* Invoking another aspect of the same image—the equation of old and extinct with inadequate—Granada exhorts us to rent rather than buy because “today’s latest models could be obsolete before you can say brontosaurus.”
† Wonderfully ironic, since the sequence showed, basically, more effective filters. Evolution, to professionals, is adaptation to changing environments, not progress. Since the filters were responses to new conditions—public knowledge of health dangers—Doral did use the term evolution properly. Surely, however, they intended “absolutely better” rather than “punting to maintain profit”—a rather grisly claim in the light of several million deaths attributable to cigarette smoking.
* A properly defined group with a single common ancestor is called monophyletic Taxonomists insist upon monophyly in formal classification. However, many vernacular names do not correspond to well-constituted evolutionary groups because they include creatures with disparate ancestries—“polyphyletic” groups in technical parlance. For example, folk classifications that include bats among birds, or whales among fishes, are polyphyletic The vernacular term animal itself probably denotes a polyphyletic group, since sponges (almost surely), and probably corals and their allies as well, arose separately from unicellular ancestors—while all other animals of our ordinary definitions belong to a third distinct group. The Burgess Shale contains numerous sponges, and probably some members of the coral phylum as well, but this book will treat only the third great group—the coelomates, or animals with a body cavity. The coelomates include all vertebrates and all common invertebrates except sponges, corals, and their allies. Since the coelomates are clearly monophyletic (Hanson, 1977), the subjects of this book form a proper evolutionary group.
† This fundamental principle, while true for the complex multicellular animals treated in this book, does not apply to all life. Hybridization between distant lineages occurs frequently in plants, producing a “tree of life” that often looks more like a network than a conventional bush. (I find it amusing that the classic metaphor of the tree of life, used as a picture of evolution ever since Darwin and so beautifully accurate for animals, may not apply well to plants, the source of the image.) In addition, we now know that genes can be transferred laterally, usually by viruses, across species boundaries. This process may be important in the evolution of some unicellular creatures, but probably plays only a small role in the phylogeny of complex animals, if only because two embryological systems based upon intricately different developmental pathways cannot mesh, films about flies and humans notwithstanding.
* Another factual irony: despite the usual picture of horseshoe crabs as “living fossils,” Limulus polyphemus (our American East Coast species) has no fossil record whatever. The genus Limulus ranges back only some 20 million years, not 200 million. We mistakenly regard horseshoe crabs as “living fossils” because the group has never produced many species, and therefore never developed much evolutionary potential for diversification; consequently, modern species are morphologically similar to early forms But the species themselves are not notably old.
* Twain used Lord Kelvin’s estimate, then current, for the age of the earth. The estimated ages have lengthened substantially since then, but Twain’s proportions are not far off. He took human existence as about 1/3000 of the earth’s age. At current estimates of 250,000 years for the origin of our species. Homo sapiens, the earth would be 0.75 billion years old if our span were 1/3000 of totality. By best current estimates, the earth is 4.5 billion years old.
* I have struggled over a proper name for this phenomenon of massive elimination from an initial set of forms, with concentration of all future history into a few surviving lineages. For many years, I thought of this pattern as “winnowing,” but must now reject this metaphor because all meanings of winnowing refer to separation of the good from the bad (grain from chaff in the original)—while I believe that the preservation of only a few Burgess possibilities worked more like a lottery.
I have finally decided to describe this pattern as “decimation,” because I can combine the literal and vernacular senses of this word to suggest the two cardinal aspects stressed throughout this book: the largely random sources of survival or death, and the high overall probability of extinction.
Randomness. “Decimate” comes from the Latin decimare, “to take one in ten.” The word refers to a standard punishment applied in the Roman army to groups of soldiers guilty of mutiny, cowardice, or some other crime. One soldier of every ten was selected by lot and put to death. I could not ask for a better metaphor of extinction by lottery.
Magnitude. But the literal meaning might suggest the false implication that chances for death, though applied equally to all, are rather low—only about 10 percent. The Burgess pattern indicates quite the opposite. Most die and few are chosen—a 90 percent chance of death would be a good estimate for major Burgess lineages. In modern vernacular English, “decimate” has come to mean “destroy an overwhelming majority,” rather than the small percentage of the ancient Roman practice. The Oxford English Dictionary indicates that this revised usage is not an error or a reversed meaning, but has its own pedigree—for “decimation” has also been used for the taking of nine in ten.
In any case, I wish to join the meaning of randomness explicit in the original Roman definition with the modern implication that most die and only a few survive. In this combined sense, decimation is the right metaphor for the fate of the Burgess Shale fauna—random elimination of most lineages.
