Bully for Brontosaurus

by Stephen Jay Gould

  In this last respect [brainpower], monotremes are inferior to typical placental mammals and, probably, to typical marsupials. The paucity of living monotremes may therefore be due to their being less bright, less adaptable in their behavior, than other mammals.

  To cite just three studies among several of similar intent and conclusion:

  1. Saunders, Chen, and Pridmore (1971) ran echidnas through a simple two-choice T-maze (down a central channel, then either right or left into a bin of food or a blank wall). They trained echidnas to move in one direction (location of the chow, of course), then switched the food box to the other arm of the T. In such studies of so-called habit-reversal learning, most fish never switch, birds learn very slowly, mammals rapidly. Echidnas showed quick improvement with a steady reduction in errors—and at typically mammalian rates. Half the experiments (seven of fifteen) on well-trained echidnas yielded the optimal performance of “one-trial reversal” (you switch the food box and the animal goes the wrong way—where the food used to be—the first time, then immediately cottons on and heads in the other direction, toward the chow, each time). Rats often show one-trial reversal learning, birds never.

  2. Buchmann and Rhodes (1978) tested echidnas for their ability to learn positional (right or left) and visual-tactile (black and rough versus white and smooth) cues—with echidnas pushing the appropriate lever to gain their food reward. As an obvious testimony to mental adequacy, they report that “unsuccessful (unrewarded) responses were often associated with vigorous kicking at the operanda.” Echidnas learned at a characteristic rate for placental mammals and also remembered well. One animal, retested a month later, performed immediate one-trial reversals.

  Buchmann and Rhodes compared their echidnas with other animals tested in similar procedures. Crabs and goldfish did not show improvement (did not learn) over time. Echidnas displayed great variation in their speed of learning—one improved faster (and one slower) than rats; all echidnas performed better than cats. Take these results (and the reward for success as well) with a grain of salt because numbers are limited and procedures varied widely among studies—but still, the single best performer on the entire chart was an echidna.

  Buchmann and Rhodes conclude: “There is no evidence that the performance of echidnas is inferior to eutherian [placental] or metatherian [marsupial] mammals.” They end by ridiculing the “quaint, explicitly or tacitly-held views that echidnas are little more than animated pin-cushions, or, at the best, glorified reptiles.”

  3. Gates (1978) studied learning in visual discrimination (black versus white, and various complex patterns of vertical and horizontal striping). His results parallel the other studies—echidnas learned quickly, at typical mammalian rates. But he added an interesting twist that confutes the only serious, direct argument ever offered from brain anatomy for monotreme inferiority—the claim that absence of a corpus callosum precludes transfer between the cerebral hemispheres, thereby compromising “higher” mental functions.

  Gates occluded one eye and taught echidnas to distinguish black from white panels with the other eye. They reached “criterion performance” in an average of 100 trials. He then uncovered the occluded eye, bandaged the one that had overseen the initial learning, and did the experiment again. If no information passes from one cerebral hemisphere to the other, then previous learning on one side of the brain should offer no help to the other, and the 100-trial average should persist. But echidnas only needed 40 trials to reach criterion with the second eye.

  Gates conjectures that information is either passing across the other two commissures in the absence of a corpus callosum, or via the few optic fibers that do not cross to the other side of the brain. (In vertebrate visual systems, inputs from the right eye go to the left hemisphere of the brain, left eye to the right hemisphere; thus, each eye “informs” the opposite hemisphere. But about 1 percent of optic fibers do not cross over, and therefore map to their own hemisphere. These few fibers may sneak a little learning to the hemisphere dependent upon the occluded eye.) In addition, direct evidence of electrical stimulation has shown that inputs to one hemisphere can elicit responses in corresponding parts of the other hemisphere—information clearly gets across in the absence of a corpus callosum.

  The solution to the paradox of such adequate intelligence in such a primitive mammal is stunningly simple. The premise—the myth of primitivity itself—is dead wrong. To say it one more, and one last, time: The reptilian features of monotremes only record their early branching from the ancestry of placental mammals—and time of branching is no measure of anatomical complexity or mental status.

