Reef Madness: Charles Darwin, Alexander Agassiz, and the Meaning of Coral - David Dobbs (2005)

Part III

Chapter 17. Entwetok

GEORGE AND MAX never discovered whether the drafts their father spoke of existed only in his head, as George half-seriously suggested, or had been roughed out and discarded. The latter seems more likely. Alexander Agassiz tended to purge anything not meant for many eyes.

His failure to publish did not mean his ideas on reefs were lost, for his theory is evident from his publications and letters. But it meant he never entered a concise form of it into the coral reef debate. That argument was a half century old when Alexander died and would simmer another half century before being resolved. Might he have changed it had he delivered? It seems likely. Even in scattered form, Alex’s theory exerted considerable influence; a succinct, well-argued version might have greatly magnified this influence, tilting further his way the balance that seesawed between Darwin’s theory and his.

As it was, despite his reticence, Alexander Agassiz (along with others advocating theories similar to his) had already made their the ory a near-equal to Darwin’s. The editors of the renowned 1911 Encyclopedia Britannica, for instance, saw fit to assign the coral reef entry to G. C. Bourne, the Oxford geologist whose examination of the Chagos archipelago (south of the Maldives) had made him a Darwin doubter back in the 1870s. Bourne cited Agassiz’s work as proof against the universality of Darwin’s theory. Expressing what the edi- tors presumably considered a judicious view of informed opinion, Bourne stated that in the face of Agassiz’s findings

it must be admitted that the subsidence theory of Darwin is inapplicable to a large number of coral reefs and islands. … In the present state of our knowledge it seems reasonable to conclude that coral reefs are formed wherever the conditions suit able for growth exist, whether in areas of subsidence, elevation, or rest… [and] the atoll or barrier reef shape is not necessarily evidence of formation during subsidence, [but] may be produced by the natural growth of coral, modified by the action of waves and currents in regions in which subsidence has certainly not taken place.

Eniwetok Atoll, in the Marshall Islands. Almost a half century after Alexander died, this small atoll, so classic in form, would yield critical information about the questions he had spent a quarter century investigating.

Bourne voices Alexander’s main contention: Reefs take hold and grow not because they are on sinking bottoms but because depth and temperature allow, and their forms are created by erosion and local conditions. If Alex had not driven Darwin from the field, he at least had forced him to share it.

After Alex died, two main figures emerged to continue this debate, a third to complicate it.

Taking Alexander’s side was J. Stanley Gardiner, the young British researcher who had just preceded Alex in both the Fijis and the Maldives. From 1897 to 1907 Gardiner saw as many reef areas as ever seen by anyone save Alex and perhaps Dana, and this, along with his fair-minded, lucid writing, made him one of his generation’s most credible reef theorists. He advocated a theory in which three foundation-building dynamics operated in various places and pro portions to build the world’s reef platforms: a Murrayesque accretion of sediment; an Agassizian erosion of upraised islands or landmasses; and Darwinian subsidence, which as a direct genitor of atolls, he said, “has probably been a purely local phenomenon.” Gardiner eventually began to doubt that any of these theories could be applied with confidence, given the ambiguous and superficial nature of the observable evidence. As he put it in his 1930 Coral Reefs and Atolls,“If we regard the question of the formation of the foundation of coral reefs honestly, we are forced to admit that all our theories and considerations are mere camouflage for our lack of knowledge.” In the 1910s and early 1920s, however, he advocated most strongly the erosion-elevation element of his trio, and he always agreed with Alex that Darwin’s theory was overextended. Through his steady pressure in the 1910s and 1920s, Gardiner forced among scientists a grudging consensus that this was probably the case. In the 1930s, as Gardiner grew doubtful, the elevation-erosion theory was taken up by Harry Ladd and J. Edward Hoffmeister, among the most prominent reef scientists of their generation, as the “antecedent platform theory,” which held that “any bench or bank … at the proper depth within the circum-equatorial coral reef zone can be considered a potential reef foundation, and … if ecological conditions permit, a reef could grow up to the surface without any progressive change in ocean-level”-an expression remarkably close to Alex’s phrasings.