* There are two in jokes in this line: orogeny is standard geological jargon for mountain building; Paleobscene is awfully close to the epoch’s actual name—Paleocene.
* The “lowest Silurian” refers to rocks now called Cambrian, a period not yet codified and accepted by all in 1859. Darwin is discussing the Cambrian explosion in this passage.
* Although the 12C/13C ratio in the Isua rocks is indicative of organic fractionation, the excess of 12C is not so high as for later sediments. Schidlowski argues that the subsequent metamorphism of the Isua rocks lowered the ratio (while leaving it within the range of organic values), and that the original ratio probably matched that of later sediments.
* Burgess himself was a nineteenth-century governor general of Canada; Walcott named the formation not for him but for Burgess Pass, which provided access to the quarry from the town of Field.
* Much material in this section comes from my previous essay on Walcott’s discovery (Gould, 1988).
* I know this so well from personal experience. People ask me all the time what 1 was thinking when Niles Eldredge and I first developed the theory of punctuated equilibrium in the early 1970s. I tell them to read the original paper, for I don’t remember (or at least cannot find those memories amidst the jumble of my subsequent life).
* These outer coverings were, of course, harder than the soft organs below. But the carapaces of most Burgess organisms were not mineralized, and therefore not formed of conventional “hard parts” that fossilize easily. These carapaces were rather like the exoskeletons of modern insects, stiffened but not mineralized. “Lightly sclerotized” might be a better term than “soft-bodied,” but the potential for conventional fossilization is nearly nil in either case.
* I asked Whittington why so little work had been done before his redescriptions, for Walcott’s specimens had always been available at the Smithsonian. He cited a number of reasons, all no doubt contributory, but not enough in their ensemble to explain this curious fact. Walcott’s wife, for one, was quite possessive and discouraging, though she held no proprietary power over the specimens. She hated Percy Raymond for collecting again at the Burgess so soon after her husband’s death in 927. Raymond, in his turn, had been no fan of Walcott’s, and taunted him as “the great executive paleontologist” for letting administrative work absorb all his time, thus precluding a proper study of the Burgess fossils. (This was an unusually acerbic assessment for Raymond, who was the most mild-mannered of men. Al Romer, who knew him well, once told me that Raymond was at the bottom of a familial pecking order, with his wife, children, and dog above him. His favorite hobby, collecting pewterware, definitely contributed to his non-macho image.) While Walcott lived, no one else would work on the specimens, for he always intended to do a proper study himself, and no one dared upstage the most powerful man in American science. (Such proprietary claims are traditionally honored in paleontology, even for scientists low on the totem pole; discovery implies the right of description, with a statute of limitation often construed as extending for a lifetime.) Walcott’s wife, and the memory of his power, managed to extend a reluctance for work on Burgess material even beyond Walcott’s grave. Moreover, as Whittington reports, although the “type” specimens were accessible (the few used in the original descriptions of the species), almost all the material resided in drawers placed high in cabinets, and therefore unavailable for casual browsing-the serendipitous mode of origin for many paleontological studies. They also were housed in a building without air conditioning (now remedied). Most paleontologists work in universities, and have substantial free time only during the summer. Need I say more to anyone who has experienced the pleasures of our nation’s capital in July or August!
* Walcott, of course, had not failed to note this prominent organ, and its uniqueness did pose a problem for his conclusion that Yohoia was a branchiopod. Walcott evaded this dilemma by arguing that the great appendage was a male “clasper,” or structure used to hold females during mating (and present in many branchiopods). But Whittington determined that all specimens bore great appendages, and disproved Walcott’s rationale.
† Walcott had placed two species in the genus Yohoia—Y. tenuis and Y. plena. Whittington realized that the two animals are distinct and belong in different genera. Y. plena, which has antennae, is a phyllocarid, one of the arthropods with a bivalved carapace soon to be studied by Derek Briggs. Whittington removed this species from Yohoia and established a new genus, Plenocaris. Yohoia tenuis is the oddball, and subject of the 1974 monograph.
* I view this as a crucial and favorable feature for the general story of this book—because you can be sure that Whittington came to his new interpretation of the Burgess from an accumulating weight of evidence, not from any a priori desire to go down in history as a radical reformer.
* The leading British professional association of paleontologists. They call themselves the “pale ass” for informal fun—a name even more humorous to an American, since the title only refers to a donkey in England (where your nether end is your arse).
