In 1683, English astronomer and natural philosopher Edmund Halley presented a paper to the Royal Society of London addressing what he called the "Magnetical System." Of this system Halley reflected:

Nearly two and half centuries later, American geophysicist Louis Agricola Bauer delivered the 1913 "Halley Lecture" at Oxford University. In this lecture he remarked on the study of terrestrial magnetism:

Both men's statements emphasized the importance of accumulated facts over time and the enduring mysteries of the subject at hand, magnetism. Also, for Halley and Bauer, understanding terrestrial magnetism required an intimate acquaintance with magnetism.

In addition to these obvious similarities, however, the differences between Bauer and Halley are revealing as well. Published only four years before Isaac Newton's Principia (1687), Halley's paper appeared near the culmination of the "scientific revolution," a period of transformation in the institutions, goals, methods, and content of natural philosophy. Be that as it may, the science emerging from this "revolution" was not modern science. Hence, the "physics" of Halley's period was quite distinct from Bauer's "physics" more than two centuries later. These general differences regarding physics applied to their specific views on magnetism as well. While both men demonstrated considerable interest in magnetism, their motives and methods diverged greatly. Halley wrote of the "Magnetical system" and the "Saylor," while Bauer spoke of the "Earth's magnetic field" and the "physicist." While Halley's prime motivation for understanding magnetism was its navigational application, Bauer stressed its inherent scientific interest. Not surprisingly, the answers to "What is a magnet?" and "What is magnetism, in general?" differed dramatically as well.

By focusing on a portion of the two-century span between Halley and Bauer, this dissertation traces the shifting motives, methods, theories, and debates regarding the study of magnetism from 1750 to 1830. Placed firmly between the late seventeenth and early twentieth centuries, this eighty-year span witnessed transformations in the understanding of magnetism as well as changing views regarding terrestrial magnetic phenomena. Throughout the period, most investigators linked their controlled magnetic experiments with the phenomena of terrestrial magnetism. In other words, they assumed that both acted in an analogous manner. Stressing, but not limited to, the specific context of British science, this dissertation examines the study of magnetism and its links to terrestrial magnetic studies.

Before delving into these topics, a wider historiographical and historical context should be established for perspective. To this end, this chapter has three primary tasks. First, it presents an historiographical overview regarding several historians’ of science attempts to categorize changes in the physical sciences from the seventeenth to the early nineteenth century. Second, the chapter discusses specific theories of magnetism and terrestrial magnetism from William Gilbert in the early seventeenth century through Halley's work about a century later. In this discussion, technical terms and historical background are introduced which will be utilized throughout the dissertation. Finally, the chapter briefly summarizes the content of the remaining chapters.

In an article published in 1976, Thomas Kuhn argued that the history of the physical sciences could be regarded in terms of two distinct traditions, the mathematical and the experimental. The mathematical tradition, contended Kuhn, encompassed the "classical sciences" many of which began in Greek antiquity (e. g., astronomy, statics, harmonics, and geometrical optics). Furthermore, this tradition included mathematically or geometrically idealized situations such as those as presented in works of Archimedes' On Floating Bodies or Ptolemy's Almagest. In contrast, Kuhn’s experimental tradition referred to topics such as electricity and magnetism which receive infrequent attention until the early seventeenth century. In studying these phenomena, experimental evidence frequently took precedence over idealized mathematical results. Hence, Kuhn contended that this tradition embraced the "Baconian sciences," named after the inductive, empirical method espoused (but not practiced) by Francis Bacon.

Throughout the seventeenth century both traditions appealed to experiment, yet used it in distinct ways. Kuhn argued that those working primarily in the classical tradition like Galileo Galilei, Johannes Kepler, and Blaise Pascal utilized "thought experiments" often intended to confirm a conclusion known by non-experimental means. For instance, although Galileo performed many experiments, he deemed some unnecessary because the power of his reasoning had ascertained the necessary outcome in advance. On the other hand, those working primarily within the experimental tradition including William Gilbert, Robert Boyle, and Robert Hooke often decried "thought experiments." Boyle, Kuhn pointed out, harshly criticized Pascal's book on hydrostatics due to its unrealizable experiments and impossible instruments. Investigators in this tradition emphasized the use of instruments and often sought to generate new effects for the end purpose of constraining and controlling nature. They were again in a sense "Baconian," because Bacon emphasized the promise of science for practical ends. As Kuhn concluded, the experimental tradition gave rise to relatively new sciences including the study of heat, electricity, and magnetism.

In 1986, Casper Hakfoort refined Kuhn's scheme by adding what he called the "natural philosophical" tradition. Providing historical continuity, this tradition included theoretical and speculative portions of natural philosophy which were not encompassed by Kuhn’s mathematical or experimental traditions. Hakfoort noted, for instance, that Aristotelian natural philosophy attempted to provide a complete account of nature closely linked to metaphysics. Of this he explained:

In the seventeenth century, René Descartes sought to replace Aristotelian natural philosophy with a quantitative, mathematical vision of the world. Nevertheless, similar to the Aristotelian system, the Cartesian system was ideally complete, certain, and partly a priori. As such, Descartes developed basic concepts and laws primarily from his "clear and distinct ideas" rather than from actual observation or experiment.

Though Aristotle allowed an important role to observation, neither he nor the Scholastic proponents of Aristotelianism had sought a quantitative, mathematical system of the world. Similarly, despite his famous dream of creating a mathematical and measurable view of the universe, the mechanical system Descartes eventually created was neither mathematical nor quantitative. As Hakfoort pointed out:

Swirling Cartesian vortices of matter in motion were not readily quantified or mathematized. Nevertheless, Descartes' vision of a universe filled with matter in motion gained many adherents. The "mechanical philosophy," Cartesian or otherwise, appealed to many seventeenth-century natural philosophers. According to Hakfoort, attempts such as Descartes' to replace Aristotelian natural philosophy do not fit readily into Kuhn's mathematical or experimental traditions. Therefore, speculative, all-encompassing explanations belong within what Hakfoort called the natural philosophical tradition.

With these mathematical, experimental, and natural philosophical traditions in mind, the history of the physical sciences can become an analysis of their shifting relationships to one another over time. As examples, Hakfoort discussed Huygens' pulse theory of light and Newton's Opticks as illustrating three-way divisions between natural philosophical, mathematical, and experimental. Using this tripartite division, Hakfoort further argued that these traditions are essential to understanding physical sciences in eighteenth-century Germany, particularly physical optics. While such an approach has its problems, it can nevertheless be used as an organizing principle for tracing historical developments in the physical sciences, including the study of magnetism.