Meanwhile emerged a new champion for subsidence. William Davis, fifteen years younger than Alex and twenty-two years older than Gardiner, had taught geology and geography at Harvard since 1885 but did not begin to study coral reefs intently until 1912, when he resigned his Harvard professorship at age sixty-two. A year later, at about the time he published a paper highlighting how Dana’s embayed-valley insight supported Darwin, Davis’s wife died. The next year, still grieving, he took a five-month tour of the South Pacific islands. He had formerly found a roughly evolutionary model helpful when studying the development of rivers, and now, like Dana, he was stunned by the explanatory power of Darwin’s coral reef theory. From then until his death in 1935, he championed an aggressively deductive application of the theory, producing more than fifty publications culminating in the six-hundred-page Coral Reef Problem in 1928. Davis advocated what he quite seriously called the “home study” of coral reefs, insisting that the key to explaining coral reefs lay not in the reefs themselves but in large patterns of the sort visible on maps viewable in any living room. Darwin himself, of course, had done most of his theorizing over maps and charts (as Alex often noted with disdain) and considered his map showing the distribution of the Pacific atolls one of his strongest proofs.

But Darwin had worked when direct reef observations were few. Davis worked when more than two dozen experienced researchers had reported closely on reefs. His insistence that their evidence was secondary struck some as perverse. (One begins to understand why he didn’t take this up while Alex was around.) He happily let powerful abstractions overrule observation and treated the latter as important when it supported subsidence but irrelevant otherwise. A later historian deemed his approach “research by debate.” Gardiner could not abide this. In his 1930 Coral Reefs and Atolls, he dismissed Davis, who for two decades had been the most prolific writer on reefs aside from Gardiner, by mentioning him only twice. He led off his more lengthy mention (a paragraph) by calling Davis’s Coral Reef Problem the book “wherein will be found almost all known facts that can be made to favour subsidence.”

It wasn’t quite the Duke versus Huxley. But these two and their allies kept the coral reef controversy very much alive through the 1930s. A full century after it was posed, the debate over Darwin’s the- ory still expressed an ambiguous body of evidence and a sharp divide over how to do science.

One more figure comes into play here. Reginald Daly Canadian born, earned a Harvard geology doctorate in 1896 at age twenty-four, instructed there from 1898 to 1901, and after six years in the field with the Canadian International Boundary Commission and four teaching at the Massachusetts Institute of Technology returned to Harvard in 1912, at forty-one, to take Davis’s place as the Sturgis Hooper Professor of Geology. He stayed another thirty years. Like Darwin, Daly was an avid field worker who valued imaginative, synthetic the orizing. In 1910 he offered a theory that threatened to completely rearrange things-and indeed would, though ultimately along lines different from what he expected.

In a 1909 visit to Hawaii, Daly had connected two key observations. First, he noticed that some of the high volcanic slopes bore the scrapes, gouges, and other marks of-of all things-glaciers. What Louis couldn’t find in Brazil, Daly had found in Hawaii. This did not mean, as Louis had absurdly argued after finding much less credible glacial signs in Brazil, that glaciers once covered every inch of the globe. Hawaii was twenty degrees north of the equator, for one thing, and the marks were at high elevations; snow still occasionally fell on these peaks, so it wasn’t stunning to find signs of former glaciers there. But this direct evidence made it clear that Hawaii had once been much cooler.

To Daly, this helped explain another thing that struck him: the relatively narrow width of the Hawaiian reefs, which suggested that they weren’t very old, not nearly as old as the islands seemed. On the evidence of the glacial marks, he proposed that glacial periods had often cooled the water along the margins of the equatorial coral belt to below coral-friendly temperatures, temporarily halting coral reef growth there. He also knew, from the work of Alexander Agassiz and others, that many reefs in Hawaii and elsewhere showed signs of significant erosion-platforms, notches, and other irregularities of the sort often created by wave erosion and beaching-in the first thirty or forty fathoms below the surface. Taking his ice age insight a step further, he proposed that many glacial periods had consolidated enough seawater into polar ice caps and glaciers to drop the sea level as much as forty or fifty fathoms, exposing the reefs to the erosive power of waves at the same time the cooling water halted coral growth. When the ice ages ended, the waters would slowly rise and warm, and the reefs, altered by erosion, would resume growing. These repeated dunkings and slow risings allowed the reefs to periodically grow both upward and outward, thickening and broadening every cycle. In short, it wasn’t sinking foundations that created the reefs; it was a slowly rising sea, or rather a series of sea rises.