* A. M. Simonetta, an Italian paleontologist, deserves a great deal more credit than this book has space to provide. He alone, after Walcott and before Whittington, attempted a comprehensive program of revision for Burgess arthropods. He worked as Walcott had, and with Walcott’s specimens, treating the fossils essentially as films on the rock surface and attempting no preparation of specimens. He therefore made many mistakes in a long series of papers published during the 1960s and 1970s. But he also provided substantial improvements upon several earlier studies, and through his comprehensive efforts reminded paleontologists about the richness of the Burgess Shale. Since science is a process of correction, Simonetta’s errors also provided an important spur to Whittington and his colleagues.
* My Catholic friends may cite Pius IX and December 8, 1854, for the last item in my list, but Ineffabilis Deus was an official resolution under the rules of the institution, and no one could pick one moment as paramount in a millennium of previous debate. On Darwin’s long and complex struggle to develop the theory of natural selection, see Howard Gruber, Darwin on Man. (New York: Dutton, 1974).
* Since Simon and Derek began working with Harry Whittington in 1972, the year of the infamous laughter over Opabinia at the Oxford meeting, I had assumed that their prodding must have convinced Harry to take the drastic step of declaring Opabinia as a unique anatomy of phylum-level status. This is how the script is supposed to go—the Young Turks dragging the old farts into the light of exciting modernity. Terrible screenplay, not at all like complex life. Simon may be ideologically radical, but he is one hell of an excellent descriptive anatomist—and anyone who would be fooled enough by externalities to rank Harry as an old fart understands nothing about the multifarious nature of genius. In any case, all three protagonists assure me that Harry worked out the interpretation of Opabinia without any hectoring or encouragement from radicals on the sidelines. The converse is equally true and contrary to script. Harry neither discouraged Simon as he wrote his five papers, nor helped with frequent counseling. Harry played virtually no role in Simon’s first forays. He can remember only one intervention—an insistence that Simon use his techniques of dissection to excavate the spines of Hallucigenia right to the point of their connection with the body. Damned good advice, but scarcely the stuff of general guidance.
* I don’t say this in a critical, revelatory, or muckraking mood. Journalistic traditions properly match their assigned roles. I only point out that different approaches see only restricted parts of a totality—as in the overworked simile of the blind men and the elephant—and that one can get something gloriously wrong by mistaking a small and biased segment for an entity.
* The mouth parts of arthropods have been given the same names as functionally comparable structures in vertebrates—maxilla, mandible, and so forth. Similarly, the parts of insect legs bear the same names—trochanter, tibia—as their vertebrate counterparts. This is an unfortunately confusing nomenclature, for whatever the functional similarities, the structures have no evolutionary connection: insect mouth parts evolved from legs; vertebrate jaws from gill arches.
* As an indication of how much struggle and effort can underlie the conclusions stated so briefly in my text, consider this interesting note that Derek Briggs wrote to me as a reaction to this passage when I sent the manuscript of this book to him: “The work on Canadaspis became a hunt for the first crustacean.… By then the expectation was that the odds on any of the arthropods falling into living groups were very low. The problem with Canadaspis was finding the critical evidence of the posterior cephalic [head] appendages. USNM 189017 [catalog number of a key specimen in the United States National Museum] is the best of only about 3 (out of thousands) specimens which show these limbs in lateral view (they are almost without exception totally obscured by the carapace, compaction etc.), and as you can see on Plate 5 (Briggs 1978) it was a huge job preparing the specimen to show them. In my view figs. 66–69 on that plate represent the peak of what can be achieved by preparing part and counterpart in tandem. I then had a major job convincing Sidnie Manton (Harry’s arthropod guru) that I did indeed have the critical evidence—at the time I considered that an enormous achievement! [Manton was the world’s greatest expert on the higher-level classification of arthropods—and one tough lady.] It was not just a case of the evidence of the specimens; it was necessary to argue that the first two pairs of a series of 10 pairs of similar biramous appendages belonged to the head—although they remain primitive in not being significantly differentiated from those which follow.”
* I am as committed as anyone to “ecology” (in the vernacular and political meaning of leaving nature alone), and I certainly believe in respecting the nearly sacred integrity of national parks. But a fossil on the ground is worth absolutely nothing. It is not an object of only pristine beauty, or a permanent part of any natural setting (especially for fossils exposed in quarry walls). If free on the ground, it will probably be cracked and frost-heaved into oblivion by the next field season. Controlled collecting and scientific study are the proper roles, intellectually and ethically, for the Burgess fossils.
