5
Late Enlightenment Meteorology

By Theodore S. Feldman

Meteorology in the 18th century meant more than observation and prediction of the weather. The complex and manifold processes occurring in the atmosphere brought it into contact with a large range of topics of natural philosophy—topics differently arranged than they are today, their boundaries less distinct and far more fluid. Meteorology naturally overlapped much of pneumatics, or the study of gases. The pneumatical parts of meteorology included the expansion of air with pressure and heat, evaporation and precipitation, latent heat, the behavior of aqueous vapor in air, and the closely related problems of the production of steam in both air and the vacuum. In the last decades of the century the discoveries of the chemical components of air and water led meteorologists to pneumatic chemistry and the chemical constitution of the atmosphere.

Meteorology also shared topics with electricity and magnetism. Atmospheric electricity, including lightning and the electricity of the air and clouds in fair and stormy weather, was a popular topic and the electrometer a familiar instrument of weather observation. Terrestrial magnetism belonged to meteorology at this time, and, especially in the last third of the 18th century, meteorologists improved the magnetic compass and discovered regularities in the magnetic variation.

These topics, coming from pneumatics, heat, electricity, and magnetism, belonged to natural philosophy or physics proper; a second grouping of subjects—including mechanics, geometrical optics, acoustics, and surveying and cartography—made up applied or "mixed" mathematics, as it was called.^[1] Meteorology had much in common with these as well. Barometry had its 17th-century beginnings in hydrostatics, which was part of mechanics. In the late 18th century, when the barometer became useful for the measurement of heights, it was taken up by surveyors. Atmospheric refraction, rainbows, haloes, the color of the sky, and the diminution of light through the atmosphere brought meteorology into contact with geometrical optics and the new science of photometry. Mixed mathematics gave these parts of meteorology a mathematical content throughout the 18th century.

Finally, meteorology was involved with parts of the life sciences, especially agriculture, medicine, and plant and human geography. The influence of the weather on agriculture and human health involved meteorologists in the numerous projects of agricultural and medical reform of the last third of the century. Botanists such as Linnæus and Willdenow began to consider the role of climate in the geographical distribution of plants. Montesquieu's discussion of the role of climate in human geography is well known.^[2]

[1] Applied mathematics was distinct from natural philosophy for most of the 18th century and was quantitative, where most of natural philosophy was descriptive and nonquantitative. As has been shown in several studies, quantification in physics in part reflected an alteration in the relation between these two fields. See J.L. Heilbron, Electricity in the 17th and 18th centuries. A study of early modern physics (Berkeley: University of California Press, 1979), 1–19. Theodore S. Feldman, "Applied mathematics and the quantification of experimental physics: The case of barometric hypsometry," Historical studies in the physical sciences, 15 (1985), 127–97, 128, lists further studies. The present chapter will not be concerned with this aspect of quantification.

[2] For pneumatical meteorology, see Feldman, The history of meteorology, 1750–1800. A study in the quantification of experimental physics (Ph.D. dissertation, University of California, Berkeley, 1983; DAI 45/03-A, 922), section I, and W. Knowles Middleton, "Chemistry and meteorology, 1700–1825," Annals of science, 20 (1964), 125–41. For atmospheric electricity, see Heilbron, Electricity. For earth magnetism, see D.H. Hall, History of the earth sciences during the scientific and industrial revolutions (Amsterdam, New York: Elsevier Scientific Pub. Co., 1976). For barometry, see Feldman, "Applied mathematics." For plant geography, see Adolf Engler, Die Entwicklung der Pflanzengeographie in den letzten hundert Jahren und weitere Aufgaben derselben (Berlin: W.H. Kuhl, 1899). For Montesquieu and other studies of the role of climate in human geography, see Clarence J. Glacken, Traces on the Rhodian shore. Nature and culture in western thought from ancient times to the end of the eighteenth century (Berkeley: University of California Press, 1967), chap. 12.

Meteorology not only overlapped these areas of natural philosophy; it enriched them. This was especially true in physics. Both the law of uniform expansion of air with heat and Dalton's law of partial pressures grew out of or were closely connected to meteorological investigations.^[3] Capital discoveries in electricity in the mid-18th century were the electrical nature of lightning and the phenomenon of electrostatic induction, which derived in part from the study of the electricity of clouds.^[4] A prize question on the magnetic compass led to Coulomb's discovery of the law of magnetic attraction.^[5] More important than these, meteorologists made substantial contributions to the design in the 1770s and 1780s of the first precise instruments of physics. These instruments and an emphasis on systematic measurement—which together may be called "exact experimental physics"—were crucial aspects of quantification in late 18th-century natural philosophy. They generated reliable quantitative data that could be used as a foundation for mathematical laws.^[6]

This chapter treats quantification in one part of meteorology: weather observation and climatology. The term "climatology" is something of an anachronism; it and its cognates are not to be found in the 18th century but made their appearance in the first years of the 19th.^[7] For convenience "climatology" is used—sparingly—to

[3] Students of barometric hypsometry—the barometric measurement of heights—investigated the expansion of air with heat in the 1770s. See Feldman, "Applied mathematics," 175–6. The law of partial pressures for mixtures of air and water vapor was discovered in the context of hygrometry in the 1770s and 1780s. Cf. Feldman, The history of meteorology , chap. 4.

[4] For discussion of induction, see Heilbron, Electricity , 374.

[5] Heilbron, Electricity , 469.

[6] For discussions of exact experimental physics, see Feldman, "Applied mathematics," 151–2, 164–77, and "The history of meteorology"; Heilbron, Electricity , 75–83; Maurice Daumas, "Precision measurement and physical and chemical research in the 18th century," in Alistair C. Crombie, ed., Scientific change (New York: Basic Books, 1963), 418-30.

[7] For example, W.A. Lampadius, Systematischer Grundriss der Atmosphärologie (Freiburg: Craz und Gerlach, 1806), chap. 4, "Climatologie"; and Leopold von Buch, "Klimatologische Betrachtungen über die Westküste Norwegens," Annalen der Physik, 25 (1807), 318–32.

denote the analysis of observations for weather patterns. In climatology the organizational imperative—the need for coordinated observations carried out in different locations according to a common plan—meant that strong meteorological institutions were a prerequisite for quantification. So was exact experimental physics. The two appeared almost simultaneously after 1770. In order to clarify their roles, this chapter begins with an overview of weather observation and climatology prior to 1770, then turns to quantification in the two decades 1770–90.

Weather Observation and Climatology Prior to 1770

From the invention of meteorological instruments in the middle of the 17th century, natural philosophers recognized that little could be won from individual weather diaries or registers; instead, organized groups of observers were needed. As G.A. Hamberger, Christian Wolff's mentor at the University of Jena, put it:^[8]

It is not enough to examine the state of the air in our own location, but we must direct our attention to the surrounding regions. This may best be done if well-informed persons throughout several provinces and neighboring kingdoms record [weather observations] simultaneously. . . . If many ephemerides of this type, from various locations, are published and compared, they will throw great light on [weather] phenomena.

Short-lived organizations or networks of weather observers had been set up around the middle of the 17th century by Périer, Pascal's brother-in-law; by Robert Hooke at the Royal Society of London; and by Ferdinand II of Tuscany, patron of the Accademia del Cimento in Florence.^[9] Little came of their efforts and interest in

[8] G.A. Hamberger, De barometris (1701), quoted in Gustav Hellman, Repertorium der deutschen Meteorologie (Leipzig: W. Engelmann, 1883), 884. As the title of the work suggests, Hamberger was discussing barometric observations in particular.

[9] Hellman, "Die Entwicklung der meteorologischen Beobachtungen bis zum Ende des 18. Jahrhunderts," Preussische Akademie der Wissenschaften, Physischmathematische Klasse, Abhandlungen , 1927, 9; W. Knowles Middleton, A history of the thermometer and its use in meteorology (Baltimore: Johns Hopkins Press, 1966), 30–2; H. Howard Frisinger, The history of meteorology: To 1800 (New York: Science History Publications, 1977), 101–2.

coordinated weather observation declined from about 1660 to the end of the century.

In the early years of the 18th century as few as three or four natural philosophers in all of Europe culled weather observations from newspapers and magazines and from reports of correspondents. After 1715 more energy is evident: at Breslau the physician Johann Kanold published in his quarterly Breslauer Sammlung observations he compiled from a dozen locations across Europe.^[10] In 1723 the Philosophical transactions of the Royal Society of London carried James Jurin's "Invitatio ad observationes meteorologicas."^[11] In it Jurin, the Society's secretary, laid out a plan for daily readings of barometer, thermometer, wind strength and direction, precipitation, and the state of the sky. Some fifteen observers, from Bengal and St. Petersburg to Cambridge, Massachusetts, responded with weather diaries, which William Derham, a Fellow of the Society and author of works on physicotheology, edited for the Philosophical transactions. This labor made Derham the most prolific meteorologist of the first third of the century.

