ALCHEMY HAS LONG PRESENTED PROBLEMS for historians of science. To the pioneering historians of the Scientific Revolution who wrote in the generation after World War II, alchemy was a “mystic science,” a “pathology of thought,” and even “the greatest obstacle to the development of rational chemistry.”1 This old habit of reflexive dismissal has largely been abandoned in recent years, and the greater community of early modernists no longer considers alchemy an annoying outlier.2 Yet even as the historiography of alchemy has undergone a sort of renaissance in the last two decades, thanks in large measure to the scholars represented in this Focus section, the topic still manages to generate considerable controversy. Despite the undeniable fact that many of the most creative minds of the Scientific Revolution were seriously interested in the aurific art—Newton, Boyle, Leibniz, Locke, and Spinoza, among others—the subject remains a flash point for dispute. Very recent critics of the new historiography of alchemy have resurrected the chestnut that alchemy was primarily a “narrative of quests for revelation,” reasserted the stereotypical view that alchemical matter theory was indistinguishable from scholastic hylomorphism seen as a single, monolithic doctrine, and insisted on the equally antiquated perspective that alchemists contributed little to the development of modern chemistry.3 Since these claims have been answered elsewhere in detail, it would be superfluous to respond to them here. Yet the very fact that the history of alchemy continues to evoke fundamental scholarly controversy suggests that the field holds much to challenge long-held stereotypes. This essay will therefore focus on a few of the fruits of the last two decades of research, providing some summary points as to what we have learned from studying the history of alchemy. I will proceed from the general to the particular, first showing how a close look at alchemy challenges the view of medieval and early modern scholasticism as being antiexperimental and then passing to the results that alchemists obtained in the realm of matter theory and experiment by means of their integration of philosophical theory with laboratory practice.

ART, NATURE, EXPERIMENT, AND THE NONINTERVENTIONIST FALLACY

One of the most general lessons to be learned from the history of medieval and early modern alchemy pertains to the role of art and nature in premodern Western thought as a whole. From late antiquity onward, alchemists had insisted on their ability to replicate—not just imitate—natural products. As a result of this persistent refrain, alchemy became the focus of a scholastic debate about the powers of art that would last from the Middle Ages until quite modern times. In pushing their claim that they could genuinely reproduce nature rather than making mere ersatz replacements of natural products, scholastic alchemical writers were, surprisingly, developing a side of Aristotelian natural philosophy that has been little studied by modern scholars. In short, they built many of their claims on the principle, enunciated in book 2 of Aristotle's Physics (2.8.199a.15–17), that human art can either mimic nature or it can perfect nature. Developing the concept of alchemy as a “perfective art” rather than one that produced mere imitations, scholastic alchemists eroded the sharp boundary between natural and artificial products espoused by many other premodern thinkers. According to the common alchemical interpretation of Physics 2.8, art could work radical changes on nature without undermining the essences of natural things.4 Already in medieval alchemy, then, one finds the principle famously pronounced by Francis Bacon that “the artificial does not differ from the natural in form or essence, but only in the efficient.” As one medieval alchemist hiding under the name of “Hermes” put it, “the works of man can be both natural with regard to essence (secundum essentiam) and artificial with regard to mode of production (secundum artificium).”5

It is important to stress that this alchemical emphasis on perfective art was not antithetical to Aristotelian natural philosophy but was actually derived from it. Indeed, the founding document, as it were, of much medieval and early modern alchemy was Aristotle's Meteorology, especially books 3 and 4. The third book contains the justification for and possibly the origin of the widespread alchemical belief that metals are composed of two principles (later identified with sulfur and mercury), and the fourth book combines a microstructural approach to matter based on particles and pores with a latitudinarian view of the powers of art. From these Aristotelian starting points, medieval alchemists developed a comprehensive view that their art could arrive at fundamental truths by means of experiment with natural materials. This position, pervasive among scholastic alchemists of the Middle Ages and their early modern heirs, brings the issue of the place of experimentation in Aristotelian natural philosophy into sharp relief. There is a widespread belief in the modern historiography of science and philosophy that intervention in nature, and experiment in particular, was dismissed by scholastic authors because of the view that it interfered with “nature in course.” This historiographical position, which I have elsewhere dubbed the “non-interventionist fallacy,” typically asserts that avoidance of artificial intervention was a necessary consequence of the Aristotelian conception of natural science, on the assumption that any attempt to isolate the object from its normal environment could only be seen as interfering with its nature and producing an artifactual result.6 One of the lessons that we have learned from the recent historiography of alchemy is that Aristotelianism, far from providing a metaphysical prohibition against experiment, actually gave alchemy the tools to justify learning from radical interventions in nature by means of laboratory-based experimentation. Ironically, the seemingly Baconian elision of the boundary between art and nature that one already finds in medieval alchemy stems in large part from the Aristotelian tradition itself.

CORPUSCULAR MATTER THEORY AND THE REDUCTION TO THE PRISTINE STATE

What exactly did alchemists learn from their interventionist manipulation of nature? The myriad and little-studied experimental techniques of medieval and early modern alchemy make this question far too broad for a simple answer. It is possible, however, to point to a particular tradition in alchemy that demonstrates one of the salient features of the art. Already in the thirteenth century one finds that alchemists employed sophisticated analytical techniques in order to determine the nature of matter. This is particularly evident in the Latin corpus attributed to Geber (Jābir ibn Hayyān), which was mostly composed in and after the thirteenth century by European authors. In the Geber corpus, assaying tests are used to determine the relative amounts and qualities of the two principles, sulfur and mercury, in the metals.7 Geber's main work, the Summa perfectionis, subjects the then-known metals to a barrage of tests, including cupellation, cementation, calcination, burning with sulfur, exposure to organic acids, amalgamation with mercury, and so forth, in order to determine the composition of the metals. Historians of technology would point out, correctly, that these techniques were nothing new in themselves—cupellation and cementation, for example, had been known since antiquity. But the self-conscious use of these artisanal techniques as tests for determining the fundamental components of matter was something that lay beyond the purview of common assaying and metallurgy. It was precisely in the integration of philosophical speculation about the nature of matter with the employment of techniques as tests for theories that alchemy distinguished itself from pure technology.