* The status of the Onychophora, probable taxonomic home of Aysheaia, remains controversial. Some experts regard the Onychophora as an entirely separate phylum, no closer to the uniramians than to any other group of arthropods. If this solution is correct, my argument here is wrong. The two other major solutions both support my argument: first, that Onychophora should rank within the Arthropoda on the uniramian line; second (and probably the predominant view), that onychophorans deserve separate status, but lie closer to the uniramians than to any other group of arthropods. (This last argument assumes a separate evolutionary origin for several, perhaps all four, of the great arthropod lines—with uniramians arising in genealogical proximity to onychophorans.)
* A small and little-known molluscan group called the Aplacophora does seem more similar in its elongate, wormlike body, sometimes covered with plates or spicules, but Conway Morris enumerates an impressive list of detailed differences in his monograph.
* If I wished to play devil’s advocate against my own framework, I would argue that the criterion by which we make the claim of twenty losers and only four winners is falsely retrospective. By patterns of tagmosis, modern arthropods are surely strikingly less disparate than Burgess forebears. But why use patterns of tagmosis as a basis for higher-level classification of arthropods? A nearly microscopic ostracode, a terrestrial isopod, a planktonic copepod, a Maine lobster, and a Japanese king crab span more variety in size and ecological specialization than all the Burgess arthropods put together—though all these modern creatures are called Crustacea, and display the stereotyped tagmosis of this class. A paleontologist living during the Burgess might consider the arthropods as less varied because he had no reason to regard patterns of tagmosis as a particularly important character (for the utility of tagmosis in distinguishing major genealogical lines only became apparent later, after most alternatives were decimated and stereotypy set in among the few surviving and highly disparate lines).
I regard this argument as a poor case. If you wish to reject tagmosis as too retrospective then what other criterion will suggest less disparity in the Burgess? We use basic anatomical designs, not ecological diversification, as our criterion of higher-level classification (bats and whales are both mammals). Nearly every Burgess genus represents a design unto itself by any anatomical criterion. Tagmosis does stabilize in post-Burgess times, as do arrangement and forms of appendages-while no major feature of arthropod design can distinguish broad and stable groups in the Burgess.
* Many of Walcott’s cruder errors, on the other hand—confusing the sclerites of Wiwaxia with setae of polychaetes, and the lateral Haps of Opabinia with arthropod segments represented a more basic failure to distinguish analogy from homology.
* Thus, we can take some steps to resolving the genealogy of Burgess organisms. We can eliminate some resemblances based on analogy-setae of polychaetes and sclerites of Wiwaxia, for example. We can also eliminate some shared-but-primitive characters that do not define genealogical groups-bivalved carapaces and “merostomoid” body form. But the identification of shared-and-derived characters has been largely unsuccessful so far. Homology of shared-and-derived frontal appendages may unite Leanchoilia with Actaeus (and perhaps also with Alalcomenaeus). The lateral Raps with gills above may be shared-andderived characters in Opabinia and Anomalocaris, thus constituting the only genealogical linkage between two of the weird wonders.
* Technical footnote: Several efforts have been made to construct a cladogram for the Burgess arthropods (Briggs, 1983, and in press). These have, so far, been conspicuously unsuccessful, as the different possibilities do not satisfactorily converge. If the grabbag model is correct, and each maior feature of each new lineage arises separately from a suite of latent possibilities common to all, then genealogical connectivity of phenotypes is broken, and the problem may be intractable by ordinary cladistic methods. Of course, some continuity in some genuinely nested sets of characters may well exist, but the appropriate features will be difficult to identify.
† I exaggerate to make a point. Rules of construction and order pervade nature. Not all conceivable combinations can work, nor can all amalgams be constructed within the developmental constraints of metazoan embryology. I use this metaphor only to express the vastly expanded range of Burgess possibilities.
* Biology textbooks often speak of variation as “random.” This is not strictly true. Variations are not random in the literal sense of equally likely in all directions; elephants have no genetic variation for wings. But the sense that “random” means to convey is crucial: nothing about genetics predisposes organisms to vary in adaptive directions. If the environment changes to favor smaller organisms, genetic mutation does not begin to produce biased variation toward diminished size. In other words, variation itself supplies no directional component. Natural selection is the cause of evolutionary change; organic variation is raw material only.
* I retranslate here, hoping not to repeat one of the greatest absurdities I ever encountered—Milton’s Paradise Lost translated into German as part of the libretto for Haydn’s Creation, then retranslated as doggerel for a performance in English that could not use Milton’s actual words and still retain Haydn’s musical values.