A related approach, explored by John L. Heilbron, examines the multifarious classifications and divisions of natural knowledge over time. For instance from the seventeenth to nineteenth centuries the shifting definition of the term "physics" reflected radically different relationships between the mathematical, natural philosophical, and experimental. From Greek antiquity through the seventeenth century, "physics" referred broadly to the study of all natural bodies, animate and inanimate. In this sense, traditional "physics" of the seventeenth century was inclusive, qualitative, and literary; the Aristotelian tradition of physics included both organic and inorganic realms. "Physics" in the seventeenth century recommended neither the use of experiment nor mathematics. As Aristotle argued, constraining nature through experimentation actually hid or distorted its true workings.

On the other hand, the lesser-esteemed study of "mixed" or "applied" mathematics included quantified physical sciences and often those in the mathematical tradition such as observational astronomy, geometrical optics, mechanics, statics, and hydraulics. It also encompassed practically oriented subjects such as geography, horology, fortification, surveying, and navigation. Thus, even though investigators including Copernicus, Galileo, and Kepler espoused the use of mathematics, introducing quantification and mathematization to seventeenth-century "physics" meant lowering its status in the established hierarchy of knowledge. Not surprisingly, most early Copernicans were mathematicians rather than natural philosophers.

By the early eighteenth century, though some continued using "physics" in its older, broader sense, many others narrowed the scope of "physics" and made it synonymous with "natural philosophy." Of vital importance in this narrowing was the work of Isaac Newton. The full English title of Newton's magnum opus, The Mathematical Principles of Natural Philosophy, signaled the rising status of mathematical methods in natural philosophy. Indeed, Newton's work elevated several areas of mixed mathematics to the level of natural philosophy. In altering its scope and methods, Newton and others restricted the domain of natural philosophy or physics. As I. Bernard Cohen explained, Newton did not merely produce mathematical constructs for "saving the phenomena," instead, he created what he considered to be "purely mathematical counterparts of simplified and idealized physical situations that could later be brought into relation with the conditions of reality as revealed by experiment and observation." In this manner, Newton successfully combined the mathematical and experimental traditions. As we shall see, he also used elements of the speculative natural philosophical tradition, particularly in the Queries to the Opticks.

During the eighteenth century, natural philosophy became ever more restricted to particular areas of study. In addition to growing mathematical content, the increasing importance of instruments and experimental techniques contributed to the transformation of natural philosophy. Heilbron argued that the introduction of the demonstration experiment or demonstration lecture (utilizing air pumps, barometers, pendulums, lodestones, etc.) contributed to the narrowing of natural philosophy for at least three reasons. First, the biological sciences did not lend themselves readily to demonstration experiments. Second, the instrument trade, long established to meet the needs of "mixed" or "applied" mathematics, could easily furnish the professor of experimental philosophy or lecturer with apparatus. Third, during the 1730s and 1740s Newton's followers, particularly Dutchmen Willem Jakob 'sGravesande and Pieter van Musschenbroek, omitted biological and geological sciences, and almost all chemistry and meteorology from their popular, widely-read textbooks. Thus, concluded Heilbron, investigators used the demonstration-lecture to spread Newton's ideas and narrow the domain of natural philosophy. Therefore, by mid-century, "physics" (or fisica, physique, physica, Naturlehre) omitted most geological, biological, and chemical topics. For many, "physics" and "natural philosophy" had become synonymous.

Although the Cartesian natural philosophical tradition survived, it drew increasing criticism from Newton and his disciples for its "system building" and for its use of unwarranted hypotheses. In the Principia, Newton clearly stated that general propositions were to be gathered by induction from the phenomena, not hypotheses. In 1709, curator of experiments at the Royal Society, Francis Hauksbee concurred:

Similarly, 'sGravesande's Newtonian textbook, Mathematical Elements of Physicks, (translated into English by John Keill) noted eleven years later:

Alluding to Descartes' use of hypotheses, 'sGravesande reiterated in a later book that Newton's predecessor did "not think the Fiction of Hypotheses was entirely to be rejected out of Natural Philosophy." Many English and Scottish Newtonians including Samuel Clarke, John Desaguliers, David Gregory, and John Freind agreed that true natural philosophy should reject hypotheses and embrace the fruitful wedding of experiment and mathematics.

The ideals espoused by Newton and his followers often failed in the actual practice of natural philosophy. Despite its narrowing scope and changing methods, natural philosophy remained divided into two domains generally practiced by different groups of people. Historians of science have pointed out the persistence of these divisions, particularly regarding eighteenth-century physical sciences. Cohen, Kuhn, and others argued that natural philosophers tended to emphasize either mathematics or experimentation, but not both, in their work. In doing so, investigators stressed the methods of the Principia or the Opticks respectively. At mid-century, mathematical subjects such as mechanics, hydrostatics, and planetary astronomy constituted the "physico-mathematical" sciences. These areas, dominated by mathematicians, concentrated on taking a single, simple generalization taken from experience and generalizing it mathematically. Subjects of lower status, dominated by experimenters, contained little, if any, mathematics. As such, the experimental branch of physics included the qualitative study of physical optics, heat, electricity, and magnetism. While eighteenth-century experimentalists frequently alluded to the desirability of quantitative data, they rarely went beyond simple numerical tables. They did not integrate numbers or quantities into their qualitative theories. In fact, until the late eighteenth century, few natural philosophers successfully combined the mathematical and experimental traditions. As we shall see, despite changes in the early nineteenth century, "mathematical" and "experimental" physics remained more or less distinct.

In tracing developments of eighteenth-century natural philosophy, John Heilbron and others have questioned historiographies that take the triumph of Newtonianism as the primary guiding principle. They have done so for several reasons. First, using the label "Newtonian" requires broadening its meaning to such an extent that it is practically useless. Although this does not imply that Newtonians did not exist, it does mean that the term "Newtonian" must be used in a qualified manner.

Not surprisingly, Newton's legacy inspired many individuals with differing educations, philosophies, goals, and methods. Some investigators, particularly French mathematicians, utilized the emerging calculus to extend the highly geometrical approach presented in the Principia. From their efforts emerged the rational mechanics of Jean d'Alembert and the celestial mechanics of Pierre-Louis Maupertuis, Alexis Clairaut and others. In contrast, experimental philosophers took their inspiration from the Opticks, virtually ignoring mathematics in their research. In the natural philosophical tradition, many investigators attempted to reduce phenomena to an all ecompassing system of particles and short-range forces of attraction and repulsion. In a similar manner, others took an interest in an active ethereal medium as the common cause of light, gravity, electricity, heat, and other phenomena. Both mathematical and experimental approaches gained inspiration from Newton's own speculations. Also claiming to follow Newton, still others eschewed all hypotheses, contending that true causes remained unknown or unknowable. All of these approaches can be designated "Newtonian," but such a label must be carefully placed in context.

