Bully for Brontosaurus

by Stephen Jay Gould

  These larval glowworms are profoundly modified members of the family Mycetophilidae, or fungus gnats. Imagoes of this species are unremarkable, but the larvae rank among the earth’s most curious creatures. Two larval traits (and nothing imaginal) inspired the name for this peculiar species—Arachnocampa luminosa, honoring both the light and the silken nest that both houses the glowworm and traps its prey (for Arachne the weaver, namesake of spiders, or arachnids, as well). The imagoes of Arachnocampa luminosa are small and short-lived mating machines. The much larger and longer-lived larvae have evolved three complex and coordinated adaptations—carnivory, light, and webbing—that distinguish them from the simpler larval habits of ancestral fungus gnats: burrowing into mushrooms, munching all the way.

  In a total life cycle (egg to egg) often lasting eleven months, Arachnocampa luminosa spends eight to nine months as a larval glowworm. Larvae molt four times and grow from 3- to 5-millimeter hatchlings to a final length of some 30 to 40 millimeters. (By contrast, imagoes are 12 to 16 millimeters in length, males slightly smaller than females, and live but one to four days, males usually longer than females.)

  Carnivory is the focus of larval existence, the coordinating theme behind a life-style so different from the normal course of larval herbivory in fungus gnats. Consider the three principal ingredients:

  Luminescence: The light organ of A. luminosa forms at the rear end of the larva from enlarged tips of four excretory tubes. These tubes carry a waste product that glows in the presence of luciferase, an enzyme also produced by the larva. This reaction requires a good supply of oxygen, and the four excretory tubes lie embedded in a dense network of respiratory tubules that both supply oxygen to fuel the reaction and then reflect and direct the light downward. This complex and specially evolved system functions to attract insects (mostly small midges) to the nest. Pupae and imagoes retain the ability to luminesce. The light of female pupae and adults attracts males, but the glow of adult males has no known function.

  The Nest and Feeding Threads: From glands in its mouth, the glowworm exudes silk and mucus to construct a marvel of organic architecture. The young larva first builds the so-called nest—really more of a hollow tube or runway—some two to three times the length of its body. A network of fine silk threads suspends this nest from the cave’s ceiling. The larva drops a curtain of closely spaced feeding threads from its nest. These “fishing lines” may number up to seventy per nest and may extend almost a foot in length (or ten times the span of the larva itself). Each line is studded along its entire length with evenly spaced, sticky droplets that catch intruding insects; the entire structure resembles, in miniature, a delicate curtain of glass beads. Since the slightest current of air can cause these lines to tangle, caves, culverts, ditches, and calm spaces amidst vegetation provide the limited habitats for A. luminosa in New Zealand.

  Carnivory: Using its lighted rear end as a beacon, A. luminosa attracts prey to its feeding threads. Two posterior papillae contain sense organs that detect vibrations of ensnared prey. The larva then crawls partway down the proper line, leaving half to two-thirds of its rear in the nest, and hauls up both line and meal at a rate of some 2 millimeters per second.

  The rest of the life cycle pales by comparison with this complexity of larval anatomy and behavior. The pupal stage lasts a bit less than two weeks and already records a marked reduction in size (15 to 18 millimeters for females, 12 to 14 for males). I have already noted the imago’s decrease in body size and duration of life. Imaginal behavior also presents little in the way of diversity or complexity. Adult flies have no mouth and do not feed at all. We commit no great exaggeration by stating that they behave as unipurpose mating and egg-laying machines during their brief existence. Up to three males may congregate at a female pupa, awaiting her emergence. They jockey for positon and fight as the female fly begins to break through her encasement. As soon as the tip of her abdomen emerges, males (if present) begin to mate. Thus, females can be fertilized even before they break fully from the pupal case. Females may then live for less than a day (and no more than three), doing little more before they expire than finding an appropriate place for some 100 to 300 eggs, laid one at a time in clumps of 40 to 50. Males may live an additional day (up to four); with luck, they may find another female and do it again for posterity.