Although Kanold and Jurin had ambitious hopes for their networks, the results fell short of expectations. Observers had difficulties in obtaining instruments. For example, Derham located few observations of the hard winter of 1709–10 made with instruments.^[12] There seem to have been no instruments in the American colonies prior to 1716, when the physician Cadwallader Colden began observing in New York; as late as 1727 neither barometer nor thermometer was to be found in the Boston area, where Isaac Greenwood, first Hollis Professor of Mathematics and Natural Philosophy at Harvard, served as Jurin's delegate on the weather watch.^[13] A number of Jurin's

[10] Sammlung von Nature- und Medicin- wie auch hierzu gehörigen Kunst- wie Literatur-Geschichten , Breslau, 1717–30.

[11] James Jurin, "Invitatio ad observationes meteorologicas communi consilio instituendas," PT, 32:2 (1722–3), 422–7.

[12] William Derham, "History of the great frost in the last winter 1708 and 1708–9," PT, 26 (1708–9), 454–78.

[13] Derham, "An abstract of the meteorological diaries communicated to the Royal Society," PT, 37 (1731–2), 261–79, on 267; J.H. Cassedy, "Meteorology and medicine in colonial America: Beginnings of an experimental approach," Journal of the history of medicine, 24 (1969), 193–204, on 197–9; Brooke Hindle, The pursuit of science in revolutionary America, 1735–1789 (Chapel Hill: University of North Carolina Press, 1956), 88.

volunteers observed without instruments, as did most of Kanold's—instruments seem to have come late and few to central and eastern Europe, where about half of Kanold's observers were stationed.^[14] A late 18th-century meteorologist at Prague reported that meteorological instruments first appeared in Bohemia in 1750 and were still rare two decades later.^[15]

Even when instruments were available, their measurements were nearly useless unless the instruments were comparable—that is, unless instrument scales were interconvertible. Comparability of barometric observations posed no difficulty in principle, since the length of a mercury column serves as a natural scale for atmospheric pressure—though the variety of national and even regional units of length confused matters considerably. But the variety of scales for the thermometer greatly diminished the usefulness of temperature readings. Although Jurin attempted to secure comparable observations by asking his volunteers to specify the scale and make of their thermometers, many failed to do so. Derham found reducing their data "a matter so perplexed and difficult, as not to answer the great trouble of it."^[16]

Lack of precision and reliability posed further problems. Instruments neither rendered accurate readings, nor could they be depended upon to render the same reading twice in identical situations. Precision and reliability, in fact, did not trouble early 18th-century natural philosophers.^[17] Imperfect instruments generated much confusion and waste—as late as 1750 measurements made with

[14] Jurin distributed thermometers to his observers (as had Hooke and Ferdinand II), but they were frequently lost or broken during transit. Cf. Louise Diehl Patterson, "Thermometers of the Royal Society, 1663–1768," American journal of physics, 19 (1951), 523–35; "The Royal Society's standard thermometer, 1663–1709," Isis, 44 (1953), 51–63; also J.L. Heilbron, Physics at the Royal Society during Newton's presidency (Los Angeles: William Andrews Clark Memorial Library, 1983), 105–6.

[15] Anton Strnad, "Meteorologische Beobachtungen auf das jahr 1775," Privatgesellschaft zur Aufnahme der Mathematik, der vaterlandischen Geschichte, und der Naturgeschichte, Prague, Abhandlungen, 2 (1775), 392–9.

[16] Derham, "An abstract of the meteorological diaries communicated to the Royal Society," PT, 38 (1733–4), 101–9, 334–44, and 458–70, on 464.

[17] However, see Heilbron, Electricity, 81.

inaccurate barometers led eminent natural philosophers to question Boyle's law.^[18] As for weather observation, "how many observations have we lost," lamented one late 18th-century meteorologist, "through the imperfection and uncertainty of Mr. Hauksbee's thermometer!"^[19] This was just the thermometer Jurin used and supplied to his network.

Observers were scarcely more reliable than their instruments. The discipline of recording daily temperature, pressure, humidity, winds, and cloud cover over a period of years did not come easily. As Derham put it, "these investigations require not only industry and inclination, but also leisure and means and opportunity, which you seldom find together."^[20] Rather than follow Jurin's instructions, several of his informants simply submitted registers they had completed in earlier years. Under these circumstances, a consistent collection of observations was unlikely. The labor of reducing the registers delayed publication, making matters worse. In 1732 and 1733, when the project had almost ended, Derham published comparisons of the weather of 1707 at Upminster and Coventry, the weather of 1715–22 at Upminster and Cambridge, New England, and the weather of 1724 in Lund and St. Petersburg.^[21] Analyses of the project's last registers, which had been submitted in 1734, did not appear until 1742.

Imprecision and unreliability of instruments, lack of agreement among scales, indiscipline in observation and inconsistency of published collections—all these factors limited the achievements of early 18th-century climatology. A thoroughgoing quantitative treatment was not possible. Instead, meteorologists described the weather and summarized observations by calculating monthly and annual means of temperature and pressure and amounts of rainfall. They aimed, as

[18] Feldman, "Applied mathematics," 137–8.

[19] Jan van Swinden, Dissertation sur la comparaison des thermomètres (Amsterdam, 1778), ix. Van Swinden was one of the chief meteorologists of the latter part of the century.

[20] Quoted in Heilbron, Physics at the Royal Society , 108.

[21] Derham, "An abstract of the meteorological diaries," PT (1731–2) and (1733–4).

Derham put it, to give "a just notion of the state of every month . . . and that which was most observable in it."^[22] Even when good quantitative data were available, they did not exploit it. From Jurin's observer at the Academy of Sciences in St. Petersburg Derham received records of the temperature, pressure, winds, and general state of the weather taken three times daily during 1724 and 1725. He felt, however, that the "observations (although very curious and useful), yet being too long, would be tedious to read at the Society's meetings."^[23] What sort of quantitative treatment could there be when, as was the practice at scientific academies, Derham read his reports at the Society's weekly meetings?

Besides describing the weather, meteorologists also drew comparisons among the locations reporting to them. Derham, for example, compared the mean annual rainfall of half a dozen European towns.^[24] Comparisons aimed particularly at coincidences in weather patterns at different locations, that is, at Meteorologica parallela .^[25] Derham repeatedly pointed out agreements among prevailing winds and storms at the towns he was comparing, as well as parallel barometric motions, which were striking. When temperature observations were not comparable, as between Zurich and Upminster, he could still see that the maxima and minima of the two series—that is, warm and cold spells—coincided. "Yea, oftentimes any remarkable weather (especially if of somewhat long continuance) affecteth one as well as the other place."^[26] During one month the weather at Zurich "constantly preceded ours here [at Upminster] by about five or more days." Pieter van Musschenbroek, who among his other services to natural philosophy sponsored a network of half a dozen Dutch observers, observed a similar parallelism: "when the south-east wind blows, it arises half a day sooner at Middelbourg than at Utrecht."^[27]

[22] Derham, "An abstract of the meteorological diaries" (1733–4), 105.

[23] Derham, ibid., 101–2.

[24] Derham, "A prospect of the weather," PT, 24 (1704–5), 1877–81.

[25] David Algöwer, Meteorologica parallela (Frankfurt and Leipzig, 1714), cited in Hellman, Repertorium , 880–1.

[26] Derham, "Tables of the barometrical altitudes at Zurich in Switzerland in the year 1708," PT, 26 (1708–9), 332–66, on 333.

[27] Musschenbroek, Essai de physique , 2 vols. (Leyden: Chez S. Luchtmans, 1739), 895ff.; Derham, "An abstract of the meteorological diaries," PT (1731–2).

These coincidences resulted, of course, from the fact that single weather systems cover large parts of Europe. But only one or two meteorologists understood this before the 19th century. Derham saw this much: that "the weather in both places was influenced by the same causes, whether the Alpine hills and the cold, or the influx of the moon and other heavenly bodies, or any other cause."^[28]

Meteorologica parallela were one type of correlation sought by early 18th-century meteorologists. The "weather rule" was another. Meteorologists hoped to discern patterns in their data that would allow them to predict the weather. Thus Kanold expected his collection to provide a "historical-theoretical attempt to predict one storm from another." Derham derived a number of rules from the observations of Jurin's network: "a cold summer is commonly a wet one"; "western clouds bring much wind"; "the falling of the quicksilver in dark and cloudy weather betokeneth rain; but the rain is always preceded by fair weather."^[29] These and other "superstitious calendar-prognostications"^[30] had long been common in almanacs and popular tradition. Meteorologists hoped to place them on a scientific footing.

These weather rules, meteorological parallela, and general descriptions and comparisons of the weather reflect an approach to the natural world that has been well characterized by Michel Foucault. In his study of 18th-century medical practice he wrote, "Disease is perceived fundamentally in a space of projection without depth, of coincidence without development." Disease appears in the "space" of the human body as in a space without character, flat; the different locations in this space do not affect the disease, so that for example dyspepsia in the lower abdomen, breathlessness in the chest, and epilepsy in the head represent the same disease.^[31] Just so, meteorologists perceived the space of the earth's surface as characterless. We

[28] Derham, ibid., PT (1733–4), 342.

