Abstract and Keywords
Meteorology in Victorian Britain was a science of expectations, not only in the literal sense in that it dealt with statements about future weather, but also figuratively, in that it summarized what Victorians thought science could or should be. The status of weather prediction as reliable knowledge grew from the difficulty of understanding atmospheric change. Throughout the nineteenth century, meteorologists struggled not only with the process of data collection but with the relationship of particular observations to the general laws of atmospheric change. In philosophical terms, this was the problem of induction: how could one build a general understanding from the accumulation of facts? Coupled to the philosophical and practical difficulty with observations in meteorology was the definition of probable knowledge. The development of weather observation networks coincided with the introduction of a new approach to probabilities: the statistical management of large amounts of data to reveal underlying patterns.
METEOROLOGY in Victorian Britain was a science of expectations, not only in the literal sense that it dealt with statements about future weather, but also figuratively, in that it summarized what Victorians thought science could or should be. The question of weather prediction became a means for contemporaries to identify critical ways in which the authority of modern science shaped the prevailing culture. Through their experiences with the science of the weather, Victorians explored ideas about expertise, the growth of public institutions, and the effects of a burgeoning print culture within which intellectual leadership took shape. None of these developments was uniquely confined to meteorology, or indeed to science, yet weather forecasting was particularly effective in outlining the claims made for natural knowledge and its leaders in modern society. Meteorology revealed these claims so well precisely because it was not conspicuously successful—because the difficulties of authoritative weather prediction exposed the uncertainties involved in defining proper science for a wide range of people and interests in the mid-nineteenth century. The weather prophet Patrick Murphy's lucky hit in his almanac of 1839, Robert Fitzroy's suicide in 1865, or the analysis of sunspot cycles to predict famine in India in the 1870s all showed meteorology's distinctive qualities. In each of these developments, Victorians negotiated the character and effects of knowledge, discipline, participation, and public life as features of modern society.
The status of weather prediction as reliable knowledge grew from the difficulty of understanding atmospheric change. It was challenging to collect enough observations over wide areas of land and sea in forms that could be coordinated and compared. But on a more fundamental level, the work of (p.286) coordinating many individual observations highlighted the effort of interpretation. Throughout the nineteenth century, meteorologists struggled not only with the process of data collection but with the relationship of particular observations to the general laws of atmospheric change. In philosophical terms, this was the problem of induction: how could one build a general understanding from the accumulation of facts? The observing programs of the 1840s through the 1860s were inspired by the assumption that more data was necessary, and the hope that such data was perhaps also sufficient, for the transformation of meteorology into a mature science. And yet by the 1870s, years of extensive and mechanized observations had not redeemed this promise. The pressure of these circumstances on meteorologists was a recurring theme. Reading the Meteorological Office publication Weather Charts and Storm Warnings in 1876, Francis Galton underlined in frustration the statement that atmospheric change was “mainly regulated by the distribution of barometrical pressure over the globe,” jotting a sardonic note about cause and effect in the margin: “The author seems to regard the pressure as a sort of wild beast having volition of its own.”1 It was all too easy to transplant instruments and observations to the center of inquiry and veer away from more fundamental questions.
Coupled to this philosophical and practical difficulty with observations in meteorology was the definition of probable knowledge. The development of weather observation networks coincided with the introduction of a new approach to probabilities: the statistical management of large amounts of data to reveal underlying patterns. As much as meteorology lent itself to the new program of statistical study of natural phenomena, however, the challenge of actually predicting the weather was much more difficult. Calculating the chances that one could expect that a particular day would bring sunshine or rain made sense within an older philosophy of probabilities. It aligned meteorological science with ideas about reason and the role of individual judgment that many Victorians continued to find sympathetic. But it was flatly out of touch with statistical thinking, the reinterpretation of probabilities as a form of precise knowledge based on analysis of a cluster of observations. At the same time, statistical techniques to manage the multiple variables that characterized a question of weather prediction had not yet been developed. Changing ideas in this period about what probability meant and what problems it could manage made it particularly difficult to speak of weather prediction as a science.
