in figure 6.3B does not include small outlines of animals, figure 6.3A provides classic Osborn elements: it radiates from an ancestral point with a series of separate lineages, and each has a vignette outline of included taxa.
Recall that Osborn was influenced by Cope’s ideas of neo-Lamarckian evolution, and he supplemented Darwin’s idea of natural selection with what he called aristogenesis, a scala naturae–like, purpose-driven concept of evolution that can be seen in many of his phylogenies, such as the two in figure 6.3A and B. Such figures are often dominated by straight-line evolutionary change (see also figure 1.7). No matter what might be the stated intent, Osborn’s views, as seen in many of his phylogenies, show much of evolutionary modification occurring along a single line of change and descent, hence the clear resemblance to the older scala naturae, or great chain of being. At the time, and for a number of years to come, his ideas had a considerable impact because from 1908 until 1933 Osborn directed the American Museum of Natural History, still one of the premier natural history museums in the world, but at that time an even more powerful institution in American science, in large measure because of Osborn. Osborn’s wealth and connections helped amass power and prestige for his museum, ensuring his place in the history of that institution, at the time a very influential purveyor of scientific ideas. Today, his reputation as an excellent administrator remains but not so his scientific reputation, in part because of his support of teleologically tainted neo-Lamarckian ideas of evolution and in part because of his racially driven ideas of eugenics (Rainger 1991; Regal 2002).
FIGURE 6.3 Osborn’s (A) “Phylogeny of Moeritherioidea, Deinotherioidea, and Mastodontoidea,” all of which are extinct relatives of elephants, showing radiation from an ancestral point with a series of separate lineages, each with a little vignette outline of included taxa, and (B) “Final Diagram (1935) Showing the Adaptive Radiation of the Forty-three Generic Phyla of the Proboscidea,” from Proboscidea (1936); (C) the phylogeny in George Hunter’s A Civic Biology (1914) placed into evidence in the “Scopes Monkey Trial” (1925), as well as being mentioned in the play Inherit the Wind (1955) and the film of the same name (1960), based on the trial. ([A] and [B] reproduced with permission of the American Museum of Natural History)
Osborn and, before him, Cope were two of the most dominant scientists in influencing tree building in the late nineteenth century and the first quarter of the twentieth, but it was a trial including an evolutionary tree that truly captured the attention of the American public, later spawning a play, a film nominated for four Academy Awards, and a variety of books. The trial is best described by Edward Larson in Summer for the Gods (1997), which details the machinations leading up to and through the “Scopes Monkey Trial,” which took place in Dayton, Tennessee, in 1925. The Tennessee state legislature had passed the Butler Act, which made it “unlawful for any teacher…to teach any theory that denies the Story of the Divine Creation of man as taught in the Bible, and to teach instead that man has descended from a lower order of animals” (Tennessee House Bill 185, 1925). As Larson relates it, the high-school teacher John T. Scopes reluctantly agreed to test the law and in so doing helped put Dayton on the map, as well as helping a fledgling American Civil Liberties Union defend one of its first cases.
A state-approved textbook, George William Hunter’s A Civic Biology (1914), became exhibit number one in the case (the Bible was number two), although it really served only as a straw man for the case. It is also mentioned in Jerome Lawrence and Robert Edwin Lee’s play Inherit the Wind (1955), based quite closely on the trial transcript, as well as in Scopes’s memoir (Scopes and Presley 1967). In a number of places, Hunter unquestionably praised and supported Darwin’s evolutionary ideas, but he also espoused racist and eugenic views. On page 193 of Hunter’s text is an orthogenetic-like diagram of horse evolution after one by the paleontologist William Diller Matthew (incorrectly called Matthews), and on page 194 is what Hunter terms an evolutionary tree (see figure 6.3C). Each branch of this quite generalized, scala naturae–like tree terminates in a circle, the circles varying in size according to the relative number of species they include, which is given within or next to each circle. The text speaks only obliquely of human evolution, stating that man at first was “little better than one of the lower animals” (196), but the tree presents nothing about human evolution or human descent from lower forms. This did not stop prosecutors in the trial from attempting to portray the book and, by implication, the evolutionary tree as showing human descent from lower forms, so that Scopes’s use of the text in class could be used to find him guilty of having violated the Butler Act. The trial transcript (1990:123) specifically refers to the evolutionary tree in a question by the prosecution regarding whether Scopes discussed it in his class.
