The Flamingo’s Smile

by Stephen Jay Gould

  5 | A Most Ingenious Paradox

  ABSTINENCE HAS its virtuous side, but enough is enough. I have always felt especially sorry for poor Mabel, betrothed to Frederic the pirate apprentice. On the very threshold of married happiness, she discovers that she must wait another sixty-three years to claim her beloved at age eighty-four—and, as could only happen in Gilbert and Sullivan, she actually promises to wait.

  The Pirate King and Ruth, Frederic’s old nursemaid and jilted paramour, present the reason for this extraordinary delay. Frederic, wrongfully apprenticed to the pirate band, has reached his twenty-first year and longs for freedom, respectability, and Mabel. But he was formally bound until his twenty-first birthday, and he was born on February 29. “You are a little boy of five,” the Pirate King informs him with glee and expectation of prolonged service. The three principals of the Pirates of Penzance then analyze the complexities of this predicament in the famous paradox song:

  How quaint the ways of paradox

  At common sense she gaily mocks.

  The classic paradox presents us with two contradictory interpretations, each quite correct in its own context. Consider our western prototypes, the so-called paradoxes of Zeno: The arrow that can never reach its destination because, at any instant, it must occupy a fixed position; and Achilles who will never catch the tortoise because he must first traverse half the remaining distance, and any gap, no matter how small, can still be halved. We delight in paradox because it appeals to both the sublime and whimsical aspects of our psyche. We laugh with Frederic, but also feel that something deep about the nature of logic and life lies hidden in Zeno’s conundrums.

  Biology too has its classical paradox. It flared as a major issue in the nineteenth century, probably because scientists then felt that they might find a resolution. All the best naturalists struggled in vain: Huxley and Agassiz lined up on opposite sides; Haeckel tried to mediate. The twentieth century has largely bypassed the conundrum, probably because we now realize that no simple answer exists. Yet, if our fascination with paradox be justified, the question can still enlighten us by virtue of its stubborn intractability.

  Physalia, the Portuguese man-of-war, embodies all this fuss. It is a siphonophore, a relative of corals and jellyfish. The old paradox addresses an issue that could not be more fundamental—the definition of an organism and the general question of boundaries in nature. Specifically: Are siphonophores organisms or colonies?

  Siphonophores belong to the phylum Cnidaria (or Coelenterata). Two aspects of cnidarian biology set the context of our paradox. First, many cnidarians live as colonies of connected individuals—our massive coral reefs are gigantic congeries made of many million tiny, conjoined polyps. Second, the cnidarian life cycle features a so-called alternation of generations. The sessile polyp, a fixed cylinder with a fringe of tentacles, reproduces asexually and generates by budding the free-swimming medusa, or “jellyfish.” The medusa produces sexual cells that unite and grow into a polyp. And so it goes.

  A Portuguese man-of-war. The creature is a colony, not a single organism. The float is a medusa person, and each “tentacle” is a polyp person. FROM LOUIS AGASSIZ’S MONOGRAPH (1862), REPRINTED FROM NATURAL HISTORY.

  T.H. Huxley’s 1849 illustration of the Portuguese man-of-war. He regarded this creature as an individual, not a colony.

  Different kinds of cnidarians may emphasize one of these generations and suppress the other. Of the three major cnidarian groups, the Scyphozoa (or true jellyfish) have abandoned polyps and emphasized medusae, while the Anthozoa (or true corals) have dispensed with medusae and constructed reefs of polyps and their skeletons. In the third group, the Hydrozoa, many members retain the full cycle, with prominent polyp and medusa. Siphonophores are hydrozoans. The technical literature, not usually noted either for charm or directness, has transcended its usual limitatons in this case: amidst a forest of formidable jargon for other parts of cnidarian anatomy, it refers to the polyp and medusa stages of a single life cycle as “persons.”