[29] Derham, ibid., PT (1731–2), 265, 267, and (1733–4), 101–2, 105. John Locke also hoped to extract weather rules from collections of observations. See Locke, "A register of the weather for the year 1692," PT, 24 (1704–5), 1917–37; Kanold, Sammlung von Natur- und Medicin- wie auch hierzu gehörigen Kunst- wie Literatur-Geschichten .

[30] Algöwer, Meteorological parallela , "so abergläubischen Calender-Prognosticis."

[31] Michel Foucault, Birth of the clinic (New York: Vintage Books, 1975), 6, 10.

would see the space between Lund and St. Petersburg as a land (and water) mass of great extent, with a richly varied topography (and depth, embracing currents, varying temperatures, varying proximity to land masses, etc.), affecting the weather across its entire compass. For Derham and his colleagues this space might as well not exist. The weather is either the same or different at Lund and St. Petersburg; Derham could compare the weather at Upminster and Coventry, or at Upminster and Cambridge, New England, as easily. So characterless, or "flat," was his perception of space that Derham could call the Alps "hills" and say that the same causes influenced the weather at Zurich, in their midst, and at Upminster, hundreds of miles away in the plains of his island nation.

Just as space, of itself, did not affect the weather, so "there is no process of evolution in which duration [i.e., time] introduces new events of itself."^[32] The weather rule, which correlates weather events at succeeding times with no sense of the creative role played by those times, exemplifies this approach. "Western clouds bring much wind," but we have no sense of the connecting skein of time, of the intervening process.

All this represents a lack of synthetic vision, which appears also in meteorologists' failure to describe the climates of places. They calculated mean temperatures and pressures, found days of monthly and annual extremes, and counted the number of days of rainfall. They did not, however, synthesize this information into a characterization of climate. To borrow once more from Foucault, their calculations named the "visible" aspects of the weather, but they did not penetrate to the hidden coherence among these aspects, that constitutes climate.^[33]

Nor did they integrate the collection of places they studied into a the notion of a region—so that, a fortiori, they did not discuss regional climates.^[34] (Two groups did discuss regions and their

[32] Foucault, Birth of the clinic , 12.

[33] Foucault, The order of things (New York: Vintage Books, 1973), 132.

[34] Malcolm Nicholson has pointed out that 18th-century botany and plant geography similarly lacked the concept of regionality. See Malcolm Nicholson, "Alexander von Humboldt, Humboldtian science and the origins of the study of vegetation," History of science, 25 (1987), 167–94.

climates: plant and human geographers, mentioned above, and mathematicians who calculated the effects of the sun's heat on different parts of the earth. They belonged to traditions distinct from meteorology; neither group drew on the weather observations considered here.) Climatology, then, existed in neither deed nor word in the early 18th century; what meteorologists gathered was a natural history–a collection of descriptions of the weather here and there, at this time and another. Given the piecemeal character of their data, it would have been difficult for them to proceed otherwise.

Weather Observation and Climatology, 1770–1790

Between the time of Jurin's and Kanold's networks in the 1720s and 1730s and the last third of the century, interest in meteorology fell off.^[35] The wars of the mid-18th century disrupted cooperative efforts and undoubtedly discouraged individual observations as well.

Around 1770 the situation changed abruptly. No meteorological treatise of Europe-wide reputation had appeared for more than a century.^[36] Between 1770 and 1790, however, half a dozen authors in as many countries published treatises of international renown and substantial papers populated the journals;^[37] meteorology was

[35] Hellmann, Repertorium , 978, 986. Hellmann provided data on numbers of individual observers only for the Germans.

[36] See Hellmann, "Entwicklungsgeschichte des meteorologischen Lehrbuches," in Hellmann, Beiträge zur Geschichte der Meteorologie, 6 , Veröffentlichungen des Königlichen Preussischen Meteorologischen Instituts, 296 (Berlin, 1917), 1–133, on 52.

[37] Important treatises of the period include Jean André Deluc, Recherches sur les modifications de l'atmosphère, 2 vols. (Geneva: J.A. De Luc, 1772) and Idées sur la météorologie, 2 vols. (London: T. Spilsbury, 1786); Louis Cotte, Traité de météorologie (Paris: Imprimerie Royale, 1774) and Mémoires sur la météorologie, 2 vols. (Paris: Imprimerie Royale, 1788); Horace Bénédict de Saussure, Essais sur l'hygrométrie (Neuchâtel: Chez S. Fauche père et fils, 1783); van Swinden, Dissertation sur la comparaison des thermomètres ; Giuseppe Toaldo, La meteorologica applicata all'agricultura (Venice: G. Storti, 1775). Hellmann documented a sharp rise around 1770 in numbers of meteorological publications and of observing stations in Germany. The number of publications more than doubled from the 1760s to the 1770s and increased by a further 25 percent in the 1780s. The number of "station-years" of observations doubled twice in the two decades. Hellmann, Repertorium , 978, 986.

"zealously pursued throughout almost the whole of Europe."^[38] Organized meteorology prospered. Throughout France, Great Britain, and the United States, economic, agricultural, and patriotic societies sponsored programs of meteorological observation and research. In France a national network of observers was established in 1778 under the auspices of the Société royale de médicine, while the Societas meteorologica palatina at Mannheim constructed an international network whose work was not surpassed for three-quarters of a century. Contemporaries spoke of meteorology as "a new science."^[39] Ludwig Kämtz, the chief authority on meteorology during the first half of the 19th century, agreed. It was in the last half of the 18th century, he wrote, that "this part of physics first began to be treated scientifically."^[40]

Interest in meteorology derived from a number of sources, of which two particularly affected weather observation and climatology. One was the application of meteorology to agriculture and public health. Prior to the advent of bacteriological theories of disease in the late 19th century, physicians followed the ancient Hippocratic doctrine that climate, topography, and living conditions—in short, the environment—are among the chief causes of disease. The particular content of the doctrine varied. In the original Hippocratic treatise "On airs, waters, and places" the seasons influence the balance of humors in the body by virtue of what was called the "constitution" of the air: the constitution of summer being hot and dry; of autumn, cold and dry; of winter, cold and wet; and of spring, a balanced mixture of all four qualities. These constitutions favor certain groups of diseases; an abnormal season or sudden changes in the weather also cause outbreaks of disease. Winds (airs) blowing from different directions, the orientation of towns (places) facing the winds, and the towns' water supplies (waters) similarly affect disease patterns. These

[38] Jean Sénébier, "Sur les moyens qu'on pourrait employer pour perfectionner la météorologie," Journal de physique, 27 (1785), 300–15, on 301.

[39] "Sur le froid de 1776," Académie royale des sciences, Paris, Histoire , 1776, 1–14, on 9; quoted by David Cassidy, "Meteorology in Mannheim: The Mannheim Meteorological Society, 1780–1795," Sudhoffs Archiv, 69 (1985), 8–25, on 8.

[40] Ludwig Kämtz, Lehrbuch der Meteorologie , 3 vols. (Halle: Gebauer, 1831–6), 1 , viii.

groups or patterns of disease attacking a population were termed the epidemic constitution.^[41]

While abandoning the theory of humors, the 17th and 18th centuries retained the notion that airs, waters, and places influence the epidemic constitution. The same four qualities—hot, cold, wet, and dry—were now held to act mechanically on the body; the air might also contain disease-causing effluvial exhalations from the interior or surface of the earth. Such theories led numerous physicians to keep weather observations in the expectation of correlating weather patterns with diseases. A typical mid18th-century product was Paul Malouin's annual "Histoire des maladies épidémiques, observées à Paris, en même temps que les différentes températures de l'air," which offered qualitative descriptions of each month's weather and epidemic diseases.^[42] Agriculture enjoyed nothing like the corpus of theory relating the weather to disease, but agriculturalists hoped that regular observations would succeed in correlating the weather with the success of crops. The most prominent midcentury effort of this type was Duhamel du Monceau's "Observations botanico-météorologiques," which for forty years presented annual tables of weather observations and general remarks on crops and public health.^[43] Duhamel's and Malouin's series became important models for later efforts.

This genre of medical and agricultural climatology became institutionalized in the last third of the century as European states increasingly intervened in matters of public welfare. Contemporaries referred to interventionist measures as "police"; "public administration" would be a modern synonym. Medical and agricultural police needed information on the environment, diet, hygiene, and living conditions of the populace, on agriculture, and on outbreaks of disease.^[44] Dozens of state and private institutions arose to fill these

[41] Genevieve Miller, "'Airs, waters and places' in history," Journal of the history of medicine, 17 (1962), 129–40; Frederick Sargent II, Hippocratic heritage: A history of ideas about weather and human health (New York: Pergamon Press, 1982).

[42] Académie royale des sciences, Paris, Mémoires , 1746–54.

[43] Ibid., 1741–81.