This shook things up. It explained every subsidence-like effect not with an unproveable subsidence but via a glacial theory that was backed by bounteous evidence and almost a century of acceptance; and it explained oddities of reef form that Darwin’s theory could not explain. Subsidence advocates didn’t like this. They had trouble finding ground from which to attack the theory, however, for its chain of deductions was nearly as imaginative and elegant as Darwin’s while resting on more empirical support. It was in fact the sort of theory Darwin would have loved. (One can imagine him dancing around a room on hearing it.) But some critics questioned Daly’s estimates of the sea-level oscillations, and others thought he overstated the foundations’ vertical stability. Davis, for instance, dismissed the theory as “largely invalidated by the evidence … of island instability” But most subsidence advocates admitted the theory was formidable, and quite a few fence sitters moved to Daly’s side.

The theory’s implications for the Agassiz-Gardiner erosion-elevation theory were less clear. It seemed at once to contradict it, support it, and make it moot. From one way of looking at it, Daly was arguing that reefs were built much as Darwin had proposed, only with the sea rising (repeatedly) instead of the reef foundations falling-the same dance, but led by ocean rather than land. Yet you could also say that Daly’s idea supported Agassiz. It accepted Alex’s preexisting or elevated platforms and alternately brought them the two reef-shaping forces he emphasized: reef-friendly conditions in warm periods and erosive forces in cooler eras. Still, some wondered whether Daly bolstered Agassiz or simply replaced him. Daly himself saw his theory as a great boost to Agassiz’s, calling the glacial-control hypothesis “a missing link in the chain of argument used by Semper, Rein, Murray, Agassiz, and Guppy against the wholesale-subsidence hypothesis.” “Correlating ice-caps and coral reefs, we use the great discovery of Louis Agassiz to support a principal conclusion of Alexander Agassiz.” In the decades that followed, however, many saw Daly’s theory as standing aside from Agassiz and Gardiner’s.

So went the debate through the first half of the twentieth century. Three well-supported theories-the first and second mutually exclusive and a third seeming to replace the first and support the sec ond-jockeyed for command over the same ambiguous body of findings. Clearly some definitive new piece of evidence was needed to settle the issue.

It had long been recognized that such evidence might be obtained by shooting a clean bore all the way through a coral reef’s layers to “basement,” or the original basalt that everyone agreed ultimately underlay all reef systems. But this peek under the floor proved elusive. The drills of the 1910s and 1920s couldn’t go deep enough, and the few borings from those years improved little on the Funafuti results. Drilling technology advanced during the 1930s, but Japan had by then declared the Pacific off limits, so the drills couldn’t be put to work where it mattered most. Reef science reached a sort of stasis. It was playing out nicely the prediction of geologist Arthur Holmes, who wrote in a 1916 review of Daly’s work that “controversy will rage for many years on the problem … presented by the origin of coral reefs, for at present the data are still insufficient to provide an adequate basis for a complete explanation.”