Probably wanting to deflect any further scandal for his textbook, in his revised edition, which appeared only one year after the trial, Hunter (1926) expunged not only the five-page section on evolution but also any mention of the word, as well as the horse evolution chart and the evolutionary tree. This recalls Hitchcock’s (Hitchcock and Hitchcock 1860) deletion of his nonevolutionary, tree-like paleontological chart (see figure 3.11) when the tree became an evolutionary icon following the publication of On the Origin of Species in 1859. The back-peddling and kowtowing to the religious right signaled a change that gained momentum in the middle of the twentieth century. The trial seemed to damp creationism for a few decades, but it returned with a vengeance in the mid-twentieth century, especially in the Unites States, along with its later spawn “intelligent design.” This time, unlike Hitchcock and Agassiz in the nineteenth century, who honestly attempted to wedge biblical creation into a scientific mold, the newcomers promoted a pseudoscience that still festers like an intellectual canker.
The New Synthesis and Neo-Darwinian Trees
Cope and Osborn influenced how the next academic generations represented biological order, but the same cannot be said of their neo-Lamarckian evolutionary ideas, which soon faded. Many trees were produced in the first two-thirds of the twentieth century, so the task of choosing among them becomes daunting, but three academic descendants and associates of Osborn in particular typify what came next.
William King Gregory (1876–1970) obtained his doctorate from Columbia University and assisted Osborn in his research and in the preparation of his publications. Gregory concentrated on the evolution and comparative anatomy of vertebrates, especially mammals. Two other persons of note with ties to the American Museum of Natural History who effectively used phylogenetic trees came in the academic generation following Gregory: Alfred Sherwood Romer (1894–1973) and George Gaylord Simpson (1902–1984). Romer was a student of Gregory’s, earning his doctorate at Columbia and then spending much of his career at Harvard, whereas Simpson, after obtaining his doctorate from Yale, spent his career in part at the American Museum of Natural Museum and at Harvard. Romer and, more so, Gregory produced copious phylogenetic trees. Romer is best known for those in his text Vertebrate Paleontology, which went through three editions and a number of printings between 1933 and 1971; Gregory compiled many such figures in Evolution Emerging (1951). Simpson did not produce nearly as many phylogenetic trees of specific groups as did Gregory and Romer, but he was one of the first and still most effective users of phylogenetic trees in the explanation of evolutionary and ecological theory and process.
Examining the contributions of Gregory, Romer, and Simpson in visualizing biological order, we start with the senior member of the group, William King Gregory. Gregory’s career began as an assistant to Osborn near the turn of the twentieth century. Osborn worked all his assistants very hard, but none more than Gregory. Osborn’s two best-known works—a two-volume study of titanotheres, distant relations of rhinos and horses (1929), and the even better-known two-volume monograph on proboscideans (1936, 1942)—can in large measure be traced to the work of Gregory. Gregory also contributed greatly to the writing of Osborn’s Age of Mammals in Europe, Asia and North America (1910; Colbert 1975). If Gregory ever supported the Cope–Osborn views of evolution, by the time he produced his magnum opus in 1951, Evolution Emerging, based on a half century of work, he largely supported the newer views set forth in the Modern Synthesis, which unified Darwinian natural selection and Mendelian genetics: “Perhaps the greater part of the causes of evolution lies deep within the nature of the hereditary mechanism and of its reactions to Natural and Artificial Selection; but…the leading principles of evolution…appear to operate all levels of organization, from inorganic material…along the diverging paths of descent with modifications” (1:552). Simpson called these two volumes “both the chef d’oeuvre and the swan song of a genius” (quoted in Colbert 1975).
Out Haeckeling Haeckel
Hands down, Gregory was the most prolific purveyor of phylogenetic trees, even bettering the number produced by Ernst Haeckel (Pietsch 2012). The first volume of Evolution Emerging (1951) presents 736 pages of text on all animals but emphasizing vertebrates; while important and worth the read, it is secondary to the second volume, whose 1,013 pages feature wondrous illustrations of all manner of animals. Except for a very imaginative, and dare I say Haeckelian, geologically based tree frontispiece titled “Procession of the Vertebrates” (figure 6.4), the first volume lacks illustrations. No images of animals are shown in the diagram, but the spindle-like aspect gives the sense of waxing and waning of various lineages, harking back to Agassiz’s creationist diagram of fish relationships produced over one hundred years earlier (see figure 3.10). Instead of the tree being anchored by a geologic time scale on one side, the vertebrate procession emerges from the depths of a Grand Canyon–like chasm; thus the appropriate name of the book, Evolution Emerging.