  Monotremes have evolved separately from placentals for a long time—more than enough for both groups to reach, by parallel evolution in independent lineages, advanced levels of mental functioning permitted by their basic, shared mammalian design. The primary evidence for parallel evolution has been staring us in the face for a century, forming part of the standard literature on echidnas, well featured even in primary documents that uphold the myth of primitivity. We know that the echidna’s brain attained its large size by an independent route. The platypus has a smooth (if bulbous) brain. The echidna evolved complex ridges and folds on its cerebral surface as a special feature of its own lineage. These sulci and gyri cannot be identified (homologized) with the well-known convolutions of placental brains. The echidna brain is so different, by virtue of a separate evolution to large size, that its convolutions have been named by Greek letters to avoid any misplaced comparison with the different ridges and folds of placental brains. And Grafton Elliot Smith, the man most puzzled by echidna brains, did the naming—apparently without realizing that the very need for such separate designations provided the direct evidence that could refute the myth of primitivity.

  In his eloquent plea for monotremes (1827), Geoffroy Saint-Hilaire wrote brilliantly about the subtle interplay of fact and theory in science. He recognized the power of theory to guide the discovery of fact and to set a context for fruitful interpretation. (“To limit our efforts to the simple practicalities of an ocular examination would be to condemn the activities of the mind.”) But he also acknowledged the flip side of useful guidance, the extraordinary power of theory to restrict our vision, in particular to render “obvious” facts nearly invisible, by denying them a sensible context. (“At first useless, these facts had to remain un-perceived until the moment when the needs and progress of science provoked us to discover them.”) Or as Warner Oland, the Swedish pseudo-Oriental Charlie Chan, once said in one of his most delightfully anachronistic pseudo-Confucian sayings (Charlie Chan in Egypt, 1935): “Theory like mist on eyeglasses. Obscure facts.”

  20 | Here Goes Nothing

  GOLIATH PAID THE HIGHEST of prices to learn the most elementary of lessons—thou shall not judge intrinsic quality by external appearance. When the giant first saw David, “he disdained him: for he was but a youth, and ruddy, and of a fair countenance” (1 Sam. 17:42). Saul had been similarly unimpressed when David presented himself as an opponent for Goliath and savior of Israel. Saul doubted out loud: “for thou art but a youth, and he a man of war from his youth” (1 Sam. 17:33). But David persuaded Saul by telling him that actions speak louder than appearances—for David, as a young shepherd, had rescued a lamb from a predatory lion: “I went out after him, and smote him, and delivered it out of his mouth” (1 Sam. 17:35).

  This old tale presents a double entendre to introduce this essay—first as a preface to my opening story about a famous insight deceptively clothed in drab appearance; and second as a quirky lead to the body of this essay, a tale of animals that really do deliver from their mouths: Rheobatrachus silus, an Australian frog that swallows its fertilized eggs, broods tadpoles in its stomach, and gives birth to young frogs through its mouth.

  Henry Walter Bates landed at Pará (now Belém), Brazil, near the mouth of the Amazon, in 1848. He arrived with Alfred Russel Wallace, who had suggested the trip to tropical jungles, arguing that a direct study of natu
re at her richest might elucidate the origin of species and also provide many fine specimens for sale. Wallace returned to England in 1852, but Bates remained for eleven years, collecting nearly 8,000 new species (mostly insects) and exploring the entire Amazon valley.

  In 1863, Bates published his two-volume classic, perhaps the greatest work of nineteenth-century natural history and travel, The Naturalist on the River Amazons. But two years earlier, Bates had hidden his most exciting discovery in a technical paper with a disarmingly pedestrian title: “Contributions to an Insect Fauna of the Amazon Valley,” published in the Transactions of the Linnaean Society. The reviewer of Bates’s paper (Natural History Reviews, 1863, pp. 219–224) lauded Bates’s insight but lamented the ill-chosen label: “From its unpretending and somewhat indefinite title,” he wrote, “we fear [that Bates’s work] may be overlooked in the ever-flowing rush of scientific literature.” The reviewer therefore sought to rescue Bates from his own modesty by providing a bit of publicity for the discovery. Fortunately, he had sufficient oomph to give Bates a good send-off. The reviewer was Charles Darwin, and he added a section on Bates’s insight to the last edition of the Origin of Species.