The chance to drill deep finally came in 1950. The U.S. Navy, preparing to test nuclear bombs on several atolls in the Marshall Islands, assembled a team of reef geologists to do some deep core drilling as part of an elaborate pretest environmental survey of the area. The geologists, led by the United States Geological Survey’s Harry Ladd (who with Hoffmeister had advocated an Agassizian “antecedent platform” theory since the 1930s), set up drills at opposite ends of Eniwetok, a circular atoll at the Marshalls’ western edge. Drilling technology had advanced since Funafuti; the drills cut through the limestone with relative ease, bringing up one-and-a-half-inch cores. The results must have been sobering for Ladd and any other subsidence doubters present. The first cores were clearly reef rock, as expected. As the drill passed the first few hundred feet and out of coral reef depth, the cores changed little. They still appeared to be reef rock. Both the immediate shipboard and later laboratory analyses found that the granular stone held fossils of foraminifera, corals, and other life that grew only in shallow water, confirming that this was coral limestone. So it went as the drills cut deeper-500 feet, 1,000, 2,000, 3,000, 4,000. Finally, at 4,200 feet, the drills hit what was unequivocally basement, a greenish basalt, the volcanic mountain on which the reef had originated. Dating of the tiny fossils in the bottommost layer of coral showed that the reef had gotten its start in the Eocene. For more than thirty million years this reef had been growing-an inch every millennium-on a sinking volcano, thickening as the lava beneath it subsided. Darwin was right, Agassiz wrong.

As data from the Marshalls and elsewhere piled up over the months and years that followed, the balance fell decisively to Dar win’s side. Drillings in the Tuamotus, at the other end of the South Sea Isles, produced results similar to those in Eniwetok, as did borings in the Caribbean and the Maldives, where a drill shot through sixty-five hundred feet of coral to basalt. Newly refined seismic techniques (first developed in the 1930s) established similar profiles in virtually all the Pacific archipelagoes. Echo soundings, which could continuously read the bottom topography, provided further proof of subsidence, showing that almost all these atolls were on steep-sided platforms, with valleys thousands of feet even between some closely spaced atolls; clearly these had not grown on available banks but had risen slowly atop their own thickening footprints. One after another of the world’s reef areas-the Marshalls, the Ellice and Gilberts, the Taumotos, the Fijis and Hawaii, even most of the Caribbean reef sys tems-was found to follow Darwin’s model. Only Florida, where Alexander started his quest and gathered his first impressions, showed itself to be built on the sort of current-shaped bank he proposed.

As for Daly, his glacial-control theory was found not to contradict Darwin’s model but to complement it. New methods of sea-level analysis showed that the ocean oscillated even more than Daly envisioned. Over the previous two million years alone, the tropical oceans had undergone numerous oscillations of up to one hundred fathoms and temperature swings of five to ten degrees Fahrenheit. Even as the sea bottom slowly dropped, these fluctuations created multiple surges of coral growth separated by intervals in which waves created platforms, notches, and other erosive forms. The most recent high-water drowning of reefs probably occurred in the Holocene, only eight thousand years ago. The reef growth since then accounts for the marked difference, emphasized so strongly by Alexander, between modern, contiguous-looking corals and the underlying lime stones. The window of growth for present-day reefs was far shorter than Alexander or anyone else then imagined.

The structure was Darwin’s, the topography Daly’s. Separately, neither theory could explain all reef areas; together they accounted for almost all of them. Today this synthesis, sometimes called the Darwin-Daly theory, is seen to fit hand in glove with the theory of plate tectonics developed in the 1960s, for it is the movement of the earth’s huge plates that explains the subsidence of the Pacific and many other reef areas. Darwin’s theory seemed to be able to take in anything. He had ignored a few details that had to be explained by Daly. But he had gotten the big picture astoundingly correct.

We needn’t ask how Alex missed all this. He faced an extraordinarily difficult observational and theoretical problem, and it was another half century before technology could reveal the evidence needed to solve it decisively. With a few painful exceptions, he read the land scape in a defensible way, and his interpretations were shaped by an empiricism true-perhaps too-to the tenets of his era. “Study nature, not books,” his father had famously urged, and Alex took to heart this elevation of observation over theory. He was determined that if he should err, it would be toward known facts and away from the big picture.

As to the painful exceptions, it’s clear now, after another century of reef science, that Alexander made four crucial field errors that allowed his misreading of atolls. Two were understandable. Two are almost inexplicable.