(p.287) Concerns with the nature of observations or probabilities cannot be separated from the history of institutions, technologies, or political culture in which they developed. Meteorological science thus gives the historian a way of analyzing the intellectual world of the Victorians and showing how scientific ideas and practices blended into broader debates. The most obvious of these contexts in meteorology was the call for discipline and organization. Observation networks were enterprises built around the ideals of voluntary participation, the control and subordination of individual parts, and the public value of natural knowledge. These elements, according to Victorian men of science, would transform fragmented individual accounts of natural phenomena into a coherent and productive science. Meteorology thus promised to demonstrate the value of collective effort, coordinated under the expert guidance of a centralized authority. That description fits the project of the marine observations that prompted the foundation of the Meteorological Department in the Board of Trade, and it fits too the private or semi-public networks like George James Symons's rainfall observers or James Glaisher's observers producing statistics for the General Register Office. A similar picture of science and community inspired Dr. Merryweather's instrument at the Great Exhibition in 1851. With his leeches arranged in a circle of counselors, linked by telegraphy to the nation as a whole, Merryweather envisioned knowledge as a vast spatial and social exchange.
The technology that had inspired Merryweather, electric telegraphy, made those visions of community and control more intense. Telegraphy spectacularly expanded the scale and speed of scientific exchange. But even more significantly, by offering the possibility of weather forecasting, it concentrated attention on the question of the immediate benefits of science, especially to the shipping industry. That in turn meant meteorology became part of longstanding discussions about how to characterize the relationship of natural knowledge to economic prosperity. The authority of Victorian men of science hinged on a careful definition of this relationship, close enough that there could be no doubt about the importance of scientific research, but not so close that it tainted the scientific gentleman with the aura of trade and business. There was no monolithic position on this relationship, within the scientific community or without, and the range of opinion on this important issue only increased the significance of meteorology as a test of how to evaluate scientific work. This was the background to the shift of meteorological science to the natural laboratory of India. Invested with these strong associations of discipline and public responsibility, the science of meteorology logically combined (p.288) with questions about the management of the empire. Famine prediction in imperial India therefore can be seen as the natural heir to the standard Victorian accounts of meteorological science as a collective enterprise, renewed by a technological conquest of distance and disciplined leadership from the center.
The emphasis on collective endeavor in meteorology, however, contained a dilemma. Wide participation in the science was critical to its development as a coherent body of knowledge, but it also disrupted that coherence. The distinctive features of meteorological science were the almanacs and astrology, on the one hand, and the installation of meteorology in a government office on the other. Because of the almanacs, meteorologists struggled to define their relationship to popular knowledge and the illegal world of astrological prophets. The government office set the stage for a clash between a conception of science as practical knowledge and a conception of publicly funded research. These circumstances increased the range of interests in meteorological science and created particularly controversial juxtapositions of individuals and claims. Eminent Victorian philosophers often carried debates about meteorology to places that seem inappropriate to the modern picture of the professional scientist: in newspapers, almanacs, or lectures on art as well as scientific journals and monographs or official commissions of inquiry. Although certainly subtle distinctions about what sort of publication to use or which scientific meetings to attend were significant, all involved shared the conviction that public exchange was the way to evaluate science. This conviction connected figures as diverse as John Herschel, Zadkiel, and John Ruskin. Meteorology was typical of Victorian scientific culture, a fluid and heterogeneous world in which messy disputes arose precisely because of commitment to reason and natural knowledge as common property, in principle accessible to all.
As revealing as is this diversity of places and people, it represents differences of degree rather than kind for a history of Victorian science. Studying meteorology does offer a particular advantage, however, for an analysis of the relations between popular and elite knowledge. This advantage emerges not because the science brought such a diverse range of individuals and visions of science into contact. Nor is it simply a question of the presence of strong traditions of popular knowledge. These qualities, as just noted, could be considered typical rather than extraordinary. Rather, meteorology was a science in which the nature of popular knowledge and its relationship to the world of observatories and precision instruments was explicitly a subject of discussion. This takes us back to the importance of failure in the science. Because weather prediction was often inaccurate, the apparent accuracy of weather wisdom (p.289) was an important consideration. It was critical for some meteorologists to establish the distinction between weather wisdom and their own approaches; others sought to incorporate the features of weather wisdom that seemed to be responsible for its authority. Investigating these relationships between organized meteorology and weather wisdom provides a richer picture of both meteorology and Victorian science in general. For instance, it suggests a new way to think about instruments and maps in the science. Rather than tools to construct modern forms of objectivity through quantitative and mechanical records, some instruments can be seen as far more tentative objects, used by meteorologists as a substitute for natural precision. They could represent connections between popular and elite knowledge, instead of a stark boundary. Similarly, weather maps had an appeal based partly on their representation of global and quantitative data, and partly on their endorsement of a weather-wise visual sensibility, meaning at a glance.