  Bates had discovered and correctly explained the major style of protective mimicry in animals. In Batesian mimicry (for the phenomenon now bears his name), uncommon and tasty animals (the mimics) gain protection by evolving uncanny resemblance to abundant and foul-tasting creatures (the models) that predators learn to avoid. The viceroy butterfly is a dead ringer for the monarch, which, as a caterpillar, consumes enough noxious poisons from its favored plant foods to sicken any untutored bird. (Vomiting birds have become a cliché of natural history films. Once afflicted, twice shy, as the old saying goes. The tale may be more than twice told, but many cognoscenti do not realize that the viceroy’s name memorializes its mimicry—for this butterfly is the surrogate, or vice-king, to the ruler, or monarch, itself.)

  Darwin delighted in Bates’s discovery because he viewed mimicry as such a fine demonstration of evolution in action. Creationism, Darwin consistently argued, cannot be disproved directly because it claims to explain everything. Creationism becomes impervious to test and, therefore, useless to science. Evolutionists must proceed by showing that any creationist explanation becomes a reductio ad absurdum by twists of illogic and special pleading required to preserve the idea of God’s unalterable will in the face of evidence for historical change.

  In his review of Bates’s paper, Darwin emphasizes that creationists must explain the precision of duplicity by mimics as a simple act of divine construction—“they were thus clothed from the hour of their creation,” he writes. Such a claim, Darwin then argues, is even worse than wrong because it stymies science by providing no possible test for truth or falsity—it is an argument “made at the expense of putting an effectual bar to all further inquiry.” Darwin then presents his reductio ad absurdum, showing that any fair-minded person must view mimicry as a product of historical change.

  Creationists had made a central distinction between true species, or entities created by God, and mere varieties, or products of small changes permitted within a created type (breeds of dogs or strains of wheat, for example). But Bates had shown that some mimics are true species and others only varieties of species that lack mimetic features in regions not inhabited by the modeL Would God have created some mimics from the dust of the earth but allowed others to reach their precision by limited natural selection within the confines of a created type? Is it not more reasonable to propose that mimicking species began as varieties and then evolved further to become separate entities? And much worse for creationists: Bates had shown that some mimicking species resemble models that are only varieties. Would God have created a mimic from scratch to resemble another form that evolved (in strictly limited fashion) to its current state? God may work in strange ways, his wonders to perform—but would he really so tax our credulity? The historical explanation makes so much more sense.

  But if mimicry became a source of delight for Darwin, it also presented a serious problem. We may easily grasp the necessity for a historical account. We may understand how the system works once all its elements develop, but why does this process of mimicry ever begin? What starts it off, and what propels it forward? Why, in Darwin’s words, “to the perplexity of naturalists, has nature condescended to the tricks of the stage?” More specifically: Any butterfly mimic, in the rich faunas of the Amazon valley, shares its space with many potential models. Why does a mimic converge upon one particular model? We can understand how natural selection might perfect a resemblance already well established, but what begins the process along one of many potential pathways—especially since we can scarcely imagine that a 1 or 2 percent resemblance to a model provides much, if any, advantage for a mimic. This old dilemma in evolutionary theory even has a name in the jargon of my profession—the problem of the “incipient stages of useful structures.” Darwin had a good answer for mimicry, and I will return to it after a long story about frogs—the central subject of this essay and another illustration of the same principle that Darwin established to resolve the dilemma of incipient stages.

  We remember Darwin’s Beagle voyage primarily for the big and spectacular animals that he discovered or studied: the fossil Toxodon and the giant Galápagos tortoises. But many small creatures, though less celebrated, brought enormous scientific reward—among them a Chilean frog appropriately named Rhinoderma darwini. Most frogs lay their eggs in water and then allow the tadpoles to make their own way, but many species have evolved various styles of parental care, and the range of these adaptations extols nature’s unity in diversity.