FIGURE 6.4 Frontispiece of the first volume of William Gregory’s Evolution Emerging (1951). (Reproduced with permission of the American Museum of Natural History)
The second volume includes almost ninety trees of various forms. In addition, Gregory presents diagrams showing a succession of species or parts of their anatomies in an almost scala naturae fashion. We can only sample a few to provide a sense of what he attempted to show. Even though Gregory was not a fan of Osborn’s views on evolution, the influence of Osborn comes through in a number of Gregory’s trees. Excluding the trees that he presents from other sources, a majority of his trees show whole organisms (extinct and extant), skulls of these organisms, or other parts of the organisms. In most of the figures, the illustration of the animal or its parts is large enough to see the anatomical aspects that Gregory wished to show his readers. The actual tree portion is often secondary, but he nonetheless was attempting to depict what he called, and we still call, adaptive evolution or radiation. This came from Gregory’s insistence on the importance of understanding the integration of the various parts of an animal’s body. Above all, Gregory was one of the twentieth century’s finest comparative anatomists. He understood what many others did not—that parts of an organism do not evolve in lock step. Rather, different parts evolve in fits and starts; thus the creationist cry of how could a bird fly with only half a wing becomes patently absurd. This idea of different rates of evolutionary change in different parts of an animal’s body is now known as mosaic evolution, with the obvious allusion to the parts of a mosaic creating the whole. Gregory recognized this, calling it his “palimpsest theory” in which later specializations overlie earlier ones, just as the term is used in the original meaning of the overwriting of older by younger texts. With this in mind, Gregory’s considerable interest in presenting phylogenies of only parts of an organism becomes obvious. Many examples occur in the second volume of Evolution Emerging, such as the evolution of marsupials in general using whole-body outlines (figure 6.5A) and, in particular, the evolution of the hind feet in Australian marsupials (see figure 6.5B) and of the occlusal, or biting, surface of a typical lower (see figure 6.5C) and upper (see figure 6.5D) molar. Gregory is attempting to show that different parts of an organism evolve at different rates and must be considered separately, but then he puts them back together as a whole to comprehend the adaptive radiation of that group.
Gregory also often employs a different sort of diagram. At first glance, it appears to be some form of scala naturae, but it is not. Figure 6.6A shows a group of skulls suggesting the march from primitive placental mammal forms to advanced elephants; they are very similar to figure 6.2B and C, which Scott (1913) and Richard Lull (1908), respectively, called evolutionary diagrams of elephants. Instead, Gregory calls this a “structural series, showing recessions of nasals”—meaning retraction of the nasal opening backward and upward onto the skull. Note the vertical line demarcating the front and back halves of the skull. Unlike earlier authors such as Lull, Scott, and Osborn, Gregory does not imply or state an evolutionary succession but rather the likely structural changes that occurred as elephants evolved. Another example shows the “general relations of the temporal and masseter muscles…from fish to man” (see figure 6.6B). The arrows track from the shark, then over, and finally down to humans. Once again, although Gregory would no doubt regard a shark as primitive relative to a human, the intent in his pseudo-ladder shows changes in the muscles of the jaw as evolutionary history unfolded, much as in his palimpsest style, but with no implication of an evolutionary series.
The Master of the Textbook Tree
Next follows Alfred Sherwood Romer, a student of Gregory’s at Columbia near the end of the second decade of the twentieth century. Romer learned comparative vertebrate anatomy from the master, which obviously influenced his work throughout his career in the study of early tetrapod evolution, including the lineages that would lead to reptiles and the others that eventually led to mammals (Colbert 1982). As with his mentor, Romer was a prolific producer of phylogenies. Among his numerous papers and books, the most popular were his textbooks The Vertebrate Body (later editions done with Thomas Parsons) and Vertebrate Paleontology, used by many students, myself included. Both have about fifteen trees of various sorts; most in The Vertebrate Body consist of lightly sketched tree outlines on which Romer placed small drawings of the animals, with no geologic time shown, whereas those in Vertebrate Paleontology consist of spindle diagrams with no animals shown but within the context of geologic time.