A history of weather prediction then leads directly to a picture of the complexities of Victorian intellectual life and the values it upheld. But that picture suggests a need to think further about the development of global sciences. Is there something paradoxical about studying a history of meteorology in Britain? Why insist on the importance of national framework for understanding a science like meteorology in this period, when contemporaries so clearly recognized that its dynamics were global and required new commitment to widespread, coordinated observation? One answer is that the organization of the science through government institutions, and the importance of public expectations for the science, reinforced the effect of specific national contexts on the science. In Britain between the 1850s and the 1880s, there was certainly increasing exchange of observations with other countries, but little real pressure to approach the study of the weather more internationally. The emphasis was rather on developing a central institution to control meteorological work on a national scale, an endeavor that meshed with the ambitions of the scientific metropolitan elite. Nearly two decades passed between the Brussels conference that proposed to coordinate marine observations, for example, and the next international meeting, in Leipzig in 1872, to set common goals and standards in meteorology. Regular international congresses followed, but coordinating meteorological work among different national observatories and standards continued to present significant obstacles.
On the particular question of weather prediction, for example, the difference of opinion within European nations precluded anything but an agreement to differ. In 1878, nearly two decades after the introduction of government forecasting, an international meeting in Vienna included a report that (p.290) presented the results of a questionnaire about storms warnings and forecasts distributed to observatories and prominent individual meteorologists. There was no agreement about what information to distribute—the observations themselves, opinions, or predictions—nor what kind of forecasting was possible (in the case of storms, for example, whether only to signal strength of wind, or direction as well). To many, the system in place in the United States was a model; others found it too ambitious. Although there was general favor for storm warnings along the lines introduced by Fitzroy in nations with strong maritime interests, like Portugal, there was more skepticism in landlocked capitals like Paris, Berlin, or St. Petersburg.2 Philosophically, the science of meteorology was global; in practice, global science developed in distinctively different political and geographical landscapes, and contemporaries insisted on the importance of the differences.
The question of global versus national contexts for the history of nineteenth-century meteorology could also take a reflexive turn. It can be argued that the idea of global science was itself a particular challenge on both a philosophical and practical level. Meteorologists were particularly aware that their science was based on connecting local experiences and global phenomena. At its most vivid level, these shifts in perspectives were the basic premise of nineteenth-century mapping techniques, the isobars and isotherms, linking up places where barometers and thermometers gave the same numbers. The arguments about sunspots hinged on similar questions about general knowledge: could the solar variation signaled by changes in sunspots explain weather, or did the particularity of terrestrial conditions make such literally over-arching explanation impossible? Such issues were related to critical narratives about the progress of meteorology, in which the relationship of particular fact to general law corresponded to the transition from a local knowledge of airs and places into a universal physical science. Victorian men of science thus wrestled hard with the problems of integrating local and general perspectives. Any irresolution in international cooperation reflected their experience with those questions as well as with the more obvious problems of agreeing on standard registration forms and instruments. For the historian, this suggests that the meaning of global science needs to be investigated through exactly these shifting contemporary characterizations of meteorology as global, national, or local science.
But there is another kind of irresolution in the history of Victorian weather prediction. How does the story end? One could point in conclusion to a particular (p.291) historical momentum that governs the narrative. The particular circumstances of the mid-Victorian era—including the economic rationale for storm forecasting and the pressure for public funding of scientific institutions—gave way in the 1880s as steam power replaced sail, and science became a more secure profession. With the passage of time, it seemed simultaneously less important to resolve the philosophical status of forecasting and less likely that a decisive and concerted effort on the part of meteorologists would produce a definitive dynamics of the atmosphere. Meteorology and weather forecasting, however it was evaluated, simply became less prominent. But fading controversy did not represent resolution. Another narrative of meteorology would represent cycles rather than a dying momentum. The Meteorological Office ceased to publish warnings and forecasts after Fitzroy's death and the inquiry of 1866, but warnings resumed at the end of 1867. By 1879, Meteorological Office forecasts for eleven different districts of the British Isles began to appear regularly, provided not only to newspapers but also available as a direct service to subscribers and for individual inquiries. Three years later, a failure to predict an October storm sparked a correspondence in Nature about observations, instruments, and public funding that could have been lifted word for word from the time of Fitzroy's experimentation with storm warnings twenty years earlier.3 The uneasiness about the scientific status of prediction remained, too. In 1926, more than seventy years after the founding of the Meteorological Department under Fitzroy, his successor Napier Shaw commented that “it is hardly an exaggeration to say that meteorologists have a natural aversion from the iteration of the duty of forecasting.”4 Many of the features of the Victorian experience with weather prediction, in other words, remain remarkably persistent.