A second major criticism launched against the eighteenth-century triumph of Newtonianism is that Newtonians, even when carefully defined, were not immune to non-Newtonian influences. Other philosophical traditions, originating on the continent, mixed and combined with Newtonianism. Elements of Cartesianism flourished as did the views of Leibnizians, Stahlians, and others. Illustrating the difficulties, various historians have called Leonhard Euler a Newtonian, a Cartesian, and a Leibnizian. Daniel Bernoulli has been classified confusingly as an advocate of "Cartesian Leibnizean Newtonianism." Euler, Pierre-Louis Maupertuis, and Johann Bernoulli accepted Newton's laws of mechanics and universal gravitation, yet rejected Newtonian optics. In developing mechanics as a branch of mathematics, Jean d'Alembert drew upon both Newtonian and Cartesian traditions. "Pure Newtonianism", whatever it might have been, rarely if ever existed. Newtonianism usually became infused with original ideas as well. In this regard, major figures such as Euler, d'Alembert, and John Dalton are especially difficult to label "Newtonian" in a meaningful manner.

A third critique of the triumph of Newtonianism is that too often historians make a general theory, a world-view, or a methodological principle the driving force for scientific change. Referring specifically to experimental physics, Heilbron argued that foundational or methodological concerns may often be too remote from actual experimental work to order it in useful ways. For instance, electrical experiments and instruments early in the eighteenth century cannot be meaningfully distinguished as "Newtonian" or "Cartesian." The same problem, noted Heilbron, extends to models and hypotheses explaining the phenomena:

In like manner, Geoffrey Sutton argued that "explanations offered by French Cartesians and British Newtonians seem essentially interchangeable . . . [finding] a pair of paradigms that differentiates the two is a quixotic task at best." In a related problem, Home remarked that applying the terms "Newtonian" or "Cartesian" to late eighteenth-century figures implies their preoccupation with the same concerns of earlier investigators. Taking this to be a faulty assumption, he concluded:

With these caveats in mind, historians of science have offered alternatives to the triumph of Newtonianism. Home, for instance, proposed to examine the actual practice of the scientists rather than stressing their prefatory methodological statements. If this is done, the mathematical, experimental, and natural philosophical traditions emerge, frequently cutting across the barriers of English and Continental intellectual schools. The history of eighteenth-century physics becomes a complex story of different developing traditions rather than the simplistic triumph of a Newtonianism.

Similar to Home, Heilbron suggested a close examination of the shifting scope, methods, and definitions of physics during the eighteenth century. To illustrate these changes in a particular case, he divided the study of electricity into three periods. From 1700 to 1740, electricity became a distinct sub-species of experimental physics stemming from the work of Stephen Gray, Charles Dufay, and others. Experimenters associated with leading scientific academies performed most of the work, establishing the basic phenomena of electrostatics. In Heilbron's second period from 1740 to 1760, information increased and qualitative theories emerged with little or no mathematical content. French experimentalist Jean-Antoine Nollet, for instance, supposed the simultaneous influx and efflux of electric matter in his theory. In part due to Nollet's exciting demonstrations, electricity became the leading branch of experimental physics, commanding lengthy sections in natural philosophy treatises. His theory, however, was eventually challenged by another qualitative theory developed by Benjamin Franklin. In the 1750s, Franklin explained that electric phenomena arose due to a subtle fluid which attracted ordinary matter, yet repulsed its own particles. Using the analogy of sponge holding water, he explained in 1753:

As the sponge was to water, common matter was to the electric fluid. Hence, when electric fluid increased beyond a body's "natural quantity", the fluid spread across its surface, forming an electric "atmosphere." Franklinian theory utilized electric atmospheres and the overabundance or deficiency of fluid to explain most electric phenomena qualitatively.

As an Enlightenment climate of polite learning stimulated a popular curiosity in science, experimentalists like Nollet and Franklin were not the only ones interested in electricity. This broader curiosity in science, generally, and electricity, specifically, ranged from common men (and women) to university professors. Although excitement and play were no doubt important motives, growing interest in electricity was not merely frivolous entertainment; electricity appeared to cause earthquakes and thunderbolts and to cure paralysis as well.

In Heilbron's final period, from 1760-1790, qualitative theories and experimentation began to be replaced by mathematical formulation and precise measurement. Professors and academicians with mathematical training, including Franz Aepinus and Charles-Augustin Coulomb, dominated the subject. As well, new instruments appeared, as did textbooks and specific monographs on electricity. By the 1790s, increasing numbers of natural philosophers attributed electricity as well as heat, light, and magnetism to the actions of distinct imponderable fluids. As French experimental physics became increasingly mathematical and quantitative, British investigators adopted and adapted French mathematics and physics. Although the divisions between experimental and mathematical traditions persisted, most physicists at the beginning of the nineteenth century attempted to subject all phenomena to careful measurement and experiment.

Despite the gradual transformation of experimental physics as a whole in the late eighteenth and early nineteenth centuries, most scholarly attention has focused on particular developments in the areas of electric, optical, and thermal phenomena. Electricity by itself, Heilbron noted in 1980, garnered over forty percent of the historical literature, as much as meteorology, optics, thermodynamics, pneumatics, and magnetism taken together. By Heilbron's calculation, electricity, light, and heat accounted for sixty-five percent of the coverage of eighteenth-century experimental physics between 1927-1965 and fifty-nine percent between 1966-1977. In contrast, magnetism received only two and three percent respectively for these periods. Books like Thomas Hankins' Science and the Enlightenment (1985) give scant coverage to the study of magnetism.

Nevertheless, the study of magnetism, like the study of electricity, light, and heat, underwent important transformations between 1750 and 1830. Neither these changes nor their connections to contemporary theories of terrestrial magnetism, however, have been examined in the existing scholarship. Although historians of science have touched upon the development of the magnetic force law and imponderable fluid theories, they have paid little specific attention to study of magnetism, particularly in

the British context before the research of Michael Faraday.

In addition to filling in a gap in the existing scholarship, examining the study of magnetism from the mid-eighteenth century through the 1820s serves several general purposes. First, it clearly illustrates the difficulties of designating certain theories of magnetism "Newtonian." Second, it demonstrates in very general terms the shifting roles of the experimental, natural philosophical, and mathematical traditions. Third, it shows the links between the understanding of magnetism and the understanding of terrestrial magnetism. Fourth, it traces the continuing importance of continental influences on the development of British experimental physics. Fifth and finally, it examines broader changes in methodology and how these influenced the study of magnetism during the period. With these goals in mind, a closer examination of the context and content of magnetic studies before 1750 will set the stage for later chapters.