  The Portuguese man-of-war, with its float above and tentacles below, looks superficially like a jellyfish (that is, a single medusa). When studied more carefully, we find that this floating weapon is a colony of many persons, both polyps and medusae. The pneumatophore, or float, is probably a greatly modified medusa (though some scientists think that it may be an even more altered polyp). The “tentacles,” though varied and specialized for different roles of capturing food, digestion, and reproduction, are not simple parts of a jellyfish but modified polyps—that is, each tentacle arises as a discrete person. (Another common siphonophore, Velella, literally the “little sail” but popularly given the lovely name of “by-the-wind sailor,” provokes even more confusion. Its persons are few enough and so well coordinated that the colony looks like a simple float surrounded by tentacles—in other words, like a simple jellyfish. But the float is a medusa person and each tentacle a polyp person.)*

  If this degree of division of labor among persons impresses you, nature has much more to offer. Physalia and Velella are simple siphonophores, with relatively few types of modified persons. The more complex siphonophores are, by far, nature’s most integrated colonies. Their parts are so differentiated and specialized, so subordinate to the entire colony, that they function more as organs of a body than persons of a colony.

  Velella, the “by-the-wind-sailor,” is a colony—the float is a medusa person, each “tentacle” is a polyp person. FROM E. HAECKEL’S Challenger MONOGRAPH (1888). REPRINTED FROM NATURAL HISTORY.

  Most siphonophores are small, transparent creatures of the open sea. They float at the surface among the plankton or swim actively, usually at shallow depths. As carnivores, they capture small planktonic animals in their net of tentacles. Larger siphonophores, Physalia among them, can ensnare and devour fish of substantial size; as many of us know to our sorrow, they can also inflict painful stings upon human bathers.

  Complex siphonophores include an imposing array of well-differentiated structures. Their bodies may be roughly divided into two parts: an upper set of bulbs and pumps for locomotion and a lower set of tubes and filaments for feeding and reproduction. Each part contains a series of different polyps and medusae.

  A relatively “simple” representative of the complex siphonophores, just for starters. Only four basic types of individuals are shown—two upper kinds of medusa persons (the pneumatophore, or float, labeled p; and a row of nectophores, or swimming bells, labeled n); and two lower kinds of polyp persons (the feeding siphons, labeled s; and the long sensory tentacles, labeled t). FROM HAECKEL’S Challenger MONOGRAPH, 1888.

  Consider first the range of forms and activities assumed by polyp persons. We find three basic types and a myriad of modifications. The feeding organs, or siphons (hence the group’s name—siphonophore means “siphon bearer”), are tubular structures each with a stomach and trumpet-shaped mouth, usually hanging in profusion below the floats and swimming persons. The siphons are minimally modified polyp persons, and we can easily comprehend their origin as complete organisms. All other types of polyps (and most medusae) are more highly altered and specialized, and therefore more difficult to link with their original personality. A second order of polyp persons, the so-called dactylozooids (“finger,” or touching, animals), capture and transport food to the siphons. Dactylozooids include the extended thin tentacles, sometimes more than fifty feet long in Physalia, that carry the painful nematocysts, or stinging cells, and form a transparent web to ensnare prey. They have retained neither mouth nor digestive apparatus and might easily be taken for parts rather than persons if we could not trace their origin as discrete buds in growth.

  These capturing parts often display a remarkable complexity of form and function. The stinging cells may be concentrated into swellings, or “batteries,” sometimes protected by a hood. In Stephanophyes, each battery ends in a delicate terminal filament and contains about 1,700 stinging cells of four different types. The terminal filament lassoes the prey and d
ischarges its few stinging cells. If these cells fail to dispatch the victim, the filament contracts and carries the prey to the far end of the battery itself, where another volley of larger stinging cells transfixes the victim. If the prey continues to struggle, another contraction moves it up the battery to the near end, where the largest and most powerful stinging cells finally end the torment before passing the vanquished prey along to the siphon for ingestion.

