by Sarah Dry
These men—today we would call them physicists, though the term was only just starting to be used—were obsessed with energy, and in particular with the transformation of heat into a form of energy that could be harnessed as work. But though their working environments were far removed from the sites of coal mines and railway cuttings, commerce played a role in their research as surely as it did in the discoveries of the geologists. The industrial revolution powered itself quite literally with the heat of the sun that was stored up in the black deposits of coal. The heat of the sun and the work it could do motivated the theoretical calculations these men made and the experiments they did on all manner of things. They studied the behavior of metal under pressure (a key factor in designing boilers that did not explode), the way a certain amount of work always generated a certain corresponding rise in temperature, and how to design steam engines that were as efficient as possible. This research led them to make predictions about how heat acted not only in the workshop or the laboratory but in the crevices and crannies of the earth itself. Deduced as they were from equations that described the behavior of energy and matter, the conclusions of men like the Thomsons furnished a kind of mathematical bedrock for the new sciences of the earth along with the increasingly consequential industries seeking ever more efficient ways to transform coal, via steam engines, into work.
One of the key ideas to emerge from this new way of thinking was that the universe, and everything in it, including the earth, was gradually and inexorably cooling down. This so-called heat death of the universe was a depressing fact of life, against which everything (with the important exception of God himself) seemed powerless. It followed that the earth’s past must consist of a very steady, very boring, and very uniform cooling like that of a long-forgotten cup of coffee. Earth, in other words, could only have been warmer in the past, not colder. Recently unearthed fossil remains of coral reefs, seashells such as the (now tropical) pearly nautilus, and warm-weather plants such as palm trees and cycads that had been found in Northern Europe provided further, seemingly incontrovertible evidence that the earth had been warmer, not colder, in the recent past.10 Taking into account the mathematical calculations of the physicists and the fossil evidence of warmer climates in the past, the cooling earth theory seemed an inescapable fact.
But the earth provided contradictory signs. Tumbling out of the crumbling walls of railway cuts and the dank tunnels of the coal mines came evidence that supported Agassiz’s theory of the ice ages. How, then, to reconcile the seeming incompatibility between these different kinds of evidence about the earth’s past? Some suggested that the continents themselves had been elevated in the past, allowing ice sheets to form at cooler altitudes and leaving the overall climate of the earth unchanged. But as the geological evidence accumulated for just how widespread ice sheets had been, this view became increasingly untenable. William Hopkins, a highly skilled mathematician who coached dozens of the best mathematicians of his generation as a tutor at Cambridge, calculated that it would be necessary to lift the entire continent of Europe above 10,000 meters to account for the ice age this way, a circumstance which, he added, “all geological experience assures us would be impossible without leaving numerous telltale signs which do not presently exist.”11 Nevertheless, the contradictory evidence of both a colder and warmer past on the planet demanded a new explanation. Hopkins had a mathematical answer to the dilemma. While it was true that the interior of the earth was cooling, Hopkins’s calculations showed that it had already cooled to such an extent that the central or “primitive” heat of the earth (leftover from its fiery formation) was contributing a vanishingly small amount of heat to the earth’s surface—just one-twentieth of a degree.12 If the leftover heat contributed such a small proportion to the earth’s surface temperature, the “problem” of heat death as a factor in terrestrial climate more or less disappeared. No longer did geologists have to find a way to link changes in the earth’s climate at its surface to the slow cooling of its molten core. Hopkins confidently, but somewhat unhelpfully, declared: “We must manifestly seek for other causes to account for the changes of temperature which mark the more recent geological periods.”13 Whatever had caused the changes in the earth’s climate that had occurred relatively recently, it was not the effect of the earth’s cooling core. It remained to be seen what those other causes might be.
Hopkins had resolved the matter of the earth’s cold recent past in relation to the overall cooling of its interior, a once insurmountable objection. By 1859, the year that Tyndall and his men braved the blizzard on the Mer de Glace, Agassiz’s ice age theory was largely accepted, but two important questions remained unanswered. First, a theory was needed to explain how, precisely, the glaciers and ice sheets that Agassiz imagined covering so much of the Northern Hemisphere had moved. Second, and even more fundamentally, scientists still lacked an explanation for what had caused the earth’s climate to cool so dramatically in the past. To answer the first of these questions, it was necessary to go, as Tyndall had, to the ice and study its motions. To answer the second would require traveling farther still, beyond the bounds of the earth and into the solar system itself.