FIGURE 6.5 Gregory’s four phylogenies of marsupials, showing (A) the evolution of marsupials in general using whole-body outlines, (B) the evolution of the hind feet of Australian marsupials in particular, and the occlusal, or biting, surface of a typical (C) lower and (D) upper molar, from Evolution Emerging (1951). (Reproduced with permission of the American Museum of Natural History)
FIGURE 6.6 Gregory’s “structural series, showing [A] recessions of nasals” and (B) “general relations of the temporal and masseter muscles…from fish to man,” from Evolution Emerging (1951). (Reproduced with permission of the American Museum of Natural History)
Figure 6.7A, from the fifth edition of The Vertebrate Body (1977), shows the “Diagrammatic Family Tree” of eutherian (placental) mammals, with an animal sketch representing each lineage. Figure 6.7B and C show the “Chronologic Distribution of Placental Mammals” from, respectively, the first edition (1933) and the third edition (1971 printing) of Vertebrate Paleontology. The first tree shows almost no extinct forms, whereas in the other two trees a number of extinct lineages appear. Although the latter two do vary in some of the relationships represented, they otherwise look rather similar except for differences in shading. The distinction between the first tree and the other two trees relates to the intended audience. For The Vertebrate Body, the audience was students of comparative anatomy interested in living forms, whereas for Vertebrate Paleontology, the audience comprised students of that discipline and hence their interest in more extinct forms.
FIGURE 6.7 (A) Alfred Romer and Thomas Parsons’s “Diagrammatic Family Tree” of placental mammals, redrawn from The Vertebrate Body (1977); Romer’s (B) “Chronologic Distribution of Placental Mammals,” from Vertebrate Paleontology (1933), and (C) “Chronologic Distribution of Placental Mammals,” from Vertebrate Paleontology (1971). ([B] and [C] reproduced from Romer 1933, 1971; used with permission of the University of Chicago Press)
Other trees showing comparable groups in the 1933 and 1971 editions differ markedly, notably those for rodents. In 1933, only ten superfamily or family-level groups appear (as well as rabbits, the duplicidentates) (figure 6.8A), compared with forty-four similarly named lineages in 1971 (see figure 6.8B). This change came from increased knowledge of rodents as well as from Romer’s desire simply to show more detail. Although Romerian in style and influential for the work of others, none of these trees broke new ground in tree construction, yet one of the smallest and simplest trees, the same in all three editions of Vertebrate Paleontology beginning with the first edition in 1933, provides some idea of how Romer and others were grappling with the issues of relating tree construction with classification. In this tree, Romer attempts to clarify how the concepts and differences between “vertical” and “horizontal” classification appear in a phylogenetic tree, portending the revolution in biological systematic thinking to soon come—phylogenetic systematics or cladistics (see figure 6.8C). In both 1933 and 1971, Romer writes:
Two types of classification are possible—“vertical” and “horizontal.”…Under the first system each family or other unit comprises all members of a known line from its first beginnings to its end or to modem times; the cleavage between lines is carried down to the very base of the evolutionary tree. But when, for example, forms are discovered seemingly ancestral to two distinct families or closely related to both, their inclusion in one or the other seems improper. Under such circumstances the best solution seems to be a “horizontal” cleavage, the erection of a stem group, including the base from which the long-lived later families may been derived. (1933:7, 1971:4)
In the 1971 edition, Romer included this additional paragraph:
“Horizontal” or “vertical” classifications of the sort just described are both acceptable. To be avoided, however, is another type of “horizontal” classification. Let us assume that of the forms shown in the diagram, A and B, although with separate pedigrees, both evolved in each the same direction. If, after setting up (as at the right of the diagram) a basal family C, a scientist decides, because of similarities, to unite A and B in a common family, he has created a family that is not merely “horizontal” but polyphyletic—that is, he has included forms of different origin in a common assemblage. Sometimes, because of inadequate knowledge, this has been done, inadvertently, by paleontologists, but such unnatural grouping should be avoided. (4)
FIGURE 6.8 Romer’s evolutionary trees for rodents (and rabbits), going from (A) only ten superfamily or family-level groups in 1933 to (B) forty-four similarly named lineages in 1971, and (C) diagram clarifying how the concepts and differences between “vertical” and “horizontal” classification appear in a phylogenetic tree, from Vertebrate Paleontology (1933, 1971). (Reproduced with permission of the University of Chicago Press)
Biological classification provides perennial problems, but the issue became particularly acute beginning in the mid-twentieth century. Romer voices this concern with an obvious expansion of these issues in the final edition of Vertebrate Paleontology.