From its earliest days, European interest in magnetism stressed navigational applications. Although known much earlier in China, the magnetic compass first appeared in Europe sometime during the twelfth century. Over the next several centuries investigators began to recognize certain irregularities in the motions of the compass. Barring few exceptions, magnetism usually gained the attention of navigators, instrument makers and practitioners of mixed mathematics rather than natural philosophers. During the fifteenth century, Christopher Columbus, Sebastian Cabot, and other explorers noticed that the magnetic needle rarely pointed to the true geographic north, an observation called magnetic variation or declination. In the sixteenth century, investigators found that the needle tilted vertically with respect to a horizontal plane. The northern end of the needle, for instance, tilted or "dipped" in the northern hemisphere. This magnetic "dip" or inclination increased in higher latitudes and diminished in equatorial regions. In 1581, Robert Norman, a London instrument maker with wide acquaintance among ships' captains published The New Attractive. Norman's book gave the first lengthy treatment of magnetic dip and described a special instrument, the dipping needle, for measuring it. Norman, however, made no attempt to explain the cause of dip.

In contrast to Norman's book, William Gilbert's De Magnete (1600), with a preface by London navigational authority Edward Wright, sought to explain all magnetic and terrestrial magnetic phenomena. Gilbert, an English physician, unrelentingly attacked the methods and contents of most earlier magnetic studies. Prior to De Magnete most writers had favored celestial or localized terrestrial origins of global magnetic phenomena. Many argued that the compass needle indicated a guiding point in the heavens, usually the Pole Star. Others proposed huge magnetic islands or rocks far to the north which guided the motions of the compass needle; tales were even told of these rocks pulling nails from passing ships and sinking them. While approving of the observations of his compatriots such as Edward Wright, Robert Norman, William Barlowe, and others connected with navigation, Gilbert dismissed most earlier theories as "figments and ravings." Such theories had been founded on reckless speculation rather than careful observation.

Reflected in the full title, "A new philosophy of the magnet, magnetic bodies and the great magnet of the Earth," De Magnete shifted the emphasis from older notions to the magnetism of the entire earth. Railing against arguments founded solely upon speculation, authority, and books rather than reason, observation, and experiment, Gilbert exclaimed:

Republished in England in 1628 and again in 1633, Gilbert's work was well-received through most of Europe for its experimental method and its magnetic discoveries. Over the next several centuries, references to Gilbert frequently cited him as the founder of experimental and magnetical philosophy.

After dismissing the "fables and follies" of earlier investigators, Gilbert distinguished between the causes of magnetism and electricity. Proposing that electrical attractions exhibited by a piece of rubbed amber arose from the material emission of effluvia, he claimed that magnetic attractions did not share the same cause. Magnetic emanations, unlike material electric effluvia, could penetrate the densest bodies and magnetize needles without adding to their weight. Therefore, Gilbert concluded, "Electrical bodies [attract] by means of natural effluvia from humour; magnetic bodies by formal efficiencies or rather by primary native strength (vigor)." Rejecting Aristotelian definitions of formal cause, he further explained, "This form is unique and peculiar: it is not what the Peripatetics call causa formalis," nor was it the specific cause alchemists associated with mixtures or the propagator of generative bodies. Asserting God had implanted this form, he remarked:

Identifying the magnetic form with an immaterial soul or earthly anima, Gilbert supposed each magnetic body, including the earth, to be surrounded by an orb of magnetic virtue (orbis virtutis) extending a certain distance in all directions. The extent of the orb depended on the purity of the magnet. Lodestone, iron and other magnetics within the surrounding orb became attracted to the body [see Figure 1].

Continuing his argument, Gilbert supposed that a suspended terrella perfectly represented the earth’s magnetism as manifested by five magnetic properties: coition, verticity, declination, dip, and rotation. First, coition, the mutual attraction between magnetic bodies, occurred within the orb of virtue. Second, verticity was a magnet’s ability to align itself in a fixed direction. Gilbert used these two properties to explain the earth’s natural ability to turn on its axis and the stability of the axis. As evidence, he presented a multitude of experiments using small mounted compass needles or versoria moved around a spherical lodestones or terrellae. Analogous to mariners’ observations with compass needles, his experiments with the terrellae indicated that the earth itself was a giant magnet. Therefore, magnetic substances within the terrestrial orb of magnetic virtue behaved analogously to substances within the orbs of ordinary, smaller magnets.

With versoria and terrellae, Gilbert duplicated all the phenomena known to navigators and mathematical practitioners including the third and fourth magnetic properties— declination and dip (which he called "variation" and "declination" respectively). Like the terrestrial globe, Gilbert’s terrella had two poles and an equator. Further linking the earth and lodestone, natural magnets had been discovered throughout the earth’s surface. Gilbert also suggested that declination and dip might be mapped out for determining longitude and latitude.

Gilbert's cosmological arguments connected terrestrial magnetism with the rotation of the earth on its axis. Assuming the coincidence of the magnetic and the rotational poles, he explained:

Magnetic declination, however, indicated that the magnetic and geographic poles did not coincide. Bypassing this difficulty, Gilbert assumed that superficial irregularities of the terrestrial surface were the source of declination. Illustrating by analogy, he constructed a deformed terrella with indentations and raised portions representing the irregular distributions of sea and land. Appealing again to the terrella-earth analogy, Gilbert noted that the dip increased to a maximum at the earthly poles. Regardless of superficial irregularities, the magnetic poles and the geographic poles coincided.

In the final book of De Magnete, Gilbert discussed the fifth magnetic property, rotation. In this book, he departed from the terrella-earth analogy by rejecting earlier speculations that a perfectly-aligned, spherical lodestone would rotate on its axis, "I omit as Petrus Peregrinus so stoutly affirms, that a terrella poised on its poles in the meridian moves circularly with a complete revolution in twenty-four hours. We have never chanced to see this: nay, we doubt if there is such movement." Lacking an analogous rotation of the terrella, he nevertheless asserted the rotation of the earth:

As we shall see, Gilbert’s experimental conclusions regarding the earth’s magnetism as well as his cosmological speculations regarding the earth’s daily rotation were embraced, modified, and criticized by numerous seventeenth-century investigators.

Early in the seventeenth century, Gilbert's "magnetical philosophy" became a popular topic, particularly among Copernicans. Proponents including Simon Stevin, Johannes Kepler, and Galileo Galilei, borrowed and extended his arguments. Kepler, for instance, extended Gilbert's system by proposing magnetic forces emanating from the sun in the plane of planetary orbits. He explained the "body of the sun is circularly magnetic" and forms a circular "magnetic river" of immaterial emanations. Later Kepler supposed that "every planetary body must be regarded as magnetic, or quasi-magnetic; in fact, I suggest a similarity, and do not assert an identity." Hence, each planet had two quasi-magnetic poles, one "friendly to the sun" and another "hostile." Kepler believed that the cause of elliptical orbits lay in these magnetic properties of the sun and planets. In Astronomia Nova (1609) he explained, "the librational force is brought about by a magnetic force which is indeed innate and solitary, without any operation of a mind, but its description depends on the external solar body. The force, in fact, is defined as sun-seeking or as sun-fleeing." Illustrating the importance of Gilbert, Kepler claimed in Epitome of Copernican Astronomy (1620) to have built his entire astronomy on Copernicus' system of the world, Tycho Brahe's observations, and "the Magnetical Philosophy of the Englishman William Gilbert."