* * *
In the late 1850s and early 1860s, a Scot named James Croll, an almost exact contemporary of Tyndall, was working quietly as a caretaker at a small college and museum in Glasgow. Unlike Tyndall, who had, by his thirties, attained one of the few full-time professorships in physics in Britain and with it a measure of personal fame, Croll was unknown. He had not had any education to speak of, but had discovered books as a young child anyway. Soon, a passion for natural philosophy emerged. This he nurtured even as he worked a series of increasingly odd jobs, including the tending of a tea shop for which the taciturn former carpenter was astonishingly ill suited. For nearly three decades, he passed his time reading, independently and widely, in the process developing a taste for theoretical rather than empirical works. Facts alone held little interest for him. He wanted the warp and woof of theory that held the facts—the world—together. In the early 1860s, taking advantage of his light work duties, he immersed himself in the study of what he called “the then modern principle of the transformation and conservation of energy and the dynamical theory of heat,” reading the works of Tyndall alongside those of Faraday, Joule, and William Thomson on heat, electricity, and magnetism. At the same time, he followed the unfolding developments in the debates over the “question of the cause of the Glacial epoch.”14
An autodidact with an idiosyncratic cast of mind, Croll was an outsider to these debates, with neither institutional affiliation nor professional training in any of the subjects. His remove gave him a perspective—and a freedom—that enabled him to make his greatest cognitive leap. In 1864, he published a paper arguing that the causes of changes in the earth’s climate are to be found nowhere on earth. Instead, Croll believed that it was in the subtly wobbly dance of the earth around the sun that the cause of the ice ages was to be found. The plural here is important—part of the reason Croll reached outside the planet was to explain why he thought there was not just one ice age in the past, but a series of alternating glacial and warmer interglacial periods (evidence for alternating glacial and interglacial periods had recently been uncovered by Archibald Geikie in the form of layers of organic matter found in deposits of glacial drift). Before Croll, eminent scientists had considered this same astronomical possibility, including Alexander von Humboldt, Charles Lyell and, most influentially, the astronomer John Herschel. Herschel had noted that given the long-term gravitational perturbations affecting the earth, there would be moments, when its orbit was especially elliptical (or squashed), that winters would be longer and summers shorter. But Herschel quickly neutralized the implications of this fact in accounting for the ice ages by pointing out that the overall amount of sunlight hitting the earth would always be the same—the long winters would be compensated for by very hot summers.
FIG. 2.6. James Croll, a sel
f-taught theorist of climate, impressed Charles Darwin and John Tyndall with his argument that changes in astronomical cycles had produced multiple ice ages on the planet via “secondary causes” on Earth, such as variations in the reflection of sunlight by ice and cloud and corresponding shifts in winds and ocean currents.
Croll’s innovation was twofold. First of all, he discounted more or less all the evidence of the geologists. He was blunt about his lack of interest in what he called the “facts and details” of science in favor of the more attractive (to him) basic “laws or principles” which underlay the empirical facts. (He remarked, positively, about a job he’d held as a geologist in the civil service that “really did not require much acquaintance with the science of geology” and therefore “relieved my mind from having to study a science for which I had no great liking, and thus allowed me to devote my whole leisure hours to those physical questions in which I was engaged.”)15 Croll was an unrepentant big-picture thinker. Having dispensed with the constricting assumptions of the geologists, confined to the lowering and rising of continents or the comings and goings of floodwaters and the pesky details of the scattered evidence, the second part of his innovation was to grasp at as big a cause as he could summon, the variable eccentricity of the earth’s orbit. What he did next was his real leap. Rather than accepting Herschel’s statement that changes in the earth’s climate could not be explained by variations in its orbit, he suggested that what he called “secondary causes” operating not on the scale of planetary orbits, but back on the earth itself, were more than capable of accounting for the ice ages.
Water was the medium by which heat traveled around the globe, and the secondary causes to which Croll now turned were a function of the complex interplay between water in all its forms. These secondary causes were primarily the result of the physical properties of water. Even if the total sunlight in a year is held constant, cooler winters would mean more snow. As snow accumulated from year to year, its reflective qualities would come into play. Because snow is white, it reflects most of the light and heat from sunlight back to space, further cooling the earth. Here was a positive feedback mechanism (though Croll did not use the term) that could begin to account for the steady increase of snow-cover needed to generate an ice age. These cooler conditions begat further cold by increasing the tendency of cooling fogs to form over the snowpack, further insulating it from the heat of the sun. As the temperature gradient between the cold poles and the warm tropics increased, the trade winds would blow more toward the equator, deflecting the Gulf Stream to the north and its sister current, the South Equatorial Current, to the south, further increasing the heat imbalance. The net result was a cooling planet that drifted into an ice age. And so it went, according to Croll’s prodigious imagining, until the gravitational forces shifted and Earth’s orbit assumed a rounder, less elliptical orbit, allowing the ocean currents to shift, snow to melt in the summer, and the feedback mechanisms to work in reverse, accelerating melting and warming where they had once accelerated cooling.
Croll’s theory of climate change was based on global factors that were not geological in origin but physical. Croll imagined the movement of heat through the atmosphere and the oceans and the ice sheets as a means of explaining dramatic and sustained changes in climate and therefore had no need for the long, slow, and monumental rising and falling of continental-scale landmasses on which Lyell had staked his reputation. “The cause of secular changes of climate,” he wrote, “is the deflection of ocean currents, owing to the physical consequences of a high degree of eccentricity in the earth’s orbit.”16 Croll was unapologetically aligned with the physicists. He was not concerned by the lack of geological evidence for his theory. In fact, he claimed that the very absence of geological data was a form of evidence in favor of his theory. It was the nature of the ice ages that the erosive action of glaciers destroyed the evidence of their own (successive) passages. Croll was, like Tyndall, comfortable in drawing inferences based on fundamental physics. He was confident enough in his physics, and the logic of his inferential apparatus, to take his assumptions to their logical extremes. Scale was no object. If the logic led him to view the ice ages on this planet as a function of astronomical shifts married to global physical dynamics, then so be it.