A Synthesizer par Excellence
Recall that Darwin’s single foldout diagram in On the Origin of Species (1859) did not attempt to explain a specific set of evolutionary relationships for a specific group of species but tried to express visually how he thought his theory of natural selection might explain the evolutionary process over great expanses of geologic time. In the first half of the twentieth century, it was George Gaylord Simpson more than anyone else who followed in Darwin’s tradition of using phylogenetic trees to deal with evolutionary processes and theory, including issues of taxonomy such as those raised by Romer.
Along with a number of mostly British and American scientists, Simpson helped to forge the Modern Synthesis of evolutionary theory, also know somewhat inaccurately as neo-Darwinism. Scientists from the Darwinian camp (who were often natural historians) and Mendelian geneticists remained at loggerheads until population geneticists and population biologists in the 1930s and 1940s demonstrated through statistical studies of plant and animal populations that Darwin’s natural selection winnowed the variation of Mendelian genetics. Darwin’s natural selection proved instrumental in sorting variations, but Mendelian genetics provided these variations that were altered by mutations. Simpson argued that the Modern Synthesis could be demonstrated through studies of the fossil record, and many tree-like figures centered on explaining this process. Simpson used tree-like diagrams to explain how the evolutionary processes unfolded, sometimes with hypothetical examples and at other times using case studies from the fossil and modern records. The examples explored here come from one scientific paper and his more popular books stretching through the middle third of the twentieth century.
In 1937, Simpson published an article ahead of its time, “Patterns of Phyletic Evolution,” in which he examined several knotty issues especially in the light of the emerging Modern Synthesis of evolution. How does one interpret evolutionarily and taxonomically a “sample of similar but varying specimens extending over an appreciable span of geologic time” (1937a:307)? This hypothetical sample is given in figure 6.9A, along with Simpson’s three interpretations gathered by his biographer Leo Laporte (2000). Simpson provides three scenarios: the first “representing two distinct and parallel phyla, with the type of frequency distribution at any one time, such as T, that would make this interpretation probable” (308) (see figure 6.9B); the second “representing a single changing, with the type of frequency distribution at any one time, such as T, that would make this interpretation probable” (309) (see figure 6.9C); and the third “interpreted as an ancestral stock giving rise to two divergent phyla with the types of frequency distributions at successive times T1, T2, T3, T4, that would make this interpretation probable” (311) (see figure 6.9D). Of course, Simpson uses the word “phylum” here to indicate a general lineage or branch rather than the formal taxonomic rank of phylum. Later in his career, Simpson credited this paper with his abandonment of typological thinking for a more statistically biometrical approach to evolutionary theory and taxonomic issues (Laporte 2000). In the same year, Simpson followed with papers that examined in a statistical framework Paleocene and Eocene fossil mammals, most notably his Crazy Mountain mammal monograph (1937a), one of the earliest uses of biometrics looking at fossil populations. Typological thinking was giving way to populational thinking in evolutionary studies, but except in the hands of scientists such as Simpson, tree-like diagrams tended not to embrace such thinking.
FIGURE 6.9 George Gaylord Simpson’s (A) sample and (B–D) scenarios for the evolution of lineages over geologic time, from “Patterns of Phyletic Evolution” (1937). (Reproduced with permission of the Geological Society of America)
Simpson’s method of explaining complex evolutionary theory frequently employed simple diagrams, including tree-like figures (Laporte 2000). He was a master at the use of these almost chalkboard-like sketches. There are many from which to choose, but a few shown here exemplify some key issues, starting with arguably his most influential and at the time most controversial book, Tempo and Mode in Evolution, first published in 1944. In this volume, we see his visual musings on how he perceives evolution. Because he was one of the architects of the Modern Synthesis of evolution, which combined Darwin’s natural selection with Mendel’s genetics, his figures must be viewed through this prism.