Victorian meteorology thus continued to shape the science long after the energy of the controversies of the 1860s had dissipated. The resilience of Victorian patterns of thinking of leadership, visual knowledge, discipline, and collective knowledge emerges in Lewis Fry Richardson's renowned 1922 book Weather Prediction by Numerical Process. Richardson's calculations of air pressure, density, and humidity in this book were a mathematical performance of the hydrodynamical theories of the atmosphere developed by Vilhelm Bjerknes and others in Bergen, Norway, in the years immediately after World War I.
(p.292) Directly contrary to his hopes, Richardson's work convinced most readers that such numerical analysis of the weather was completely impractical: his six-hour forecast for a single location had taken him six weeks to calculate. Serious application of numerical methods did not emerge until after another world war, when John von Neumann at Princeton chose weather prediction as a project to demonstrate the potential of the electric computer, and thereby turned meteorology, as Richardson had foreseen, into a science of calculations.5
Richardson's book is a recognized landmark in the history of meteorology and its transformation in the early decades of twentieth century. Less evidently but equally importantly, the book reveals how the framework for the development of modern meteorology was created in the nineteenth century. In anticipation of the critics who found his methods impossibly laborious and slow, alongside his own calculations Richardson described a model for a large calculating center. Working on a scale that any one individual could not match, this imagined system would make weather prediction an attainable science. In Richardson's scenario, a large number of human computers raced through differential equations for several layers of blocks of the atmosphere, altogether representing two hundred square kilometers. The remarkable details of his center summarize the concerns we have just finished tracing in this history of Victorian meteorology. Here then is another ending, one which looks both forward and backward.
“Imagine,” Richardson wrote, “a large hall like a theatre” whose walls, ceiling and floor are all “painted to form a map of the globe.” Each computer works upon equations for the regions corresponding to his or her place in the theater, and their work is fed to overhead displays so that adjacent computers can incorporate their results instantly. In the center, “a tall pillar” supports a “large pulpit” in which stands the central director and his assistants. From his pulpit, the director “maintain[s] a uniform speed of progress in all parts of the globe … like the conductor of an orchestra in which the instruments are slides and calculating machines.” Instead of a baton, however, this scientific conductor has “a rosy beam of light” to wave on the walls representing a region whose computers are moving too quickly, and a “beam of blue light” for those regions whose computers need to improve their speed. Pneumatic tubes transfer the calculations to another room, from which a coded version is sent by telephone to a radio station. In an adjacent building, researchers toy (p.293) with small hydrodynamic models, in the unlikely event that such laboratory experimentation can be channeled to help the work of the computers.6
Richardson's fantasy suggests the enduring effect of the features that were explored and established by Victorian meteorology. The need to work on a global scale, to coordinate and control individual activities, to create a hierarchical system that could decipher knowledge and then transmit it to waiting audiences—these were the challenges of modern weather prediction. The leader who is simultaneously a director, a preacher in his pulpit, and a conductor creating harmony from a miniature world of instruments and calculators was an evocative image. As an account of the scientist, it responded to interpretations of collective knowledge and scientific leadership that had developed in the mid-nineteenth century. The tension familiar to Victorians between researchers, working on their models, and the practical computers, working with their data, is there as well, as is the integration of the whole system with modern communication technologies. Finally, the visual and panoramic sensibilities associated with the study of the weather appear dramatically in the mapped walls, the beams of light, and the commanding panoptic pillar. In 1922, Richardson intended his calculating theater as a prophecy, marking out the future of meteorology as a science of computers. But it was also testimony to the past, and to the influence of Victorian experiences with meteorology. (p.294)
(1.) See the copy of Scott's Weather Charts and Storm Warnings, 121, annotated in Galton's hand, Galton Papers.
(2.) Report on Weather Telegraphy and Storm Warnings.
(3.) Carlisle, “Weather Forecasts,” Nature, November 2, 1882, 4; Ley, “Weather Forecasts,” Nature, November 9, 1882, 29; Carlisle, “Weather Forecasts,” Nature, November 16, 1882, 51–52; Carlisle, “Weather Forecasts,” Nature, November 26, 1882, 79.
(4.) Shaw, Manual of Meteorology, 1:9.
(5.) See Fleming, ed., Historical Essays on Meteorology; Friedman, Appropriating the Weather, Nebecker, Calculating the Weather.
(6.) Ashford, Prophet or Professor? 91–92. This famous fantasy of numerical methods deserves comparison with other, fictional fantasies of the control of the weather in the early twentieth century, like Jones, The Great Weather Syndicate. For a related discussion of the conductor and scientific knowledge, see Winter, Mesmerized, 309–20.