  As a final and grisly irony, emphasizing larval dominance over the life cycle of A. luminosa, a rapacious glowworm will eat anything that touches its feeding threads. The much smaller imagoes often fly into the lines and end up as just another meal for their own children.*

  Please do not draw from this essay the conclusion that larvae are really more important than imagoes, either in A. luminosa or in general. I have tried to show that larvae must not be dismissed—as preparatory, undeveloped, or incomplete—by false analogy to a dubious (but socially favored) interpretation of human development. If any “higher reality” exists, we can only specify the life cycle itself. Larva and imago are but two stages of a totality—and you really can’t have one without the other. Eggs need hens as much as hens need eggs.

  I do try to show that child-adult is the wrong metaphor for understanding larva-imago. I have proceeded by discussing a case where larvae attract all our attention—literally as a source of beauty; structurally in greater size, length of life, and complexity of anatomy and behavior; and evolutionarily as focus of a major transformation from a simpler and very different ancestral style—while imagoes have scarcely modified their inherited form and behavior at all. But our proper emphasis on the larva of A. luminosa does not mark any superiority.

  We need another metaphor to break the common interpretation that degrades larvae to a penumbra of insignificance. (How many of you include maggot in your concept of fly? And how many have ever considered the mayfly’s longer larval life?) The facts of nature are what they are, but we can only view them through spectacles of our mind. Our mind works largely by metaphor and comparison, not always (or often) by relentless logic. When we are caught in conceptual traps, the best exit is often a change in metaphor—not because the new guideline will be truer to nature (for neither the old nor the new metaphor lies “out there” in the woods), but because we need a shift to more fruitful perspectives, and metaphor is often the best agent of conceptual transition.

  If we wish to understand larvae as working items in their own right, we should replace the developmental metaphor of child-adult with an economic simile that recognizes the basic distinction in function between larvae and imagoes—larvae as machines built for feeding and imagoes as devices for reproduction. Fortunately, an obvious candidate presents itself on the very first page of the founding document itself—Adam Smith’s Wealth of Nations. We find our superior metaphor in the title of Chapter 1, “On the Division of Labor,” and in Smith’s opening sentence:

  The greatest improvement in the productive powers of labor, and the greater part of the skill, dexterity, and judgment with which it is anywhere directed, or applied, seem to have been the effects of the division of labor.

  By allocating the different, sometimes contradictory, functions of feeding and reproduction to sequential phases of the life cycle, insects with complete metamorphosis have achieved a division of labor that permits a finer adaptive honing of each separate activity.

  If you can dredge up old memories of your first college course in economics, you will remember that Adam Smith purposely chose a humble example to illustrate the division of labor—pin making. He identifies eighteen separate actions in drawing the wire, cutting, pointing, manufacture of the head, fastening head to shaft, and mounting the finished products in paper for sale. One man, he argues, could make fewer than twenty pins a day if he performed all these operations himself. But ten men, sharing the work by rigid division of labor, can manufacture about 48,000 pins a day. A human existence spent pointing pins or fashioning their heads or pushing them into paper may strike us as the height of tedium, but larvae of A. luminosa encounter no obvious psychic stress in
a life fully devoted to gastronomy.

  Hobbyists and professional entomologists will, no doubt, have recognized an unintended irony in Smith’s selection of pin making to illustrate the division of labor. Pins are the primary stock-in-trade of any insect collector. They are used to fasten the dry and chitinous imagoes—but not the fat and juicy larvae—to collecting boards and boxes. Thus, the imagoes of A. luminosa may end their natural life caught in a larval web, but if they happen to fall into the clutches of a human collector, they will, instead, be transfixed by the very object that symbolizes their fall from conceptual dominance to proper partnership.

  Output from Ed Purcell’s computer program for arranging dots by the “stars,” ABOVE (random), and the “worms,” FACING PAGE (ordered by fields of inhibition around each dot), options. Note the curious psychological effect. Most of us would see order in the strings and clumps of the figure just above, and would interpret the figure on the opposite page, with its lack of apparent pattern, as random. In fact, the opposite is true, and our ordinary conceptions are faulty.