[44] George Rosen, "Cameralism and the concept of medical police," and "Mercantilism and health policy in eighteenth-century French thought," From medical police to social medicine: Essays on the history of health care (New York: Science History Publications, 1974), 120–41, 211–9, resp.; see André Bourde, Agronomie et agronomes en France au XVIII siècle , 3 vols. (Paris: S.E.V.P.E.N., 1967), 3 , 1533ff., for a discussion of agricultural police in late 18th-century France; L.J. Jordanova, "Earth science and environmental medicine: The synthesis of the late Enlightenment," in L.J. Jordanova and Roy Porter, eds., Images of the earth (Chalfont-St. Giles: British Society for the History of Science, 1979), 119–46, esp. 136; Caroline Hannaway, "Discussion," in Abraham Lilienfeld et al., eds., Times, places, and persons, aspects of the history of epidemiology , supplement to the Bulletin of the history of medicine (Baltimore: Johns Hopkins University Press, 1980), 39–42, on 40; Jean-Pierre Peter, "Disease and the sick at the end of the eighteenth century," in Robert Forster and Orest Ranum, eds., Biology of man in history , transl. Elborg Forster and Patricia M. Ranum (Baltimore: Johns Hopkins University Press, 1975), 103–5.

needs. Climatology became an essential component of their programs. So strong, in fact, was the link between climatology and medicine in France during the 18th century that the French term "température" retained its ancient medical significance: it meant not the degree of heat but the "temperament" or constitution of the atmosphere.^[45]

The second stimulus to climatology was exact experimental physics. A call for reform of instruments was sounded by the Genevan meteorologist Jean-André Deluc in 1772, in his Recherches sur les modifications de l'atmosphère , which included extensive historical and critical surveys of barometers and thermometers and his own design for a portable barometer, accurate to between one-eighth and one-sixteenth of a line.^[46] Deluc discussed problems of parallax in taking readings and the proper point of the meniscus from which to read the mercury column. He demonstrated the importance of boiling the mercury in the barometer tube, a procedure that multiplied the accuracy of the instrument by a factor of 10. He showed how to use the barometer and thermometer in extensive, systematic measurement, taking all possible precautions to avoid disturbing factors. In order to establish a barometric rule for heights (i.e., a formula relating barometric pressure to altitude), he took over 400 measurements of temperature and pressure at fifteen stations on a mountain near Geneva, using great care in the exposure of the instruments and correcting for the effect of heat on both the barometer and the

[45] Note the title of Malouin's memoirs, "Histoire des maladies épidémiques . . . en même temps que les différentes températures de l'air." See also the usage of the term by Turgot and Cotte and in medical topographies, as discussed below.

[46] Deluc, Recherches . A line is 1/12 inch.

column of air whose height he was measuring. Later he made eighty-seven observations atop Geneva's cathedral. Such methods were a great novelty, and the Recherches were hailed as "a revolution in this part of physics."^[47]

The revolution spread quickly. Within a few years Jesse Ramsden, the great English instrument-maker, was constructing barometers accurate to 1/500 or 1/1,000 inch.^[48] With them two surveyor-mathematicians, William Roy and George Shuckburgh, measured mountain heights to within 0.2 and 0.7 percent, adopting Deluc's methods and his rule for heights.^[49] Hygrometry—the science of measuring humidity—presented greater difficulties than barometry and thermometry, but in the 1770s and 1780s Deluc and Horace Bénédict de Saussure designed reliable, reasonably accurate, and in de Saussure's case, comparable hygrometers. By the 1790s Alessandro Volta was measuring saturation quantities of water vapor and aqueous vapor pressures to within 4.5 percent.^[50] Around the same time he and other electricians devised sensitive electrometers, although they were not sure just what their instruments measured.^[51] Everyone knows about the adoption of exact instruments and methods in chemistry in France.^[52]

These methods transformed experimentation from a descriptive art to a quantitative science. They provided the kind of numerical data that both inspired and confirmed mathematical laws. With their emphasis on discipline, rationalization, and standardization, they were closely related to the bureaucratic impulse of the late Enlightenment.^[53]

[47] Cotte, Traité , xii.

[48] Liverpool Papers, British Library, Shuckburgh in Add. MS. 38,481, ff. 1v–4r, Roy Papers, Public Record Office, Kew, MS. WO30/119, f. 26v.

[49] Feldman, "Applied mathematics," 178.

[50] Feldman, The history of meteorology , 106, 71–2.

[51] Heilbron, Electricity, 451.

[52] French chemists relied on instruments made by compatriots—Alexis Megnié and Nicolas Fortin, among others. Their instruments were of a precision comparable to Ramsden's. Cf. Daumas, "Precision measurement."

[53] Charles Gillispie has pointed out that quantification in the late Enlightenment often meant standardization, rationalization, and a "spirit of accountancy." Science and polity in France at the end of the old regime (Princeton: Princeton University Press, 1980), esp. chap. 1, sect. 6 and p. 65.

The two reform movements—exact experimental physics and medical and agricultural reform—offered solutions to the most pressing problems of observation and climatology. Instruments were generally more reliable and accurate. Exact experimenters stressed careful specification of their make and scales. Together with the general adoption of the Fahrenheit and Réaumur scales after midcentury, this meant that the readings of different instruments could be reliably compared. The emphasis on discipline in measurement meant that observers would now read their instruments several times daily instead of the single reading common earlier in the century. (Jurin had requested one daily reading by his observers.) They would keep their weather diaries over a period of years rather than submit an out-of-date register covering a single year or less. Bureaucrats in medical and agricultural police were anxious to recruit these observers into their programs and, especially in France, had the means to enforce a proper "labor discipline." A coherent collection of weather observations became possible, and with it a quantitative climatology. A survey of late 18th-century meteorological institutions will show just where and to what degree all these possibilities were realized.

The Smaller Societies

In the German states a number of agricultural, economic, and patriotic societies took up meteorology.^[54] Their emphasis on a friendly, amateurish atmosphere and a preponderance of bureaucrats among their membership^[55] did not favor rigorous observation. Typically they published occasional weather observations, made with or without instruments, and reports of unusual or hard weather and its effects on crops and public health. The Gesellschaft der Naturforschenden Freunde at Berlin was one of the more active groups. Over two decades (1775–95) the society published some twenty reports of unusual weather, accounts of fog, snow, northern lights,

[54] Henry Lowood gives an excellent account of the more than 200 such societies in Patriotism, profit, and the promotion of science in the German Enlightenment: the economic and scientific societies, 1760–1815 (Ph.D. dissertation, University of California, Berkeley, 1987; DAI 49/06-A, 1563).

[55] Ibid., 39, 76.

and the like, as well as descriptions of improved meteorological instruments. These were brief, elementary discussions. A paper of 1787 used rainfall measurements made in Berlin in the 1730s to find monthly precipitation, numbers of rainy days, and the average precipitation on rainy days of each month.^[56] Another contributor described a lightning rod that doubled as an electroscope for atmospheric electricity. He used the apparatus intermittently: "I had no opportunity to observe during the whole of 1791; there were few storms, and [they occurred] at inconvenient times."^[57] The use of data half a century old and a casual attitude toward observation suggest that exact experimental physics had not penetrated the Gesellschaft der Naturforschenden Freunde.

Several German societies organized meteorological networks. In Silesia Ignaz Felbiger, abbot of the Monastery of Our Lady at Sagan, established a network under the auspices of the Patriotische Gesellschaft at Breslau. Felbiger's own enthusiasm for meteorology yielded papers on lightning rods, on the art of weather observation, and on the cold winters of 1783–5,^[58] and prompted him to correspond frequently with Johann Heinrich Lambert, who advised him about organizing observers.^[59] Felbiger saw the project as a second but better Breslauer Sammlung : "with the help of mathematics the necessary instruments and methods of observation have attained a far higher level of perfection since [Kanold's] time; without this exactness and precision it is impossible to compare observations."^[60] Felbiger was ambitious: his observers were to record the temperature in Fahrenheit degrees and tenths; the barometer in Paris inches, lines, and tenths; weather conditions and cloud cover; quantity of rain; wind direction and strength; humidity, by means of Lambert's new hygrometer; optical phenomena such as rainbows and haloes; and the phase of the moon. These specifications, which

[56] Rosenthal, "Bestimmung des Ganges des Niederschalges zu Berlin," Gesellschaft der Naturforschenden Freunde, Berlin, Beobachtungen, 7 (1787), 484–9.

[57] J.P. Pelisson, ibid., 398–407.

[58] Hellmann, Repertorium , 126.

[59] Lambert's considerable contributions to meteorology are discussed in Feldman, "Applied mathematics" and The history of meteorology .