Though rejecting Kepler's elliptical orbits, Galileo accepted some of Gilbert's magnetic arguments in a Dialogue Concerning the Two Chief World Systems (1632). Toward the conclusion of the third day in the Dialogue, he wrote:

Salviati and Sagredo then demonstrated to Simplicio, the Aristotelian character, that true elemental earth or lodestone, in contradiction to Aristotelian physics, had a circular motion. Furthermore, Salviati, Galileo's spokesman, endorsed Gilbert's argument for the magnetic stabilization of the Earth's axis. Later, Salviati noted: "That which I could have desired in Gilbert is that he had been a somewhat better mathematician and particularly well grounded in geometry." In this manner, Galileo, working in the mathematical tradition, criticized Gilbert, working in the experimental tradition.

Also in the experimental tradition (if in word, not deed), Francis Bacon, though not a Copernican, praised Gilbert's "many exquisite experiments" and lauded the compass as one of the most important discoveries of modern times. Despite his high praises, Bacon consistently condemned the magnetic philosophy. In Novum Organum (1620), he rejected the philosophy of Gilbert as well as the dogmas of alchemists. Constructing an entire system upon his favorite subject, Gilbert had, in Bacon's opinion, "become a magnet; that is, he has ascribed too many things to that force, and built a ship out of a shell." In contrast to Gilbert, Bacon rejected the diurnal rotation of the earth and supposed a material cause for magnetism. Though frequently mentioning Gilbert, Bacon exhibited no deep knowledge of his work as did several of his fellow countrymen including Mark Ridley, William Barlowe, and Henry Gellibrand, all of whom worked on magnetism within the experimental tradition.

In 1635, Gellibrand, professor of astronomy at Gresham College and an self- admitted Copernican, recognized that the declination in London had changed from past measurements. This slow, gradual change, which he called the "variation of the variation" gave impetus to additional magnetic observations. Of this secular variation Gellibrand remarked:

Since Gilbert's theory did not allow compass needles to change direction with the passage of time, Gellibrand's "variation of the variation" clearly conflicted with the Gilbert’s theory. The discovery of secular variation seemed to require that terrestrial magnetic poles become distinct from the geographical poles, thereby it cast doubts on the cosmological conclusions of the magnetical philosophy. Among other difficulties, secular variation did not mesh with Gilbert's explanation for diurnal rotation of the earth. As such, these factors contributed to the demise of Gilbertian cosmological arguments.

In the meantime, British mercantilism, overseas colonies, and trade wars continued, providing economic stimuli to the navigational application of magnetic studies. In their efforts, Gellibrand and other Gresham professors closely cooperated with members of the naval community such as John Wells, the Keeper of His Majesty's Naval Stores at Deptford. The national importance of navigation led many to propose magnetic solutions for determining latitude and longitude. For instance, Henry Bond, a navigation teacher, put forth a scheme in 1648 which eventually appeared in The Longitude Found (1676). Also with navigation in mind, the Royal Society of London established a Magnetics Committee in 1664 for studying and measuring magnetic variation. When the committee discovered instrumental errors too large to confirm Bond's variation prediction, they assumed that unknown variables had altered the directive property of magnetic needles. President of the Royal Society, Sir Robert Moray, conjectured that perhaps different lodestones induced different directions or that the mechanical process of magnetization affected the needle’s magnetism. During the second half of the seventeenth century, opinions akin to Moray's illustrated the growing acceptance of magnetism's mechanical origins and a concomitant rejection of Gilbert's immaterial magnetic souls.

Magnetism and the Mechanical Philosophy

Although many continued using Gilbertian terms like "magnetic virtue", "magnetical vigor", and "intrinsic energy" to describe magnetic effects, numerous Englishmen after 1650 reduced natural phenomena, including magnetism, to matter and motion. According to this mechanical-atomical philosophy, all phenomena were explicable in terms of the size, shape, number, and motion of particles of matter. As one of the leading advocates of mechanical philosophy, René Descartes sought to rescue magnetism from occult or animistic explanations such as Gilbert's. Though he accepted that the earth contained or behaved as a giant magnet, Descartes vigorously rejected Gilbert's incorporeal emanations and orbs of magnetic virtue. The Cartesian alternative utilized tiny screw-like particles which circulated through and around all magnets, including the earth, forming magnetic vortices. When iron filings were sprinkled onto a piece of paper placed over a magnet, their patterns indicated the underlying flow of these invisible particles. The minute particles of Descartes' first element which made up these vortices became grooved as they squeezed through gaps in clusters of his larger, spherical second element. As the particles emerged, they naturally rotated and twisted, resembling either right-handed or left-handed headless cylindrical screws. Cartesian theory posited the two-way circulation of effluvia through appropriately threaded pores of iron. To explain all magnetic phenomena, Descartes depended exclusively on the motion of these threaded particles of matter. For its adherents, the theory persuasively described why magnets acted continuously without the loss of power or weight.

Explaining terrestrial magnetism, Descartes' theory supposed one type of particles entered at the north pole, traveled through the terrestrial interior, and returned via the earth's atmosphere to its point of origin. Similarly, the oppositely-threaded particles entered at the south pole and circulated in the opposite direction [see Figure 2]. To accommodate the passage of particles, grooved or threaded channels ran along the entirety of the earth's interior. Magnetic needles aligned with the effluvial flow of particles circulating around and through the earth, giving them a general north-south direction. Magnetic attractions occurred when opposite poles faced each other and the particles circulated as if around a single magnet, eventually closing the gap between them. Similarly, Magnetic repulsions occurred when the threaded particles could not enter the channels made for them, thereby pushing objects apart.

Though the specific details of Descartes' magnetic theory were often rejected, the general concept of circulating effluvia became popular for several reasons. First, it plausibly explained why metals extracted from the earth became easily magnetized; these metals, it was supposed, simply had the correct kind of grooves. Second, it explained the irregularity of magnetic declination as originating from uneven deposits of iron and lodestone, thereby disturbing the symmetry of effluvial flow. Third, the "variation of variation" arose due to mankind's mining activities and numerous natural processes which altered the global distribution of ferrous materials. Finally, and most importantly, Cartesian theory appealed to an intellectual climate in which animistic forces and Aristotelian "substantial forms" were becoming anathema to many.