  Postscript

  Nothing brings greater pleasure to a scholar than utility in extension—the fruitfulness of a personal thought or idea when developed by colleagues beyond the point of one’s own grasp. I make a tangential reference in this essay to a common paradox—the apparent pattern of random arrays versus the perceived absence of sensible order in truly rule-bound systems. This paradox arises because random systems are highly clumped, and we perceive clumps as determined order. I gave the example of the heavens—where we “see” constellations because stars are distributed at random relative to the earth’s position. I contrasted our perception of heavenly order with the artificial “sky” of Waitomo Cave—where “stars” are the self-illuminated rear ends of fly larvae. Since these carnivorous larvae space themselves out in an ordered array (because they eat anything in their vicinity and therefore set up “zones of inhibition” around their own bodies), the Waitomo “sky” looks strange to us for its absence of clumping.

  My favorite colleague, Ed Purcell (Nobel laureate in physics and sometime collaborator on baseball statistics), read this tangential comment and wrote a quick computer program to illustrate the effect. Into an array of square cells (144 units on the X-axis and 96 on the Y-axis for a total of 13,824 positions), Purcell placed either “stars” or “worms” by the following rules of randomness and order (following the heavens versus the fly larvae of Waitomo). In the stars option, squares are simply occupied at random (a random number generator spits out a figure between 1 and 13,824 and the appropriate square is inked in). In the worms option, the same generator spits out a number, but the appropriate square is inked in only if it and all surrounding squares are unoccupied (just as a worm sets up a zone of inhibition about itself). Thus, worm squares are spaced out by a principle of order; star squares are just filled in as the random numbers come up.

  Now examine the patterns produced with 1,500 stars and worms (still less than 50 percent capacity for worms, since one in four squares could be occupied, and 3,456 potential worm holes therefore exist). By ordinary vernacular perception, we could swear that the “stars” program must be generating causal order, while the “worms” program, for apparent lack of pattern, seems to be placing the squares haphazardly. Of course, exactly the opposite is true. In his letter to me, Ed wrote:

  What interests me more in the random field of “stars” is the overpowering impression of “features” of one sort or another. It is hard to accept the fact that any perceived feature—be it string, clump, constellation, corridor, curved chain, lacuna—is a totally meaningless accident, having as its only cause the avidity for pattern of my eye and brain! Yet that is perfectly true in this case.

  I don’t know why our brains (by design or culture) equip us so poorly as probability calculators—but this nearly ubiquitous failure constitutes one of the chief, and often dangerous, dilemmas of both intellectual and everyday life (the essays of Section 9, particularly number 31 on Joe DiMaggio’s hitting streak, discuss this subject at greater length). Ed Purcell adds, emphasizing the pervasiveness of misperception, even among people trained in probability:

  If you ask a physics student to take pen in hand and sketch a random pattern of 1,500 dots, I suspect the result will look more like the “worms” option than the “stars.”

  18 | To Be a Platypus

  LONG AGO, garrulous old Polonius exalted brevity as the soul of wit, but later technology, rather than sweet reason, won his day and established verbal condensation as a form of art in itself. The telegram, sent for cash on the line and by the word, made brevity both elegant and economical—and the word telegraphic entered our language for a style that conveys bare essentials and nothing else.

  The prize for transmitting most meaning with least verbosity must surely go to Sir Charles Napier, who subdued the Indian province of Sind and announced his triumph, via telegram to his superiors in London, with the minimal but fully adequate “Peccavi.” This tale, in its own telegraphic way, speaks volumes about the social order and education of imperial Britain. In an age when all gentlemen studied Latin, and could scarcely rise in government service without a boost from the old boys of similar background in appropriate public schools, Napier never doubted that his superiors would remember the first-person past tense of the verb peccare—and would properly translate his message and pun: I have sinned.

  The most famous telegram from my profession did not quite reach this admirable minimum, but it must receive honorable mention for conveying a great deal in few words. In 1884, W. H. Caldwell, a young Cambridge biologist, sent his celebrated telegram from Australia to a triumphant reading at the Annual Meeting of the British Association in Montreal. Caldwell wired: “Monotremes oviparous, ovum meroblastic.”