[60] Patriotische Gesellschaft, Breslau, Ökonomische Nachrichten, 1 (1773), 2.

would have been impossible a decade earlier, reflect Felbiger's reading of Deluc's newly published Recherches .^[61] Felbiger also stipulated standard times of observation—soon after dawn, midday (1–4 p.m.), and evening (around 10 p.m.)—and he provided scales for cloud cover, fog, rain, wind, and snow. He had set his sights too high, however. The few observers responding submitted only occasional contributions and several did not use instruments.^[62] An invitation to cooperative weather observation from Prague's patriotic society to "all Bohemian patriots" met with even less success: the organizer found himself alone in supplying the society's journal with observations.^[63]

Western European projects achieved better results than did those in central and eastern parts. In France the Société royale d'agriculture de Paris, founded in 1761, had lapsed into inactivity by the 1770s. It was revived in 1785 by a drought and by the growing urgency of agricultural reform. The Intendant of Paris, under whose jurisdiction the Society fell, appointed the energetic Auguste Broussonet as its permanent secretary, and the Academy began publishing Mémoires , holding public meetings, awarding prizes—in short, adopting the demeanor of a learned academy of the Enlightenment.^[64] Between 1785 and its demise in 1793, the Society published in its Mémoires eight sets of "Observations géorgico-météorologiques," elicited by questionnaires distributed by the local authorities, and a paper on the cold winter of 1789. The observations, more georgical than meteorological, included detailed topographical descriptions, general monthly accounts of weather, descriptions of the effect of weather on crops and animals, harvest quantities, and grain prices.^[65]

[61] Ibid., 79.

[62] For example, Kretzschmer, "Wetterbeobachtungen in Braunau, 1778," Ökonomische Nachrichten , 6 (1778) and Freytag, "Einfluss der Witterung," ibid., 113–4, 180–1, 220–2.

[63] Strnad, "Meteorologische Beobachtungen," 397–9.

[64] Gillispie, Science and polity , 370–1, 375–6.

[65] De Courset (le Baron), "Observations géorgico-météorologiques faites dans le Boulonnais," Société royale d'agriculture, Paris Mémoires , 1786–8; Gallot, "Observations géorgico-météorologiques," ibid., 1787.

The mix of topics illustrates nicely the affinities between meteorology and agriculture.

Bern's Ökonomische Gesellschaft, which included among its members Albrecht von Haller, Daniel Bernoulli, and the important instrument-maker Michel du Crest, organized on a more ambitious scale. Since 1760 the Gesellschaft had published "Observations rurales" from seven towns and villages of the Canton. The observations included qualitative descriptions of the weather and its medical and agricultural effects; tables of monthly rainfall; degree summations of heat and cold at morning, noon, and night; and extremes of temperature and pressure. (Degree summations are sums of temperature readings taken over a given period. They indicate the total amount of heat available to plants and are a characteristic innovation of late 18th-century climatology, with its interest in agricultural applications.) In 1763 the Society resolved "to establish meteorological observers in at least six different places in the Canton, and to supply them with exact instruments." The network was part of a plan to collect information on topography, climate, and disease (i.e., on "airs, waters, and places") and on agricultural and industrial resources. It lasted a decade before submissions petered out.^[66]

The Big Projects

These smaller societies lacked personnel and resources for significant meteorological research. The great projects of the late Enlightenment were carried out by institutions that enjoyed generous funding and the full support of the state: the Societas meteorologica palatina and the Société royale de médecine. The Royal Society of London, whose connection with the state was more tenuous, provides an interesting counterpoint.

The Royal Society had maintained its interest in meteorology after the end of Jurin's group, publishing about 150 meteorological papers between 1750 and 1770.^[67] Most of these were brief notices of spells

[66] Ökonomische Gesellschaft, Bern, Mémoires , 1763–72; quotation is in Mémoires , 1763, xvii; cf. Lowood, Patriotism , 202–3.

[67] Feldman, The history of meteorology , 213–4.

of unusual or hard weather or of rainbows, haloes, and other optical phenomena, submitted by provincial correspondents. Typical were "Observations of the late severe cold weather" of the winter of 1753–4 and "Concerning a very cold day, and another very hot day" in 1748—observations of at most a few days' weather, reporting only its temperature and general character and using outdated instruments.^[68] Between 1750 and 1770 brief reports like these outnumbered meteorological registers (weather diaries covering a longer period) by almost two to one.

The contrast with the two decades after 1770 is striking. Between 1770 and 1790 registers and theoretical and experimental papers on meteorology outnumber brief, descriptive reports by eight to five. Several registers covered five or six years' weather; one spanned ten times as much. Contributors used instruments from the best artisans—Nairne and Ramsden, for example—and took care to describe their construction, calibration, and exposure.^[69] Several contributors were on assignment for the East India or Hudson's Bay Companies and observed the weather at their employers' and the Royal Society's request. The arrangement reflects the kind of relations the Society enjoyed with semigovernmental commercial and exploring ventures.^[70] The Society published their registers and those of provincial English correspondents in full.

In 1774 the Royal Society began keeping its own register of observations. Henry Cavendish submitted the Society's instruments to an

[68] William Arderon, in PT, 48 (1753–4), 249–50, and Henry Miles, in PT, 46 (1749–50), 208–13, resp.

[69] The six-decade register is Thomas Barker, "Register of the barometer, thermometer, and rain at Lyndon," PT, 61–88 (1771–98); the first year included a register of rainfall from 1736 to 1771. William Roxburgh, "A meteorological diary kept at Fort St. George in the East Indies," PT, 68 (1778), and 80 (1790), observed at three fixed times daily with a Nairne thermometer and Ramsden barometer.

[70] The best known example involving meteorology is reported in Thomas Hutchins, "An account of the success of some attempts to freeze quicksilver, at Albany Fort in Hudson's Bay," PT , 1776, 174–81, and Thomas Hutchins and Charles Blagden, "Experiments for ascertaining the point of mercurial congelation," ibid., 1783, 303–70; Roxburgh at Fort St. George provides another example. See also "Weather observations from Nain and Okak at Labrador," "Meteorological journal kept at Port Jackson in New South Wales," and others, Royal Society of London, Meteorological Archives, mss. 143, 146, etc.

exhaustive exact experimental investigation prior to the institution of twice-daily observations, which continued until 1843.^[71] The registers, printed in full in the Philosophical transactions , included readings of exterior and interior thermometers, barometers, hygrometers, and instruments to measure wind, magnetic variation and dip, and rainfall. Together with the contributions of the Society's correspondents, they constituted a far more extensive, detailed, and reliable source of data than had been available earlier in the century. The Society's registers were analyzed by its secretary Samuel Horsley, who calculated monthly mean, extreme, and mean morning and evening temperatures, extreme and mean pressures, the number of days on which the wind blew from each of the sixteen points of the compass, and the proportional rainfall in each month and season of the year.

Horsley also tested for the influence of winds and of the moon on the barometer. The moon was supposed to affect the weather by causing atmospheric tides; because its motions were periodic, a long enough series of observations might reveal correlations that would enable meteorologists to predict the weather. The theory ("so improbable, so destitute of all foundation," wrote Horsley) was eventually discredited, although adherents could be found well into the 19th century. As for the influence of wind direction on the barometer, this hypothesis contributed in the 19th century to the theories of wind rotation of H.W. Dove and others. Counting mean monthly barometer heights and corresponding directions of monthly prevailing winds, Horsley found a correlation: the barometer stood higher in months with winds in the semicircle WSW-W-N-NE; W and NE winds accompanied the greatest mean heights, and in seven months out of twelve the highest barometer reading occurred during a NE wind. Horsley must have been surprised when his count of the number of changes in the weather occurring within three days of the moon's syzygies and quadratures supported the hypothesis of lunar influence. He remarked only that the observations had not continued long enough to draw a firm conclusion.^[72]

[71] Henry Cavendish, "An account of the meteorological instruments used at the Royal Society's house," PT, 66 (1776), 375–401; Cavendish, The scientific papers , vol. 2, Chemical and dynamical (Cambridge: The University Press, 1921), 53.

[72] Samuel Horsley, "An abridged state of the weather at London in the year 1774," PT, 65 (1775), 167–93.

While Horsley was calculating, an epidemic—or, more properly, an epizootic—of cattle plague was raging through France. The quarantine and slaughter of thousands of infected cattle became necessary. Anne-Robert Turgot, Louis XVI's minister of finance, encountered bitter resistance from the peasantry; faulty communications between the central authorities and physicians in the provinces further inhibited effective action. In order to direct and enforce the necessary measures, Turgot called forth a Commission de médecine aux maladies epidémiques et epizootiques in 1776. It received letters patent in 1778 as the Société royale de médecine.^[73]

The Society was charged with the medical police of France, formerly the responsibility of the intendants. Working through them, it centralized public health policy in a single state agency staffed by trained physicians. Its duties were varied: to regulate the distribution of patent medicines and mineral waters; deploy a network of provincial physicians to gather information on public health and the environment; establish sanitary measures and regulations; advise provincial physicians on the prevention and treatment of epidemics; and, in cases of epidemic outbreaks, coordinate and enforce appropriate measures.^[74] The Society also took on all the attributes of a learned academy: titles of membership signed by its president and secretary, medals, prize questions, éloges , and a journal whose title, Histoire de la Société royale de médecine, avec les mémoires de médecine et de physique médicale, echoed that of its elder sister, the Paris Academy of Sciences.