  In R. darwini, males ingest the fertilized eggs and brood them in the large throat pouches usually reserved for an earlier act of courtship—the incessant croaking that defines territory and attracts females. Up to fifteen young may fill the pouch, puffing out all along the father’s ventral (lower) surface and compressing the vital organs above. G. B. Howes ended his classic account of this curious life-style (Proceedings of the Zoological Society of London, 1888) with a charming anthropomorphism. Previous students of Rhinoderma, he noted, had supposed that the male does not feed while carrying his young. But Howes dissected a brooding male and found its stomach full of beetles and flies and its large intestine clogged with “excreta like that of a normal individual.” He concluded, with an almost palpable sigh of relief, “that this extraordinary paternal instinct does not lead up to that self-abnegation” postulated by previous authors.

  But nature consistently frustrates our attempts to read intrinsic solicitude into her ways. In November 1973, two Australian scientists discovered a form of parental care that must preclude feeding, for these frogs brood their young in their stomachs and then give birth through their mouths. And we can scarcely imagine that a single organ acts as a nurturing uterus and a site of acid digestion at the same time.

  Rheobatrachus silus, a small aquatic frog living under stones or in rock pools of shallow streams and rills in a small area of southeast Queensland, was first discovered and described in 1973. Later that year, C. J. Corben and G. J. Ingram of Brisbane attempted to transfer a specimen from one aquarium to another. To their astonishment, it “rose to the surface of the water and, after compression of the lateral body muscles, propulsively ejected from the mouth six living tadpoles” (from the original description published by Corben, Ingram, and M. J. Tyler in 1974). They initially assumed, from their knowledge of Rhinoderma, that their brooder was a male rearing young in its throat pouch. Eighteen days later, they found a young frog swimming beside its parent; two days later, a further pair emerged unobserved in the night. At that point, they decided (as the euphemism goes) to “sacrifice” their golden goose. But the parent, when grasped, “ejected by propulsive vomiting eight juveniles in the space of no more than two seconds. Over the next few minutes a further five juveniles were ejected.” They then dissected the parent and received their biggest surprise. The frog had no vocal sac.
It was a female with “a very large, thin-walled, dilated stomach”—the obvious home of the next generation.

  Natural birth had not yet been observed in Rheobatrachus. All young had either emerged unobserved or been vomited forth as a violent reaction after handling. The first young had greeted the outside world prematurely as tadpoles (since development clearly proceeds all the way to froghood in the mother’s stomach, as later births demonstrated).

  Art then frustrated nature, and a second observation also failed to resolve the mode of natural birth. In January 1978, a pregnant female was shipped express airfreight from Brisbane to Adelaide for observation. But the poor frog was—yes, you guessed it—“delayed” by an industrial dispute. The mother, still hanging on, eventually arrived surrounded by twenty-one dead young; a twenty-second frog remained in her stomach upon dissection. Finally, in 1979, K. R. McDonald and D. B. Carter successfully transported two pregnant females to Adelaide—and the great event was finally recorded. The first female, carefully set up for photography, frustrated all hopes by vomiting six juveniles “at great speed, flying upwards…for approximately one meter…a substantial distance relative to the body size of the female.” But the second mother obliged. Of her twenty-six offspring, two appeared gently and, apparently, voluntarily. The mother “partially emerged from the water, shook her head, opened her mouth, and two babies actively struggled out.” The photo of a fully formed baby frog, resting on its parent’s tongue before birth, has already become a classic of natural history. This second female, about two inches long, weighed 11.62 grams after birth. Her twenty-six children weighed 7.66 grams, or 66 percent of her weight without them. An admirable effort indeed!

  Rheobatrachus inspired great excitement among Australian scientists, and research groups in Adelaide and Brisbane have been studying this frog intensively, with all work admirably summarized and discussed in a volume edited by M. J. Tyler (1983). Rarely has such extensive and coordinated information been presented on a natural oddity, and we are grateful to these Australian scientists for bringing together their work in such a useful way.