In spite of his lack of standing in the scientific community, Croll had come up with a theory too compelling to be ignored. It piqued the interest—and raised the hackles—of some of the most prominent thinkers of the day. In the process of revising the tenth edition of his great work, Principles of Geology, Lyell wrote to ask his friend the great astronomer John Herschel what he thought of Croll’s theory. Convinced of his own ideas about the gradual (and uniformly acting) changes that affected the earth’s climate, Lyell couldn’t completely disregard what he acknowledged was compelling evidence to the contrary. “I feel more than ever convinced that changes in the position of land and sea have been the principal causes of past variations in climate, but astronomical causes must of course have had their influence and the question is to what extent have they operated?”17 That indeed was the question. To Lyell, climate change occurred primarily as a result of geographical change—the rising or falling of landmasses, the subsequent changes of sea level, and the blocking or opening up of ocean currents.18 He considered neither astronomical nor the secondary physical mechanisms of warming ocean currents and coolly reflective ice to be sufficient to account for the changes in climate the earth had witnessed. Herschel’s reply was not reassuring. Croll’s astronomical causes were, according to the astronomer, “quite enough to account for any amount of glacier and coal fields.” Given the right astronomical conditions, Herschel continued in grudging acceptance of Croll’s theory, “any amount of glacier you want is at your disposal.”19
As well as the reluctant acceptance of geologists such as Lyell, Croll found other allies who were more enthusiastic. Theory could offer a powerful way out of the conundrums caused by too much, and too complex, data. Reading the geological evidence that took the form of messy, complicated drift deposits had challenged scientists for decades. Drift was chaotic, disordered, and largely devoid of fossils. All the tools that geologists had so far developed to enable them to make sense of the structure of the earth depended on the presence of fossils (to allow comparative dating) and the assumption that sediments were gradually deposited over time and could thus serve as standardized indexes of past change. Drift obeyed none of these rules and offered none of these tools. Charles Darwin was forever embarrassed by his failure to see the evidence for glacial motion in the escarpments of Wales that he visited in 1831. Once he’d learned to read the landscape, it seemed clear to him that such features could only have been caused by massive ice sheets. Darwin was an enthusiastic reader of Croll’s work, writing to tell him that “I have never, I think, in my life, been so deeply interested by any geological discussion. I now first begin to see what a million means, and I feel quite ashamed of myself at the silly way in which I have spoken of a million years. . . . How often I have speculated in vain on the origin of the valleys in the chalk platform round this place, but now all is clear.”20
Less easy to see, even for those who were trained to look, was evidence of both advancing and retreating glaciers. In 1871, James Geikie (who had worked with Croll at the Geological Survey of Scotland since 1867) published a book outlining his own theory of the ice age—or, rather, ice ages, because Geikie’s signature claim was that the ice age was actually a series of glacial periods punctuated by warmer interglacial periods. Thanks to Croll, Geikie was emboldened to look for the causes of terrestrial climate in the solar system. Dissatisfied with the idea that the rise and fall of great landmasses could explain such dramatic changes in climate as the fossil and plant evidence from the past suggested, Geikie wondered, “is it not possible that a solution of the problem may be found in the relations of our planet to the sun?”21
Geikie was a geologist by training and inclination, but he would not
have arrived at his theory of alternating warm and cold periods on the basis of the complex and fragmentary geological record alone. Without Croll’s big idea, Geikie would have had neither the confidence nor the insight to make his claim that there was not simply an ice age but, as one historian has called it, an “eventful” series of ice ages. When he published a series of seven papers on the topic, Geikie was careful to lead with the geological evidence for his ideas—glacial deposits found in Scandinavia, Switzerland, and North America—only referring to Croll’s theory of climate in the latter papers. In doing so, he deliberately made it seem as if he was proceeding inductively, as geologists did, by building up a theory out of the many bits of evidence about glacial deposits. That way was safer, and more convincing to the geological community, than proceeding assuming the theory was correct and using it to make sense of the complex records of glacial deposits—a deductive form of logic.22
The need to bring other kinds of thinking to bear on the problem of the ice ages made many people uncomfortable. It was difficult, when different methods of science were brought together, to know what counted as evidence, or facts, anymore. How valuable were theories that couldn’t be tested? Sometimes, as with the case of Croll, it could be very useful indeed to grab hold of an idea, such as alternating glacial and interglacial periods, that could be used to sort a planet’s worth of evidence. There would always be anomalies. The trick was to determine when, if ever, the anomalous data was heavy enough to bring the whole theoretical structure tumbling down. In the meantime, if the structure held and enough data fit, it seemed reasonable to ignore the few pieces that didn’t.