From 1650 onward, many natural philosophers sought to elaborate what Robert Hooke called "the real, the mechanical, the experimental Philosophy." Mechanical philosophers, whether in agreement with Descartes or not, believed that all phenomena could be explained in terms of matter and motion. Referring to the electrical hypotheses of Gassendi, Descartes and others, Robert Boyle explained that each had attempted "to solve the phaenomena in a mechanical way, without recurring to substantial forms, and inexplicable qualities." Electricity, he noted, resulted from a material effluvium issuing from electrified bodies. In this and many other instances, British mechanical philosophers rejected Aristotelian explanations. Utilizing elements from both Cartesianism and revived Epicurean atomism, Boyle explained in 1660:

Thereby, Boyle, among other mechanical philosophers, sought to explain magnetism by the shape, size, and motion of particles.

By the second half of the seventeenth century, mechanical explanations for magnetism had become fairly commonplace. French atomist Pierre Gassendi, for instance, proposed the continuous emission of hooked particles from the lodestone which become anchored in iron, thereby pulling the two together. While Gassendi, Christiaan Huygens, and others on the continent advocated mechanical theories of magnetism, Henry More, John Wilkins, Walter Charleton, and Thomas Hobbes supported similar mechanical explanations in England. Criticizing those who embraced Gilbert's argument for terrestrial rotation, Hobbes noted:

Furthermore, he explained that the lodestone's attractive power arose from "nothing else but some motion of the smallest particles thereof."

Though wary of Aristotelian "forms" and "qualities", many English mechanical philosophers did not entirely reject action-at-a-distance. Astronomer John Wilkins, for example, kept alive the analogy between magnetic forces and those which governed celestial bodies. An uneasy coexistence persisted between the aspects of the magnetical philosophy and aspects of the mechanical philosophy. As well, different elements of these traditions sometimes intermingled. For instance, building upon Henry Power's Cartesian magnetic research, Boyle upheld the tenets of mechanical description in a qualified manner. In 1675, Boyle divided the general properties of bodies into several categories. Although his fourth category, "occult qualities," included electricity and magnetism, Boyle's Experiments and Notes about the Mechanical Production of Magnetism, published the following year, explained:

Hence, while generally favoring mechanical explanations, Boyle did not wholeheartedly embrace them when it came to magnetism.

While Boyle did not explain how corporeal effluvia produced magnetic effects, he did briefly describe the mechanical generation of several specific phenomena. Repeating some of Gilbert's experiments, Boyle argued that red-hot iron bar acquired magnetism when cooled in a north-south direction because it became "pervaded by the magnetical effluvia of the earth, which glide perpetually through the air from one pole to another, and by the passage of these steams [sic] it becomes endowed with a magnetical property, which some call polarity." Admittedly, Boyle wrote, it might seem strange to attribute to "so gross and dull a body as the earth" the invisible power to communicate magnetism; in fact, he concluded, that we probably would not have dreamed of this "if our inquisitive Gilbert had not happily found out the magnetism of the terrestrial globe." Hence, Boyle retained elements of both Gilbert and Descartes in his writings.

Robert Hooke, Curator of Experiments for the Royal Society, also gave equivocal support to effluvial explanations. In the early 1670s, Hooke remarked of magnetical effluvia bending or inflecting themselves in different directions. Magnetic power, he argued, came from the motions of "an Aethereal subtil Matter, which penetrates and pervades, and fills the Interstices of all Terrestrial Bodies." While espousing the mechanical philosophy, Hooke simultaneously appealed to attractive forces. In Lectures and Collections (1678), he noted, without qualification, that the "attractive power" between the planets and the sun, was "as the Load-stone hath to Iron, and the Iron hath to the Load-stone." In the very same work, however, he supposed, "all things in the Universe that become the objects of our senses are compounded of these two . . . namely, Body, and Motion." Several years later, Hooke illustrated his growing doubts regarding the powers of magnetism and gravity:

As well, speculation and confusion continued regarding the mechanism of secular variation of terrestrial magnetism. In 1670, Boyle supposed that unknown internal changes were the cause, while Hooke suggested several years later that a magnetic pole rotated around the geographic pole once every 370 years. In the 1680s and 1690s, astronomer Edmond Halley proposed the existence of four magnetic poles to explain the slow westward drift of variation. In 1683, Halley's first hypothesis conjectured the motion of magnetic poles, two in the northern hemisphere and two in the southern, on the terrestrial surface. Nine years later, he proposed a mechanism for the motion of the poles— an internal magnetic nucleus with two poles surrounded by a magnetic shell also with two poles. Thus, secular variation resulted from a slight difference in rotational periods for the inner kernel and the outer shell. Despite these clever attempts to reduce secular variation to a physical mechanism, the myriad variables, complexities of effluvial magnetism, and scarcity of reliable terrestrial magnetic data tested the confidence of Hooke, Halley, and other investigators as well. In 1683, Halley listed a multitude of unknowns:

He warned that investigators should be wary of accepting any hypothesis (including his own), no matter how plausible it might seem.

Continuing the connections developed between Gresham professors and the Navy, the Admiralty in 1698 instructed Halley upon the wishes of King William III to "improve the knowledge of the Longitude and Variations of the Compasse" and to find Terra Incognita (the southern land mass whose existence had been postulated by ancient geographers). After three Atlantic voyages (1698-1700), Halley failed to locate the great southern continent, yet he collected several hundred magnetic observations. These were used for a magnetic chart of the Atlantic (1700) showing lines of equal magnetic variation (later called isogonic lines). This chart and an extended world chart published in 1702 were intended to help solve the longitude problem. Of his magnetic chart and its uses, Halley wrote:

As we shall see in the next chapter, few were willing or able to make the necessary observations to periodically update Halley's magnetic charts.

Late in the seventeenth century the complexities of magnetic phenomena and its terrestrial manifestations led to the collapse of Gilbert's magnetic philosophy. Separate and distinct from Newton's force of gravity, magnetic phenomena lost their cosmological significance. Further hastening the decline of the magnetic philosophy, mechanical philosophers supposed that compass needles were subject to mechanical effects of heat, cold, and hammering, as well as irregular atmospheric and geological disturbances. Fraught with complexity and uncertainty, the study of magnetism became subsumed to the general study of mechanical effluvia. Although the compass continued to guide mariners through rough seas, terrestrial magnetic irregularities and a lack of reliable data increasingly ruled out the application of magnetic measurements for determining longitude.

In the early eighteenth century the study of magnetism and its application to the longitude problem garnered less attention then previously. Despite continued, even frenzied, interest in the longitude problem (particularly fostered by the Longitude Act of 1714), eighteenth-century investigators, including Halley, gave less attention to magnetic solutions. Magnetic longitude schemes persisted, but no longer with the frequency of those put forth in the seventeenth century. Other solutions developed during the eighteenth century used astronomical observations or marine chronometers. These eventually proved more convenient and more accurate than magnetic charts. Until the nineteenth century, very few showed enthusiasm for collecting magnetic observations on a global scale. By then, the motivations for collecting magnetic data had changed.