  Jennifer E. Purcell (see bibliography) has recently presented further evidence that feeding and capturing persons do not form a simple passive network, like the web of a spider, but play an active role in obtaining food. She found that the stinging cell batteries of two species function as lures by resembling, in both form and motion, small zoo-plankton that serve as prey for animals eaten by siphonophores. The batteries of Agalma okeni look like a copepod with two long antennae; each contracts independently at varying intervals of five to thirty seconds, creating a web of motion that simulates the darting and swimming of a copepod school (or whatever you call an aggregation of these tiny planktonic arthropods). To seal the story, Purcell opened the stomachs of Agalma and found the remains of three creatures, all predators of copepods. The batteries of another species, Athorybia rosacea, resemble the planktonic larvae of fish. They also contract rapidly, mimicking the swimming and feeding motions of their models.

  Gonozooids, the third category of polyp persons, are reproductive structures. They are usually short, simple tubes, without mouth or motion. But they bud off the medusa persons, which then make reproductive cells to produce the next generation of siphonophores.

  The medusa persons of a complex siphonophore include four basic types: swimming, floating, protection, and reproduction. The swimming organs, or nectophores, are minimally modified medusae—basically the upper swimming bells without the lower tentacles. Some siphonophores carry several orderly rows of nectophores; their rhythmic muscular contractions propel the creature, often in elaborate, looping trajectories. The passive floats, or pneumatophores, are filled with gas (of a composition near ordinary air) and maintain the siphonophore at the surface or at some intermediate depth. Their origin is a matter of controversy. Long interpreted as modified medusa persons, some biologists now regard pneumatophores as even more elaborately transformed polyps. The two most familiar siphonophores, Velella and Physalia, build large floats but contain no nectophores. They therefore move passively on winds and currents, often drifting into bays and beaches in vast accumulations.

  The covering organs, or bracts, are the most curiously modified structures of all. They are usually flat, shaped like a prism or a leaf, and so different in form and function from a medusa person that we would scarcely suspect their origin if we could not follow their growth and budding.

  The reproductive medusae, or gonophores, are budded off from polyp persons, the gonozooids discussed earlier. In a few species, gonophores are liberated to float in the ocean as independent objects. But they cannot feed, and die soon after releasing their sex cells. In most siphonophores, however, gonophores never separate from the parent colony and remain attached as a kind of sexual organ.

  The middle figure shows a complete and complex siphonophore. The colony includes the following modified persons, from the top to the bottom: the single float, or pneumatophore (p); rows of swimming organs, or nectophores (n); fingerlike sensory projections, or palpons (q); clusters of reproductive parts (g); feeding siphons with trumpet-shaped mouths (s); and long, twisted strands of food-capturing filaments (t). Other figures are parts or early growth stages of the complex colony. FROM E. HAECKEL’S Challenger MONOGRAPH, 1888. REPRINTED FROM NATURAL HISTORY.

  One more complex siphonophore for good measure, and with yet another kind of person added (protective bracts). From top to bottom: a single pneumatophore; two vertical rows of nectophores; protective leaflike bracts; feeding siphons with trumpet-shaped mouths; and finally, long food-capturing filaments. FROM E. HAECKEL (1888). REPRINTED FROM NATURAL HISTORY.

  The paradox of the Siphonophora expresses an issue that I have been avoiding, or rather skirting about, in presenting this taxonomy of persons or parts. I have described the various swimming, floating, protecting, feeding, capturing, and reproducing structures as persons—that is, as individual polyp or medusa organisms. Using evolutionary history as a criterion, this designation is almost surely correct and accepted by nearly all biologists. By history, siphonophores are colonies; they evolved from simpler aggregations of discrete organisms, each reasonably complete and able to perform a nearly full set of functions (as in modern coral colonies). But the colony has become so integrated, and the different persons so specialized in form and so subordinate to the whole, that the entire aggregation now functions as a single individual, or superorganism.