The former first. Alex erred when he read the raised limestone beds of Fiji and other Pacific atolls as old sedimentary limestone rather than more recent reef limestone. His reading was encouraged by the rock’s clearly uplifted state (limestone can’t be formed above water); the marked difference between the chunky, coral-filled veneer and the finer conglomerate beneath it; and the reef rock’s extreme similarity to older, marine limestone. This key misreading suggested to Alex a relatively large uplift (since marine limestones typically form in deeper water than coral limestones) and let Alex discard the implication of subsidence posed by a reading of the limestone as reef rock.

He also overestimated the effect and duration of subsurface erosion. Alex wasn’t alone here; many of his colleagues struggled to explain the platforming, notching, and other erosive forms that complicated the idealized picture of smoothly contoured reefs, and no one really succeeded until Daly. Alex read these erosive patterns as evidence that the reef banks had been in place long enough to allow currents to create not just the topographical irregularities but the shapes of the reef foundations.

Both of these errors-his emphasis on the distinction between surface coral and underlying limestone and his vision of long periods of subsurface erosion-were attempts to interpret features that would be clearly explained only by Daly’s glacial-control theory. Both were misreadings that other skilled observers might and did make.

More troubling are a pair of major errors that defy explication. One, the more amateurish, he made repeatedly in the lagoons of the archipelagoes that most strongly suggested subsidence (the Tuamo-tus in the east and the Ellice, the Gilberts, and the Marshalls in the north), where he mistook old, storm-deposited slabs of reef lime stone, blackened by sun and algae and welded by calcification to the lagoon floors, for remnants of eroded volcanic islands. This allowed his mischaracterization of these islands as composed largely of volcanic rock, suggesting an island of lava that rose to the surface, rather than thickening reef limestone, which would indicate a slowly sinking island composed mainly of coral. The error astonishes trained geologists today. It seems the sort of mistake a neophyte or student might make. How did Alexander misread these slabs? It’s possible he sampled them and made a misidentification. But it seems more likely that he looked too briefly. He kept a brisk pace through much of his longer Pacific trip and devoted much of the time to soundings. Rushed and having drawn fairly firm conclusions from his Fiji trip, he appears to have taken the shortcut for which he so disdained Dar- win and Dana: From the deck of a boat at too great a distance, he observed poorly and concluded hastily.

His other inexplicable error was grossly misinterpreting his own soundings. He took considerable trouble to measure scores of depths in all the Pacific archipelagoes, yet he overlooked implications from the results that might have struck others as obvious. In almost all the archipelagoes, he found ravines up to forty-five hundred feet deep between many of the atolls-a finding that supported Darwin’s the ory and argued strongly against his own notion that most atolls grew from raised areas on broad banks. Yet he discounted these ravines by considering them very deep channels cut by current and chose to emphasize instead the shallower, inter-atoll banks he also discovered, which “go to show,” as he wrote Murray, “that those atolls are not so immensely steep but rise from a great plateau.” He was practicing a Davis-like selectivity here, filtering his evidence to yield the interpretation most favorable to his own view.

That he would look too briefly, that he would strain evidence- these failures are harder to rationalize than his misreadings of geology. Indeed you can’t rationalize them. Alex fell to the same fault he found so unforgivable in others: He saw what he wanted to see.

Was it some troubled awareness of this failure that kept him from writing up his theory? Did some part of him, realizing he was pushing the facts out of shape, balk at the sin he seemed to feel greatest, that of theorizing too soon and with too little proof? In 1904 he complained to Murray that a certain geologist carrying on an argument with Gardiner in the pages of Nature“is the kind of man who knows it all and has seen nothing of coral reefs. It really seems as if the less a man has seen the more anxious he is to write.” Perhaps at some level he recognized, as Gardiner would later, that there weren’t really enough facts to go on.

It’s stretching things, of course, to make his failure to publish his theory a virtue, the happy result of a healthy inhibition. That relieves him of the job he set out to do, which was to see reefs clearly and describe their genesis. If he saw the evidence as inconclusive, he should have said so. He can’t be absolved of the obligation he carried, as so prominent, privileged, and experienced a scientist, as the man who had seen more reefs than anyone else, to declare his findings.