Experimental research early in the century further complicated matters. Like the efforts of Hooke and Halley, eighteenth-century attempts to determine a magnetic force law led to uncertainty, even frustration. In the second edition of the Principia (1713) Newton noted that the power of the magnet diminished "not as the square but almost as the cube of the distance," yet neglected to explain how he had come to this conclusion. Around the same time, Francis Hauksbee and Brook Taylor saw fit to confine their experimental results to tables of raw data. In the 1720s, Taylor complained of the difficulties of locating the centers of magnetic power in his magnets and needles. He concluded that magnetism's power did not change according to any particular power of the distance. Similarly, the Dutch experimentalist Pieter van Musschenbroek contended that there no constant law related magnetic force to distance. Illustrative of the confusion and complexity, proposals for the law of magnetic force also included simple inverse, inverse to the 3/2 power, inverse square, inverse to the 5/2 power, and inverse to the fourth power.

Complexity continued to confound terrestrial magnetic measurements as well. In 1724, London clockmaker George Graham noticed with a specially-made compass that magnetic variation fluctuated over the course of the day. This diurnal variation of the magnetic needle and other irregular fluctuations merely added to the difficulties of explaining terrestrial magnetic phenomena. In the 1740s and 1750s, John Canton performed many experiments on diurnal variation. First, he placed a small magnet near a compass and noted the needle's deflections. Next, he covered the magnet with a brass container and poured hot water into the container. Heating the magnet, Canton noted, caused the nearby needle to fluctuate. Like Gilbert, Canton extended his observations to the entire earth so that, by analogy, he postulated that the sun heated different terrestrial regions during the day, thereby weakening the magnetic forces and deflecting the needle. Using this analogy, Canton attempted to explain the small diurnal variations of the magnetic needle. These tiny, yet noticeable variations and other irregular fluctuations in magnetic measurements gained much greater attention in the nineteenth century.

Adding to early eighteenth-century confusion, Isaac Newton never gave a straightforward discussion of his theoretical views on the causes of magnetism. Without putting forth a detailed theory, he espoused conflicting, even contradictory messages on the subject. Several times, Newton compared magnetism to gravity and other action-at-a-distance forces whose causes remained unknown. In the 31st Query to the Opticks (1730) he asked:

Nonetheless, Newton conjectured immediately following these remarks that what he called attraction "may be perform'd by impulse, or by some other means" unknown to him. In fact, most evidence (published and unpublished) indicates that he, like many others, supported a Cartesian description of magnetic phenomena. For most of his career, Newton espoused circulating magnetic "streams" or "effluvia."

The situation in the early eighteenth century aptly illustrates the difficulties of defining a "Newtonian" in a straightforward manner. With respect to magnetism early followers of Newton might take one of several paths including: accepting the circulating effluvial theory; claiming that magnetism's causes remained unknown; or endorsing an action-at-a-distance force like gravity. Newtonians Edmond Halley, George Graham, and Colin Maclaurin accepted effluvial theories, yet William Derham, Brook Taylor, and John Desaguliers, also dedicated Newtonians, avoided writing about the underlying causes of magnetism. Rather than speculate about the underlying physical mechanisms, the latter investigators used terms such as "power", "attractive virtue", "polarity", and "magnetic virtue." Desaguliers, a popularizer of experimental natural philosophy, wrote in 1730:

Other Newtonians such as William Whiston, John Keill, and Musschenbroek openly disapproved of corpuscular-mechanical theories. Keill, an opponent of the mechanical philosophy, commented in his natural philosophy lectures at Oxford in 1700 that:

Appealing to experimental evidence, Musschenbroek argued into the 1740s that magnets did not act by circulating material effluvia. As these and other examples illustrate, the label "Newtonian," even if chosen by the investigators themselves, did not presuppose one specific view with respect to magnetism. Hence, Newtonians explained magnetic phenomena in several ways and remained consistent with what Newton himself espoused.

Despite a lack of consensus among Newtonians, the Cartesian notion of a subtle magnetic effluvia circulating around and through magnets retained widespread acceptance in England and on the Continent. In 1696, German mathematician and philosopher Gottfried Wilhelm Leibniz, wrote of Descartes' theory:

Despite Leibniz' objections, he and many others retained the basic elements of the circulating fluid theory. John Harris' Lexicon Technicum (1704) supported such a theory and the Gilbertian analogy as well. From all experiments, Harris concluded, "'tis plain (as Mr. Boyle concludes) That Magnetism doth much depend upon Mechanical Principles. As also, That there is such a Thing as the Magnetism of the Earth; or that there are Magnetical Particles, which continually are passing from Pole to Pole." In an English translation of Cartesian Jacques Rohault's A System of Natural Philosophy (1723), Newtonian Samuel Clarke, who annotated much of the text with objections, did not object to Rohault's presentation of the Cartesian magnetic theory. Five years later, Ephraim Chambers' popular Cyclopaedia (1728) commented, "The opinion that principally prevails among the moderns [on magnetism] is that of Des Cartes." The fifth (1741-43) and seventh (1752) editions of Chambers' Cyclopaedia reiterated this claim. At mid-century, Britain's leading magnetic researcher Gowin Knight clearly embraced the effluvial theory; he wrote, "The magnetic Matter of a Loadstone moves in a Stream from one Pole to the other internally, and is then carried back in curve[d] Lines externally, till it arrives again at the Pole where it first entered, to be again admitted." Though the theory was not Cartesian in the strict sense of following the details of Descartes' theory, it nonetheless retained the notion of a mechanical circulating effluvia traversing around and through all magnetic bodies. Variants of this general idea, for sake of convenience, be designated "Cartesian" throughout the remaining chapters.

In 1746, Leonhard Euler, Daniel and Jean Bernoulli (II), and Etienne-François DuTour split the prize offered by the Paris Academy of Sciences for the best paper on theory of magnetism. Though rejecting grooved particles and rifled channels, each paper proposed a qualitative, non-mathematical explanation appealing to a circulating subtle magnetic matter. In the Cartesian tradition, able mathematicians such as Euler and the Bernoullis accepted that they were dealing with an irreducible, erratic phenomena not amenable to mathematical analysis. Hence, the study of magnetism remained firmly within the experimental tradition. As we shall see, the dominance of "Cartesian" effluvial theories continued for most of the eighteenth century, even in "Newtonian" Britain.