  The persons of a siphonophore no longer maintain individuality in a functional sense. They are specialized for a single task and perform as organs of a larger entity. They do not look like organisms and could not survive as separate creatures. The entire colony works as a single being, and its parts (or persons) move in a coordinated manner. Although each nectophore (or swimming bell) maintains its own nervous system, a common nerve tract connects the entire set. Impulses along this pathway regulate the rows of nectophores in an integrated manner that permits the whole colony (or animal) to move with precision and grace. Touch the float of Nanomia at one end, and nectophores at the other extremity contract to remove the animal (or colony, if you will) from danger. Siphons pump their digested food along the common stem to the rest of the colony, but empty siphons also join in the general peristalsis, and food, as a result, reaches the entire colony (or organism) more effectively.

  My studied parentheticals of the last paragraph underscore the fundamental paradox. Shall we call the entire siphonophore a colony or an organism—for it is a colony by evolutionary history but more an organism by current function. And what of the parts or persons? By history, they are modified individuals; by current function, they are organs of a larger entity. What is to be done?

  This issue fueled the great siphonophore debate of nineteenth-century natural history. T.H. Huxley studied siphonophores during his long apprenticeship at sea on H.M.S. Rattlesnake (less celebrated than Darwin’s adventure on the Beagle, but another example of the same extended, exemplary, and largely extinct style of training in natural history). He interpreted siphonophores as conventional organisms, their parts as true organs and not modified persons. Huxley used siphonophores as his primary example in a famous essay on the nature of individuality in biology.

  Louis Agassiz studied the Portuguese man-of-war on the shores of his adopted America (I have included his beautiful lithograph of Physalia with this essay) and decided that siphonophores are colonies, their integration a sign of divine handiwork.

  Ernst Haeckel, artist and naturalist extraordinaire, described the siphonophores collected on that most celebrated of scientific expeditions in oceanography, the voyage of H.M.S. Challenger, 1873–1876. He published with his report a series of plates (including all other illustrations in this essay), unmatched ever since for beauty (though a bit short on accuracy, since Haeckel often added a touch of heightened symmetry for artistic effect). Haeckel also included several plates of siphonophores in his Kunstformen der Natur (Art Forms in Nature) of 1904—the great series of 100 lithographs, with plants and animals arranged in weirdly distorted form and swirling symmetry, in the best tradition of reigning art nouveau so well embodied in contemporary kiosks of the Paris Métro.

  Haeckel’s theory of siphonophores would require an essay in itself to explain and explore, but he tried to mediate between Huxley and Agassiz by viewing these creatures in part as colonies (the poly-person theory in his words), in part as organisms (the poly-organ theory). Haeckel also used siphonophores, as Huxley had, to illustrate by dubious analogy his views on the proper organization of human societies. In his Über Arbeitstheilung in Natur und Menschenleben (On the Division of Labor in Nature and Human Life), he comp
ared the simple colonies of other cnidarians with the life styles of “primitive” humans and their limited division of labor for repetitive tasks performed by all: “The wild people of nature, who have remained on the lowest level right to our own day, lack both culture and division of labor—or they limit division of labor, as do most animals, to the different tasks of the two sexes.” He then compared complex colonies of siphonophores with the “advances” that division of labor permits in “higher” human societies—including modern warfare, where instruments of destruction “require hundreds of human hands, working in different ways and manners.”

  Can we now suggest any resolution to this ancient debate, any possible mediation between two legitimate criteria that seem to yield opposite results—the criterion of history supporting the poly-person theory (siphonophores are colonies and their parts are persons) and the criterion of current function upholding the poly-organ theory (siphonophores are organisms and their parts are organs)? Can we tip the balance in favor of one view or the other by invoking the third major criterion of natural history—growth and form?

  And finally, just for its aesthetic value, another Haeckel (1888) plate of complex siphonophores.

  Growth and form provide us with an embarras de richesses by presenting evidence for and against both theories. As strong support for the poly-organ theory, siphonophores develop from a single fertilized egg cell. A siphonophore begins life as an unambiguous person—should we not regard any later development as an elaboration of this founding individual? Moreover, the adult siphonophore acts as a discrete object. Many species exhibit definite and complex symmetry governing all parts considered together. Some Portuguese men-of-war, for example, come in right- and left-handed versions.