Two aspects of the Society's operations concerned climatology. The first was the construction of "a medical and topographical map of all of France," toward which physicians "of all the cities of the realm" would contribute memoirs "on the nature of their climate (de aere, acquis, et locis ), and on the temperament of those who live

[73] Gillispie, Science and polity , 24–33; Jean-Paul Desaive et al., Médecins, climat et épidémies à la fin du XVIII siècle (Paris: Mouton, 1972), 9; Caroline Hannaway, "The Société Royale de Médecine and epidemics in the Ancien Régime," Bulletin of the history of medecine, 46 (1972), 257–73, on 259ff.

[74] Gillispie, Science and polity , 219ff. and 226ff.; Hannaway, "The Société Royale de Médecine," 258ff.; Desaive et al., Médecins, climat et épidémies , 11–2.

there."^[75] The best of these medical topographies, as they were called, won annual prizes from the Society. They might cover nearly 100 pages of text with complete environmental and geographical descriptions of the author's region, including climate and topography, economy, hygiene, and endemic diseases.^[76] By the outbreak of the Revolution, 226 topographies had been collected.^[77]

The Society's network of provincial physicians also carried out a grand program of meteorological observation. Already in 1775, in the midst of the epidemic, Turgot had tested the waters by distributing questionnaires via the provincial intendants, asking physicians in their jursidictions for "the temperature and the [epidemic] constitution of the years 1772, 1773, 1774, and 1775."^[78] The resulting medicometeorological correspondence was continued by the Commission for Epidemics. After 1778, the authors of this correspondence became the Society of Medicine's network of observers.^[79]

The direction of the network was undertaken by Louis Cotte, Oratorian priest and France's foremost meteorologist. Just a few years earlier Cotte had published the first textbook of meteorology based on observations—that is, the first to include many observations at all and to derive its discussions of the weather from them.^[80] Cotte was especially interested in agricultural applications of meteorology. He had originally planned his Traité de météorologie as an extension of Duhamel's "botanico-meteorological" investigations; it is agriculture, he wrote, "which I have had principally in view in this work." He included medicometeorology as well, excerpting Malouin's

[75] Société royale de médecine, Paris (SRM), Histoire , 1776, xiv; ibid., 1782–3, 6–7.

[76] For example, de Brieude, "Topographie médicale de la Haute-Auvergne," SRM, Mémoires , 1782–3, 257–340.

[77] Hannaway, "The Société Royale de Médecine," 267.

[78] Ibid., 3; "Questions à faire aux médecines," France. Académie nationale de médecine, Archives, SRM Carton 189, dossier 13, number 16.

[79] France. Académie nationale de médecine, Archives, SRM Carton 172, dossiers 1–22 cf. Gillispie, Science and polity , 33, 202–3.

[80] Cotte, Traité de météorologie . Cf. Gustav Hellmann, "Entwicklungsgeschichte des meteorologischen Lehrbuchs," in Hellmann, Beiträge zur Geschichte der Meteorologie, 3 (Berlin, 1922), 1–14, on 1.

memoirs extensively. The goal of meteorology, he believed, is "the perfection of the sciences of agriculture and medicine."^[81]

In the Royal Society of Medicine, Cotte had observers to pursue that goal. The requirements he imposed upon them were, in the words of one modern meteorologist, "very strict," and, in the judgment of another, comparable to mid19th-century standards.^[82] "It is necessary first of all," Cotte warned, "to have good instruments." A Réaumur mercury thermometer was to be calibrated if possible against the Society's standard; the barometer, calibrated "with the greatest exactitude," should be equipped with a vernier and readable to an accuracy of tenths or twelfths of a line. The Society would recommend reliable artisans on request. Cotte specified the proper exposure for the instruments and set fixed times for observation.^[83] "Great exactitude and a spirit of order—these are the principal requirements of the physicist who devotes himself to these sorts of observations."^[84] These words might serve as a motto for late-Enlightenment physical science.

Cotte complained more than once that his observers' instruments were not comparable, that observers were not providing "an exact description" of their instruments, that the instruments were "defective. . .supplied by travelling barometer-peddlers." But he had to admire his observers' zeal.^[85] By 1785, 150 provincial physicians were participating in the project; about 50 of them observed for more than a decade.^[86] Modern climatologists agree that the Society's observers were the elite of the medical profession—otherwise they could not have followed Cotte's instructions at all. They generated a great mass of observations over the whole of France—the largest collection, most likely, prior to the foundation of national meteorological bureaus in the mid19th century. From these observations

[81] Cotte to Macquer, 25 Aug 1768, France. Bibliothèque nationale, ms. Fr. 12305, ff. 209–11; Cotte, Traité de météorologie , ix, xvii.

[82] Desaive et al., Médecins, climat et épidémies , 90; J.A. Kington, "A late eighteenth-century source of weather data," Weather, 25 (1970), 169–75.

[83] SRM, Histoire , 1776, xi–xiv.

[84] Cotte, Traité de météorologie , 517.

[85] SRM, Histoire , 1777–8, 104; ibid., 1782–3, 245–6; ibid., 1784–5, 204.

[86] Ibid., 203; Peter, "Disease and the sick," 85, 87.

meteorologists have been able to reconstruct the climate of France in the last years of the Old Regime.^[87]

Cotte did not publish these observations in full; "Tables are not pleasant for the reader," he said, echoing Derham across a half-century.^[88] Instead he published monthly summaries of the weather at each reporting location, including its "temperature" in terms of the Hippocratic categories cold and wet, cold and dry, warm and wet, and warm and dry, along with the usual means, extremes, and descriptions of diseases. Although the summaries resembled those of Derham's time, they were more plentiful and more consistent. Cotte arranged them in great tables (tables were not to be dispensed with entirely) that amounted to the first descriptions of the climate of a nation based on detailed, regular observations.

Cotte's tables represented a collection of data dense enough over a sufficient area to reveal something of the importance of space for the weather. "These extreme temperatures take place," he said, "at the same time over a very great extent of country."^[89] This remark, made early in the Society's career, sounds like one of Derham's coincidences, only it covers more ground. A few years later Cotte wrote that the tables showed clearly "the influence of great variations of the atmosphere over a very great extent of country on the thermometer, and principally on the barometer."^[90] These isolated suggestions mark the limit of Cotte's insight. He did not draw from the tables a description of the climate of France. Moreover, the tables' monthly means and extremes could not reveal the density of the weather in time. Because Cotte did not publish the full record of observations, his observers' efforts were of as little use to other meteorologists as they seem to have been to himself. They lay hidden in the Society's archives until climatologists began to explore them around the middle of the present century.

[87] Desaive et al., Médecins, climat et épidémies , 40–2, 97. O. Muller, ibid., has submitted the Society's observations to correlation analysis. Many, especially those of Paris and the Ile de France, show a correlation coefficient of 0.8 and higher.

[88] Cotte, Traité de météorologie , 517.

[89] "Temperature" here is taken in a broad sense to include barometric pressure, rains, storms, and drought. SRM, Histoire , 1777, 105.

[90] SRM, Histoire , 1780, 261.

The Societas meteorologica palatina has enjoyed greater fame than its French sister, in part because it was an international organization whose members were the chief scientific institutions of Europe (more on this below). More significantly, the Palatine Society published the full record of its members' observations, which were excellent and became a valuable resource for 19th-century meteorologists.^[91] The Society was a product of a late 18th-century enlightened state, the principality of the Palatinate or Kurpfalz. Cultural life flourished there under the patronage of the Elector Karl Theodor, who was keen on science and its contribution to his subjects' welfare.^[92] Mannheim, Karl Theodor's capital, was one of Germany's centers of culture; the Elector lavished on it an Akademie der Wissenschaften, a Deutsche Gesellschaft, and a Deutsches Theater, and brought Gotthold Ephraim Lessing, Friedrich Gottlieb Klopstock, and Wolfgang Amadeus Mozart to write and play for him. He furnished an astronomical observatory with English instruments, then the best in the world, and directed his court chaplain and ecclesiastical councillor, Johann Jakob Hemmer, to assemble a cabinet de physique and offer lectures and demonstrations.^[93] In the 1770s Hemmer investigated the electricity of flames, of dew, and of the atmosphere, and he was largely responsible for the erection of lightning rods throughout the Elector's extensive realm. His book on the subject went through two editions. Many German rulers bought it for distribution in their principalities.^[94]

The idea for a Palatine meteorological society may have come from neighboring Baden. There, in 1778, the Karlsruhe professor of mathematics and physics Johann Böckmann founded the Badische

[91] For example, H.W. Brandes derived the first synoptic analysis of European weather in 1823 from the Society's record for the year 1783. See Brandes, Beiträge zur Witterungskunde (Leipzig, 1820).

[92] Hellmann, Repertorium , 896.

[93] David Cassidy, "Meteorology in Mannheim," 12; Hellmann, Repertorium , 193.