How does the British study of magnetism fit within the mathematical, experimental, and natural philosophical traditions? Before Gilbert's magnetical philosophy, navigators and instrument makers dominated the study of magnetism. Magnetic study was considered a part of navigational science under the domain of mixed mathematics. As navigators and instrument makers continued to study magnetism in the seventeenth century, De Magnete assisted in creating a new "Baconian science," which was soon taken up by mathematicians and some natural philosophers. Despite the involvement of mathematicians and magnetism's classification as part of mixed mathematics, the actual study of magnetism remained devoid of mathematical content. Efforts to quantify were limited to tables of numbers, either derived from experiments or from observations of magnetic dip or variation. As such, these studies remained squarely within the experimental tradition.

The rise of the mechanical philosophy in the mid-seventeenth century, however, placed the study of magnetism within the natural philosophical tradition as well. Descartes' screw-shaped particles, Gassendi's hook-shaped particles, and Boyle's magnetic corpuscles, although appealing to experimental evidence such as the patterns of iron filings, did not arise solely within the experimental tradition. Instead, magnetic effluvia arose from an a priori metaphysical assumption that all phenomena could be explained mechanically, in terms of matter and motion. Making simplistic mechanical interpretations of magnetism more complex, Boyle, Hooke, and Newton supposed action-at-a-distance forces while simultaneously espousing mechanical descriptions. Adding to the magnetic perplexities were numerous problems including the complete lack of a magnetic force law, the unpredictable effects of magnets being hammered, heated and cooled, and the unreliable, insufficient terrestrial magnetic data which grew ever more intransigent with the accumulation of measurements.

By 1750, circulating effluvia dominated theories of magnetism and terrestrial magnetism alike. Despite this dominance no consensus existed among British investigators during the eighteenth century. The study of magnetism, divided between the experimental and natural philosophical traditions, continued to be a mystery with tremendous philosophical and navigational potential. Nonetheless, magnetism was less studied than other areas of experimental physics, particularly electricity. It remained a riddle which even the incomparable Sir Isaac Newton had not satisfactorily reduced to a mathematical law. Despite numerous experimental attempts, the magnetic force law, if it existed at all, remained elusive. British attempts to resolve the mysterious natures of magnetism and of terrestrial magnetism are the focus of the remaining chapters.

The second chapter examines the impetus behind magnetic data collection beginning circa 1750. Although the close connections between the Royal Society, the Admiralty, and the Board of Longitude persisted, very few natural philosophers continued Halley's program of magnetic mapping. Most often, navigators such as James Cook collected data, while natural philosophers and instrument makers stayed at home to design, construct, and test a variety of magnetic and meteorological instruments. Throughout the voyages of the 1760s and 1770s, magnetic observations took a back seat to the more vital tasks of astronomical collecting and testing chronometers. Following the Napoleonic wars, magnetic collecting in Britain continued with new prominence as Arctic voyages resumed the search for the North-West Passage and the North Pole with renewed vigor. The division of labor continued, as scientific servicemen in the Royal Navy or Royal Artillery collected magnetic data in the frigid polar climate, while natural philosophers and mathematicians performed magnetic experiments at home. Tracing the collection of magnetic data up to 1835 reveals its shifting practical, scientific, and symbolic importance.

Returning to the theoretical scene of mid-eighteenth century, the third chapter closely examines the continuing division of magnetic studies between experimental and natural philosophical traditions. By the 1790s, some investigators slowly merged the mathematical with the experimental tradition. This happened for several reasons, including the acceptance of imponderable fluid theories which successfully quantified and mathematized the study of magnetism and other areas of experimental physics. Earlier in the century, speculations abounded, some even challenged the Gilbertian notion of a giant terrestrial magnet. Others conjectured that an electric fluid or solar rays were responsible for terrestrial magnetic effects. Finally the chapter discusses the influential magnetic and electric theories of German mathematical physicist Franz Aepinus. It examines how Aepinian theory diverged from circulating fluids and how it was initially ignored by British experimenters.

The fourth chapter continues the story of the wedding of experimental and mathematical traditions for the period c. 1780 to 1820 by examining several key transitional figures. The first, Tiberius Cavallo, accepted the one-fluid theory of Aepinus, while remaining in the experimental tradition. As the blending of experimental and mathematical traditions continued, the natural philosophical tradition took an important yet subsidiary role. Using Newton's ether or other basic unifying principles, this speculative tradition continued speaking of nature's intimate connections. The second important transitional figure, John Robison, the professor of natural philosophy at the University of Edinburgh, was one of few British investigators who enthusiastically embraced Aepinian theory and scientific style. The development of Robison's scientific methodology, his influences, and his theory of magnetism are discussed in great detail. By bringing continental mathematics and physics to a wider British audience through encyclopedia articles and other writings, Robison and his successor at Edinburgh, John Playfair, helped alter the face of British experimental physics. In doing so, they brought closer together the use of mathematics and experiment.

As the fifth chapter demonstrates, other important factors in this transformation were the magnetic researches of Charles Augustin Coulomb and Laplacian scientists. This chapter discusses the development of Coulomb's magnetic theory, Laplacian science, and the growing impact of French physics in Britain during the early nineteenth century. Supporters of Aepinus' theory such as Thomas Young increasing questioned and rejected circulating effluvial theories. Despite their long-held popularity Cartesian theories yielded neither to mathematical analysis nor quantification, two desiderata of the emerging style of experimental physics. Though British investigators preferred Aepinian theory during the first decades of the century, the growing influence of Laplacian physics contributed to increasing approval for Coulombian theory by the early 1820s. The chapter suggests that Laplacian physics, particularly that of Jean Baptiste Biot, had certain similarities with the Scottish approach to physics. These similarities contributed to both personal connections and methodological affinities between Laplacian and British experimental physicists.

The final chapter examines how theories of magnetism dramatically changed in Britain from 1820 to 1835. The experimental tradition of magnetic research gained particular impetus from the discovery of electromagnetism in 1820. This discovery initiated a wave of experimental work and fostered further speculation about nature's unity and interconnectedness. The experimental, mathematical, and natural philosophical traditions all played roles in these developments. Meanwhile, continued research and new theories brought about a decline of Laplacian views. Affecting theories of terrestrial magnetism, Ampère, Arago, and others subsumed magnetic effects to the circulation of electric currents in all magnets, including the earth. Other researchers, including Humboldt, Oersted, and Hansteen supposed the effects of electricity, heat, chemical action, or rotation were intimately connected with magnetic and geomagnetic phenomena. This chapter examines various British responses to these experimental discoveries, while emphasizing their impact upon the understanding both magnetism and terrestrial magnetism.

This dissertation adds to the relatively small body of scholarship dealing with the history of magnetism and terrestrial magnetism. In particular, it emphasizes the shifting understanding of magnetism within the context of British experimental physics. Methodological concerns, personal links, changing instrumentation, navigational practices, and other factors such as educational context and nationalism are incorporated into the primary discussion of theoretical changes. This dissertation's main purpose is tracing the shifting theoretical understanding of magnetism and its relationship with changing theories of terrestrial magnetism.