* Perhaps the most touching document in the Walcott archives at the Smithsonian Institution is the highly personal note of condolence written to Walcott by Roosevelt upon the accidental death of Walcott’s second wife.
* Yes, this is William Howard Taft, then ex-president, and acting chief justice of the United States, who introduced this memorial meeting for Walcott.
* I do not like to discuss intellectual issues as abstract generalities. I believe that conceptions are best appreciated and understood through their illustration in a person’s idea, or in a natural object. Thus, I am charmed and fascinated by Walcott. I have rarely “met” a man so out of tune with my own view of life—and I do feel that I know him after so much intimacy from the archives. Yet 1 have gained enormous respect for Walcott’s integrity and demoniacal energy in research and administration. I do not particularly like him (as if my opinion mattered a damn), but I am mighty glad that he graced my profession.
* Walcott is identified on this manuscript as “of the Geological Survey and Honorary Curator of Paleozoic Fossils in the National Museum.” He held the honorary curatorial post from 1892 until he became secretary of the Smithsonian in 1907. I assume that he had not yet been appointed director of the Survey, for he would have been so identified. Since he became director in 1894, the date of the lecture must be between 1892 and 1894.
* One tangential point before leaving this rare example of a public address by such a private and imperious man. Walcott was a clear but uninspired writer. So many professionals make the mistake of assuming that popular presentations of science—particularly writing about nature—must abandon clarity for overblown, rapturous description. A Wordsworth or a Thoreau can pull it off; the great majority of naturalists, however great their emotional love for the outdoors, cannot—and should not try, lest the ultimate in unintended parody arise. Besides, audiences do not need such a crutch. The “intelligent layperson” exists in abundance and need not be coddled. Nature shines by herself. But, in any case, and with some embarrassment, I give you Charles Doolittle Walcott on the Grand Canyon at sunset:
The Western sky is all aflame. The scattered banks of clouds and wavy cirrus have caught the warring splendor, and shine with orange and crimson. Broad slant beams of yellow light, shot through the glory-rifts, fall on turret and tower, on pinnacled crest and wending ledge, suffusing through with a radiance less fulsome, but akin to that which flames in western clouds. The summit band is brilliant yellow, the next below is pale rose. But the grand expanse within is deep, luminous, resplendid [sic] red. The climax has now come; the blaze of sunlight poured over an illimitable surface of glowing red is flung back into the gulf, and, commingling with the blue haze, turns it into a sea of purple of most imperial hue. However vast the magnitudes, however majestic the forms or sumptuous the decoration, it is in these kingly colors that the highest glory of the Grand Canyon is revealed.
* Our agreement on the theme, if not the terminology, provides hope that even the most implacable differences in style and morality may find a common meeting ground on this most important of intellectual turfs—for Steve is the most fanatical Red Sox booster in New England, while my heart remains with the Yankees.
* “Holotype” is taxonomic jargon for the specimen designated to bear the name of a species. Holotypes are chosen because concepts of the species may change later and biologists must have a criterion for assigning the original name. (If, for example, later taxonomists decide that two species were mistakenly mixed together in the first description, the original name will go to the group including the holotype specimen.)
* Mass extinctions do not negate the principle of natural selection, for environments can change too fast and too profoundly for organic response; but coordinated dyings do run counter to Darwin’s preference for seeing the large in the small, and for viewing organic competition, group by separate group, as the primary source of life’s overall pattern.
* The repetition of the Burgess pattern by conventional groups with hard parts is very fortunate and favorable for testing the main issue presented by the phenomenon of decimation: Do losers disappear by inferiority in competition, or by lottery? Unfortunately, we can learn little about this key question from the Burgess Shale itself, for this soft-bodied fauna is only a spot in time, and we have virtually no evidence about the pattern of later decimation. (One Devonian arthropod, Mimetaster from the Hunsrückschiefer, is probably a surviving relative of Marrella; most other Burgess anatomies disappear without issue, and we have no evidence at all for how or when.) But patterns of extinction in groups with hard parts can be traced. Paradoxically, therefore, the best and most operational way to test for sources of decimation in the Burgess would be to study the parallel and tractable situation in echinoderms. My first question: do echinoderm “failures” tend to disappear at full abundance during mass extinctions, or to peter out gradually at different uncoordinated times? The former situation would be strong evidence for a substantial component of lottery in decimation. We do not know the answer to this question, but the solution is obtainable in principle.
* Geographic range is a property of populations, not of individual clams or snails. Hence, even if survival is correlated with geographic range, a species’fate may be random with respect to the anatomical virtues of its individuals.