In 1972, Niles Eldredge and Stephen Jay Gould proposed the theory of “punctuated equilibrium,” which argues that most evolutionary change (usually meaning morphological change in fossils) occurs during rapid bursts of speciation followed by long periods of stasis when rates of evolutionary change and speciation are much lower (figure 6.10A). Simpson proposed a precursor to punctuated equilibrium in Tempo and Mode. Figure 6.10B shows his hypothetical phylogeny of this process. Simpson (1944) called this “explosive” (his quotation marks) evolution in which there are “multiple quantum steps into varied adaptive zones, followed by extinction of unstable intermediate types and phyletic evolution in each zone” (213). By phyletic evolution, Simpson meant gradual change mostly in a single lineage, or what we now call anagenesis. Punctuated equilibrium and “explosive” evolution models arguably differ in that the former addresses what might be called normal evolution, whereas the latter addresses the origin and rapid radiation of major groups in a short period of time. We do know more clearly now that evolution occurs at and below the species level, whether it results in so-called normal evolution or a major radiation over longer intervals of time. Possibly this is a distinction without a difference; nevertheless, Simpson’s tree motif explores this concept.
A final set of Simpson’s trees showing the adaptive radiation of primates tells an interesting story of changing ideas, specifically for Simpson and generally for evolutionary biology, regarding interpretations of evolution and the nature of taxonomy in the mid-twentieth century. These primate phylogenies come from Simpson’s books The Meaning of Evolution (1949) (figure 6.11A) and Principles of Animal Taxonomy (1961) (see figure 6.11B). These two trees, now outdated on several levels, show primate evolution as viewed at the time. The two phylogenies are similar in general pattern but differ in two important ways. First, the 1949 tree shows a noticeably staccato appearance of the four labeled groups, whereas the 1961 tree evens out these sharp breaks. The 1949 representation is closer to Simpson’s view in Tempo and Mode (1944), in which evolution was regarded as happening both rapidly and slowly. By 1961, the revolt against the early-twentieth-century view of rapid, mutational change advocated by the early Mendelian geneticists was over. An interpretation—whether incorrect or not—of slow, stately evolutionary unfolding was in vogue. Simpson’s (1961) phylogeny reflects this perspective, whether intended by Simpson or not. He had been criticized for advocating rapid evolutionary change in Tempo and Mode, and his later writing and his trees reflect this. The second difference between these two phylogenies comes in the form of the suggested taxonomies, echoing the vertical versus horizontal classifications discussed by Romer (1933, 1971). In the two phylogenies, prosimians (a now discarded taxon) are shown both as a grade of early primate and as a clade or lineage. The phylogeny from 1949 recognizes four major named radiations of primates. The phylogeny from 1961 shows this but also puts a dotted line around prosimians, making them both a grade (horizontal) and a clade (vertical) of primates.
FIGURE 6.10 (A) Niles Eldredge and Stephen Jay Gould’s diagram of what they termed “punctuated equilibrium,” from “Punctuated Equilibria” (1972); (B) Simpson’s figure depicting a similar idea, which he called “explosive” evolution, modified from Tempo and Mode in Evolution (1944). ([A] reproduced with permission of Niles Eldredge)
FIGURE 6.11 Simpson’s trees separated by a decade and reflecting changing ideas about the adaptive radiation of primates, going from (A) a noticeably staccato, bunched appearance of the four labeled groups, from The Meaning of Evolution (1949), to (B) an evening out of the sharper breaks, from Principles of Animal Taxonomy (1961). ([A] reproduced with permission of Yale University Press; [B] reproduced with permission of Columbia University Press)
The trees of the mid-twentieth century, with those of scientists such as Simpson leading the charge, emphasized the grand unfolding radiations of life over geologic time, very much in accord with perceptions of a Darwinian view of evolution. But perceptions and research programs in the 1970s, including how to build and show trees, were about to change radically for three reasons: first, the emergence of a new school of phylogenetic reconstruction; second, the development of vastly increased computer power and algorithms; and third, the explosion of molecularly based phylogenetic analyses primarily by dint of a procedure called polymerase chain reaction (PCR) that made the analysis of vast quantities of genetic information possible. Our perceptions of evolution, how we show this in trees, and even our view of ourselves within this scheme would never be the same.