[94] J.J. Hemmer, Anleitung, Wetterleiter an allen Gattungen von Gebäuden auf die sicherste Art anzulegen (Mannheim, 1786, 1788). Cited by Adolf Kistner, Die Pflege der Naturwissenschaften in Mannheim zur Zeit Karl Theodors (Mannheim: Selbstverlag des mannheimer Altertumsverein, 1930), 87; Cassidy, "Meteorology in Mannheim," 14–5.

Witterungsanstalt, a network of sixteen observers within the margravate.^[95] The project failed for lack of funds, but it probably inspired Hemmer, who hired Böckmann's instrument-maker and placed his own plans before the Elector.^[96] Karl Theodor, who himself kept a weather diary, approved, and in 1780 the Societas meteorologica palatina received its charter as a three-member "Meteorologische Klasse" in the Mannheim Akademie der Wissenschaften. Hemmer, the Society's secretary, recruited observers, oversaw the construction of instruments for them, and edited the annual Ephemerides meteorologicae .

The Society's work was a model of exact experimental method. Hemmer followed Deluc closely in specifications for instruments. "We shall always search for ways to make observations more exact," he declared, "both for the sake of agriculture and for our health." Tubes of barometers and thermometers were to be exactly cylindrical, carefully cleaned, and filled with pure, boiled mercury, and the scales carefully prepared and exactly measured. Barometers were to be read to tenths of a line with the vernier against the cusp of the meniscus and thermometers to tenths of a degree. Care was to be taken to prevent disturbances from the observer's breath or candle. Observers could use their own instruments if they met these standards. Hemmer nonetheless dispatched to each of them a packet of instruments "whose comparability has been assured, sparing no expense": a barometer with correcting thermometer (the thermometer to correct for the effect of heat on the barometer), two thermometers (for exposure in sun and shade), and a hygrometer of Deluc's design (using a goose-feather quill that expanded with humidity). Selected observers received a magnetic declination compass as well. Hemmer also asked participants to observe wind direction and strength, cloud cover, precipitation and evaporation, river level, lunar phase, and medical and agricultural conditions. Hours of observation were set at 7 a.m., 2 p.m., and 9 p.m.; Hemmer supplied scales for wind

[95] J.L. Böckmann, Beiträge zur neuesten Geschichte der Witterungslehre (Karlsruhe, 1781), cited in Hellmann, Repertorium , 889.

[96] Kistner, Die Pflege der Naturwissenschaften , 96–8; Cassidy, "Meteorology in Mannheim," 17–8.

strength and cloud cover and symbols for precipitation and other "meteors."^[97]

Along with these exacting specifications went a new strategy of recruitment. In order to secure continuity in observations, Hemmer invited scientific institutions rather than individuals to participate; each institution would appoint one of its members to observe. No fewer than thirty-seven academies, universities, and monasteries in Europe and America responded, and Hemmer soon had an international network staffed by the world's capable workers. Their registers—reliable, uniform, and, in a number of cases, extending over a decade—appeared in extenso in the Society's Ephemerides . This was a "princely plan of operations," in the words of the 19th-century meteorologist John Daniell,^[98] carried out at princely expense: 1,500 gulden annually for the Ephemerides alone, or about a third more than the salary of a senior member of France's Académie royale des sciences.^[99] The resulting collection, wrote Daniell, "contain[ed] more data for a correct history of the weather than all other works on the same subject taken together."^[100] Such was the stature of the Palatine Society that its hours of observation—the so-called "Mannheim hours"—and the "Mannheim cloud cover and wind scales" were widely adopted and remained in use for more than a century.

Several of Hemmer's observers carried out extensive climatological analyses of their own weather records. Typical was the contribution of Nicholas de Béguelin of the Academy of Sciences in Berlin. He reported monthly maxima and minima for the thermometer and barometer, their monthly ranges, and annual and monthly means; the same for morning and evening thermometrical observations, together with maximum diurnal range; for the hygrometer and magnetic needle monthly mean morning, afternoon, and evening values, the mean of all three, and monthly extremes and extreme ranges; and monthly

[97] Societas meteorological palatina, Ephemerides meteorologicae, 1 (1781), 21, 7–14. See Hellmann, Repertorium , 898.

[98] John Daniell, Meteorological essays , 3d ed., 2 vols. (1845), 305.

[99] Cassidy, "Meteorology in Mannheim," 19; for salaries, see Heilbron, Electricity , 117, 122.

[100] Daniell, Meteorological essays , 305.

frequencies for each direction and degree of wind strength and for each degree of cloud cover, all according to Hemmer's scales.^[101] Participants contributed a wealth of other material, which Hemmer published indiscriminately: ten years' tidal observations from Padua, hourly barometric measurements over an entire month, hourly observations of magnetic declination from 6 a.m. to 10 p.m., Hemmer's own measurements of atmospheric electricity, and van Swinden's collection of temperature observations taken four and five times daily in nine Belgian towns during a cold spell in December 1783. From analyses of contributors' registers, Hemmer's assistant Karl König published tables like Cotte's but far more complete. A single table for each reporting location presented monthly summaries. A "general table" of all locations showed annual extremes, means, and ranges; extremes of monthly means for barometer, thermometer, and declination-needle; annual rainfall and prevailing winds; monthly sums of degrees of heat; and frequencies of wind strength and direction, "meteors," and cloud cover.

Although the collection was something of a miscellany, the observations and analyses were more consistent and rigorous than anything earlier in the century. In them we can see the first hints of a discipline of climatology: a standard and consistent practice in the gathering and analysis of weather data. This practice is reflected in the emergence of a standard set of climatological variables calculated by meteorologists such as Béguelin, König, and Horsley. Their work marks the beginning of a quantitative climatology.

The new climatology provided some definitive answers to old questions. Van Swinden, in an independent investigation, and two contributors to the Society's Ephemerides examined barometric observations taken five, six, and eighteen times daily over a year or more. They refuted the hypothesis of lunar influence but confirmed the diurnal variation, a daily fluctuation of the barometer that several earlier meteorologists thought they saw in their data.^[102] These analyses

[101] Ephemerides , 1784, appendix, 3–6.

[102] Van Swinden, Mémoire sur les observations météorologiques faites à Franeker pendant le courant de l'année MDCCLXXIX (Amsterdam: M.M. Rey, 1780), 34; Joh. Jac. Planer, "Observatio oscillationis mercurii," Ephemerides , 1783, 250–7; Vincent Chiminiello, "De diurna nocturnaque oscillatione barometri," ibid., 1784, 230–4.

exploited the density of observations in time . The coverage of observations in space led meteorologists to a greater appreciation of its role in the weather. They could now trace the paths of weather events from one end of Europe to the other: the cold wave of the winter of 1775–6, for example, and the famous hailstorm of 13 July 1788, which by destroying crops across the most fertile regions of France contributed significantly to the trouble preceding the Revolution.^[103] And they arrived at a clear statement of the spatial extent and of the motion of barometric variation. "Anyone who carefully examines and compares the barometric observations in volume I of the meteorological Ephemerides ," proclaimed Coelestin Steiglehner, "cannot fail to conclude that oscillations longer than one day extend over many places of diverse longitude and latitude."

From observations at London, Regensburg, and St. Petersburg, Steiglehner determined that a barometric minimum occurring around Christmas 1775 had traveled from west to east, striking each town in succession. Further analysis of data from London, Regensburg, St. Gotthard, Buda, Mannheim, and Vienna confirmed the rule "early in the west, late in the east."^[104] F.X. Epp had drawn a similar conclusion from the observations of the network of twenty-four Bavarian observers he directed. The Churbayrische Akademie der Wissenschaften, a sister society to the Mannheim Academy after Karl Theodor inherited Bavaria in 1777,^[105] established the network in the same year that the Palatine Society was chartered. From parallel barometric motions recorded by his observers Epp concluded that the causes of barometric variation extend over wide regions, perhaps over

[103] Van Swinden, Observations sur le froid rigoreux du mois de janvier, MDCCLXXVI (Amsterdam: M.M. Rey, 1779), 5–6; Cotte, in the "Feuille météorologique" of the Journal de france , cited in Buissart, "Mémoire sur l'orage du treize juillet, 1788," France. Académie nationale de médecine, SRM carton 141, dossier 40. On effects of the hailstorm, see Robert D. Harris, Necker and the revolution of 1789 (Lanham, Maryland: University Press of America, 1986), 273; Michel Vovelle, The fall of the French monarchy 1787–1782 , transl. Susan Burke (Cambridge: Cambridge University Press, 1984), 86.

[104] Coelestin Steiglehner, "Excerpta ex dissertatione: atmosphaerae pressio varia," Ephemerides , 444–57.