  This message may lack the ring of peccavi and might be viewed by the uninitiated as pure mumbo jumbo. But all professional biologists could make the translation and recognize that Caldwell had solved a particularly stubborn and vexatious problem of natural history. In essence, his telegram said: The duckbilled platypus lays eggs.

  (Each word of Caldwell’s telegram needs some explication. Oviparous animals lay eggs, while viviparous creatures give birth to live young; ovoviviparous organisms form eggs within their bodies, and young hatch inside their mothers. Sorry for the jargon so early in the essay, but these distinctions become important later on. Monotremes are that most enigmatic group of mammals from the Australian region—including the spiny echidna, actually two separate genera of anteaters, and the duckbilled platypus, an inhabitant of streams and creeks. An ovum is an egg cell, and meroblastic refers to a mode of cleavage, or initial division into embryonic cells, after fertilization. Yolk, the egg’s food supply, accumulates at one end of the ovum, called the vegetal pole. Cleavage begins at the other end, called the animal pole. If the egg is very yolky, the cleavage plane cannot penetrate and divide the vegetal end. Such an egg shows incomplete, or meroblastic, cleavage—division into discrete cells at the animal pole but little or no separation at the yolky end. Egg-laying land vertebrates, reptiles and birds, tend to produce yolky egg cells with meroblastic cleavage, while most mammals show complete, or holoblastic, cleavage. Therefore, in adding “ovum meroblastic” to “monotremes oviparous,” Caldwell emphasized the reptilian character of these paradoxical mammals—not only do they lay eggs but the eggs are typically reptilian in their yolkiness.)

  The platypus surely wins first prize in anybody’s contest to identify the most curious mammal. Harry Burrell, author of the classic volume on this anomaly (The Platypus: Its Discovery, Position, Form and Characteristics, Habits and Life History, 1927), wrote: “Every writer upon the platypus begins with an expression of wonder. Never was there such a disconcerting animal!” (I guess I just broke tradition by starting with the sublime Hamlet.)

  The platypus sports an unbeatable combination for strangeness: first, an odd habitat with curiously adapted form to match; second, the real reason for its
special place in zoological history—its engimatic mélange of reptilian (or birdlike), with obvious mammalian, characters. Ironically, the feature that first suggested premammalian affinity—the “duckbill” itself—supports no such meaning. The platypus’s muzzle (the main theme of this column) is a purely mammalian adaptation to feeding in fresh waters, not a throwback to ancestral form—although the duckbill’s formal name embodies this false interpretation: Ornithorhynchus anatinus (or the ducklike bird snout).

  Chinese taxidermists had long fooled (and defrauded) European mariners with heads and trunks of monkeys stitched to the hind parts of fish—one prominent source for the persistence of mermaid legends. In this context, one can scarcely blame George Shaw for his caution in first describing the platypus (1799):

  Of all the Mammalia yet known it seems the most extraordinary in its conformation, exhibiting the perfect resemblance of the beak of a Duck engrafted on the head of a quadruped. So accurate is the similitude, that, at first view, it naturally excites the idea of some deceptive preparation.

  But Shaw could find no stitches, and the skeleton was surely discrete and of one functional piece (the premaxillary bones of the upper jaw extend into the bill and provide its major support). Shaw concluded:

  On a subject so extraordinary as the present, a degree of scepticism is not only pardonable but laudable; and I ought perhaps to acknowledge that I almost doubt the testimony of my own eyes with respect to the structure of this animal’s beak; yet must confess that I can perceive no appearance of any deceptive preparation…nor can the most accurate examination of expert anatomists discover any deception.

  The frontal bill may have provoked most astonishment, but the rear end also provided numerous reasons for amazement. The platypus sported only one opening, the cloaca, for all excretory and reproductive business (as in reptiles, but not most mammals, with their multiplicity of orifices for birth and various forms of excretion; Monotremata, or “one-holed,” the technical name for the platypus and allied echidna, honors this unmammalian feature).