[105] See James E. McClellan III, Science reorganized. Scientific societies in the eighteenth century (New York: Columbia University Press, 1985), 118.

hemispheres. He confirmed this conclusion by examining one month's barometric variations in five European towns, the data for which he found in the Palatine Society's Ephemerides.^[106]

Finally, a few meteorologists began to synthesize descriptions of regional climates from observations. The Society of Medicine's medical topography project aimed at a "medical and topographical map of all of France." Van Swinden planned to determine the climate of Frisia from his own observations, and from others' observations throughout Holland he began to construct a picture of its provincial climates.^[107] Epp expected from his collection "a more exact knowledge of [Bavaria's] climate" and of the "physical character of the land."^[108] If the stations were properly distributed the observations would yield information on local climates as well.^[109] None of these projects, however, was completed.

Conclusion

The French revolutionary period brought a precipitous decline to meteorology. The correspondence of the Société royale de médecine ceased with the onset of the Revolution in 1789; along with all the Old Regime academies the Society was suppressed in 1793.^[110] The Palatine Society, already in decline after the death of Hemmer in 1790, collapsed in 1795 when the French army crossed the Rhine, occupied Mannheim, and closed the Academy of Sciences.^[111] In Germany the number of stations observing the weather dropped during the 1790s to a third of its peak value in the previous decade.^[112]

[106] Akademie der Wissenschaften, Munich, Ephemerides, 1 (1781), 14; ibid., 2 (1782), 39–43.

[107] Van Swinden, Observations sur le froid rigoreux , 134–5.

[108] Akademie der Wissenschaften, Munich, Ephemerides, 1 (1781), 4–5.

[109] Akademie der Wissenschaften, Munich, Anzeige an das Publikum, von den Gegenstände der Witterungslehre, und von der Art und Weise die Witterung zu beobachten (Munich, 1781).

[110] Busquet, "Les origines de l'Académie de médecine," Société française d'histoire de la médecine, Bulletin, 20 , (1926), 391–6; Cotte to Thouret, Montmorenci, 9 thermidor an 4 (1796), France. Académie nationale de médecine, Archives, SRM carton 197, dossier 3, number 3.

[111] Cassidy, "Meteorology in Mannheim," 19–20.

[112] Hellmann, Repertorium , 985.

Meteorological contributions to the Royal Society of London's Philosophical transactions had already diminished in the second half of the 1780s.

No doubt other factors besides the physical disruption of war and revolution contributed to the decline. Politics distracted natural philosophers; van Swinden, for example, became involved in administrative and educational reform. When in 1800 Cotte requested some earlier observations from him, van Swinden wrote in regret of "the oblivion to which I have consigned them, regretting all the while that so much research produced so few useful results."^[113] His remarks suggest a further reason for the loss of interest in meteorology. Meteorologists like Cotte had promised "to prove the usefulness of meteorological observation to those who deride [it]."^[114] Nothing positive seemed to have come from the expenditure of so much effort on what many considered a tedious and trivial activity.

The program of late Enlightenment climatology had failed. Meteorology had not brought "the perfection of the sciences of agriculture and medicine." No useful correlations had been discovered between the weather and agriculture or disease, and correlations like the lunar influence that might lead to predictive rules remained in doubt. "All attempts at rules governing the weather have been in vain," remarked the astronomer Bode.^[115] There were a number of causes for this failure. The meteorological projects of the late Enlightenment still confronted some of the limitations of the earlier decades of the century. Instruments had neared perfection but remained unavailable in the more remote French provinces and German states. Rigorous standards had been established for observers but were not always followed outside scientific centers. Institutional arrangements were weak in many areas; this was especially true in central and eastern Europe, where projects suffered from the small size and political fragmentation of the German states. And editors of

[113] Van Swinden, "Lettres sur les grands hivers," Journal de physique, 50 (1800), 277–96, on 277–8.

[114] Cotte, Traité de météorologie , 248.

[115] Rudolf Wolf, Biographien zur Kulturgeschichte der Schweiz , 4 vols. (Zurich: Orell, Füssli & Co., 1858–62), 1 , 451.

meteorological ephemerides retained the practice of publishing summaries rather than the full record of observations, to the frustration of generations of meteorologists.

Cotte and his colleagues no doubt believed that they might have reached some of their goals if the Revolution had not interrupted them.^[116] They were mistaken. Their search for correlations between weather patterns and agriculture and disease belonged to the "classical" episteme , as Foucault called it, the episteme of natural history and nosology.^[117] Meteorologists working within this episteme associated the weather with diseases and agriculture according to superficial correlations, without penetrating to the interior forces that govern their interactions. But climatology (parallel with the rest of Western knowledge, if Foucault was right) was moving in the last decades of the 18th century away from this approach toward the perception of these interior relations.

It was approaching this goal through quantification. Quantification here did not include mathematical theories of the weather. Other parts of meteorology did develop mathematical theories. In hygrometry, thanks to a new hygrometer and the techniques of exact experimentation, de Saussure and others announced Dalton's law of partial pressures for the special case of aqueous vapor: the total pressure of moist air at a given temperature is the sum of the pressures of its components, water vapor and dry air. From this elementary arithmetical relation de Saussure calculated the relative weights of dry and saturated air, using nothing more complicated than simple proportions.^[118] More sophisticated mathematical laws were worked out in the 1790s governing the relation of vapor pressure to temperature.^[119] In barometry Lambert, working in the tradition of mixed mathematics, had applied the integral calculus to the variation of air pressure with height in the 1760s.^[120] Studies of the distribution of

[116] This appears from Cotte's papers and his correspondence with van Swinden during the 1790s. See, for example, Cotte, ms. preface to Nouveaux mémoires sur la météorologie (1810), Laon, France. Bibliothèque municipale, ms. 568^[2] .

[117] Foucault, Order of things .

[118] See Feldman, The history of meteorology , sect. I, chap. 4.

[119] Ibid., chap. 7.

[120] Feldman, "Applied mathematics," 145–9.

heat over the earth employed trigonometry to calculate the effect of the sun's heat in different latitudes.^[121] All these results lay in areas of contact of meteorology with the numerous disciplines discussed in the introduction to this chapter. Except for one or two isolated examples,^[122] meteorology proper—the study of weather phenomena such as rain and snow, winds, and storms, as well as climatology—lacked any mathematical theory until well into the nineteenth century.

In climatology quantification meant precise instruments and a rigorous discipline of observation. The example of the Société royale de médecine shows how closely tied this discipline was with the bureaucratizing impulse of enlightened absolutism. Once observations were collected, meteorologists analyzed them using the most elementary statistical techniques: counting (the number of days of rainfall, the number of days on which the wind blew from different quarters), taking averages and ratios (means of temperature, pressure, etc., the proportion of rain falling in each season, the proportion of changes in the weather occurring within three days of the moon's syzygies and quadratures), and establishing correlations (between the weather and diseases, among the weather at different locations and times, the weather rule). It is no accident that a discipline so close to public health and administration should rely on statistical techniques.

Limited as it was, this type of quantification pushed climatology to the verge of a new existence. The "discipline" of regular observation and intensive calculation led toward the "discipline" of climatology: the practice of deriving a standard set of variables representing the typical weather of a location. By enforcing a finer coverage in time and space, this discipline led meteorologists' perception (or "gaze," as Foucault called it) toward the creative functions of time and space in weather and climate.

[121] For example, Tobias Mayer, "A more accurate definition of the variations of a thermometer," in Eric G. Forbes, ed., transl., Tobias Mayer's Opera inedita (New York: American Elsevier Pub. Co., 1971), 53–61.

[122] For example, James Hutton, "The theory of rain," Royal Society of Edinburgh, Transactions, 1 (1788), 41–86. The mathematics involved here in elementary.

Even so, there is a tremendous gap between the climatology of the late Enlightenment and that of the early years of the 19th century. The most impressive result of late 18th-century climatology, the diurnal barometric variation, remained a correlation. The discovery of the great extent of barometric variation did not capture the full role of space in modifying weather and climate. Steiglehner's "early in the west, late in the east" still sounds like a weather rule.

The climatology of the first years of the 19th century looks vastly different. Alexander von Humboldt may stand as a representative of the new point of view. Humboldt's isotherms (lines of equal temperature), presented in an essay of 1817, are syntheses of mean annual temperatures across the globe, which present immediately to the eye the role of space in the development of climate.^[123] In the same essay Humboldt speaks naturally and comfortably of the climates of the different regions of the earth; develops comparisons among the climates of eastern and western shores of continents and among coastal, continental, and island climates; and arrives at the notion of a "climatic system"—a region in which the different factors of climate vary in a continuous manner, so that in specifying for example the mean annual temperature one has fixed by implication the range of other climatic factors such as winter and summer temperature. Foucault argued that between 1775 and 1795 naturalists found the key to the new episteme —the notion of organic structure, in the case of biology—but still applied it to the old task of classification.^[124] During the same decades exact experimental physics had provided the key to an appreciation of the role of space and time in climate and the weather, but meteorologists still applied that key to the search for correlations. A mature climatology awaited the age of Humboldt.

[123] Alexander von Humboldt, "Sur les lignes isothermes et la distribution de la chaleur sur le globe," Société d'Arcueil, Mémoires, 3 (1817), 462–602.

[124] Foucault, Order of things , 226ff.