But where did this experimental probing of matter lead? From a modern perspective, of course, the theory of sulfur and mercury was incorrect, even if it did produce the influential phlogiston theory of eighteenth-century chemistry. It is far more revealing, however, to look at alchemical debates about matter in the context of their own time and to see how experimental data was used to combat coeval theories. One prominent example of such a debate may be found in the work of Paul of Taranto (who may also be the real author of the Summa perfectionis), whose Theorica et practica was probably composed near the beginning of the fourteenth century. Paul explicitly argues against the view of Thomas Aquinas and his followers that a given substance can have only one substantial form, which is necessarily destroyed and replaced by another substantial form during its generation or mixture. The Thomistic viewpoint, if upheld strictly, would imply that there can be no principles intermediate between the Aristotelian first matter (prima materia) and the fully formed substance, hence no mercury and sulfur in a fully formed metal. But if that were the case, much of alchemical practice would be rendered ineffectual, since the attempted transmutation of metals at this time depended largely on rectifying their mercury and sulfur and changing their relative quantities. Rather than arguing against the Thomistic position from first principles, Paul employs an early form of an experiment that would later achieve fame as the “reduction to the pristine state.” Using calcination along with traditional “sharp waters” derived from Arabic alchemy, which included organic acids alongside various alkalis, Paul points out that various metals can be incinerated, dissolved, and sublimed, but that the resulting products can always be reduced back into the original metal that one started with. Hence he concludes that the metals are composed of robust intermediate constituents, combined in a “very strong composition,” and that, contrary to the Thomistic belief, the juxtaposed particles can be separated and recombined without losing their being.8

Here we see a type of philosophical engagement with experiment that is unusual indeed in medieval writings. The closest analogue to Paul's approach, it would seem, lies in the area of scholastic optics, where a sort of “maker's knowledge” is sometimes explicitly invoked in comparing the artificial production of rainbows by means of sprays and water-filled flasks to the generation of alchemical gold. Not coincidentally, the optical treatment of the rainbow in medieval texts, like much of alchemy, descended ultimately from Aristotle's Meteorology, where the Stagirite speaks of artificially produced rainbows. The link between optics and alchemy requires further exploration, but one thing is clear. The interaction between philosophical reasoning and experiment evident in the reduction to the pristine state would have striking repercussions in the early modern period. Daniel Sennert, a well-known professor of medicine at the University of Wittenberg, made this principle an important feature of his influential book, De chymicorum cum Aristotelicis et Galenicis consensu ac dissensu (1619). Like Paul of Taranto several centuries before, Sennert used dissolution and reduction of metals to demonstrate that they were composed of robust corpuscles that defied the supposed corruption required by the Thomistic theory of generation and mixture. Unlike Paul, however, Sennert was able to use the powerful mineral acids (nitric, sulfuric, hydrochloric, and aqua regia) to provide much more dramatic demonstrations of metallic destruction than had been possible with the weaker acids and bases available in the thirteenth century. Sennert's work would in turn be appropriated and reworked by Robert Boyle in his early atomism and later corpuscular theory. Like Paul of Taranto and Sennert, Boyle would use the reduction to the pristine state as a means of demonstrating the reality of robust corpuscles beneath the level of sense. Hence this inheritance of medieval, scholastic alchemy would resurface as an essential feature of Boyle's mechanical philosophy. Needless to say, this continuity belies the traditional view of Boyle's mechanical chemistry as a radical break tout court with alchemy on the one hand and scholasticism on the other.9

ANALYSIS, SYNTHESIS, AND MASS BALANCE

The early use of the reduction to the pristine state by scholastic alchemists in their attempt to derail the Thomistic theory of generation and mixture, followed by the subsequent adoption of this technique and its interpretation by nonscholastic authors such as Boyle, raises an interesting question. What further aspects of the long alchemical engagement between philosophy and the laboratory may have entered into chemistry in the seventeenth century? In a sense the question is misplaced, since one thing that we have learned in the last twenty years is that there was no broadly accepted, sharp demarcation between alchemy and chemistry until the eighteenth century. “Chymistry,” for the most part, still included the transmutation of base metals (chrysopoeia) as a goal, along with other traditional alchemical desiderata.10 But the question still lingers: What did alchemy or chymistry supply to the very field that would in the late eighteenth and early nineteenth centuries see itself as something quite distinct from the aurific art?

The alchemical emphasis on laboratory analysis provides an important starting point. The reduction to the pristine state, as employed by Boyle's immediate predecessors such as Sennert, consisted of both a synthetic and an analytic stage. First, the metal or other material was dissolved in a solvent, thus forming an apparently homogeneous mixture between the solvent and the dissolved. After this, the dissolved material was reduced, often by means of an added base such as salt of tartar (potassium carbonate). In effect, the mixture was analyzed into its components. Hence already in Sennert one sees a paired cycle of synthesis and analysis used in a demonstrative proof. The work of the Flemish chymist Joan Baptista Van Helmont also capitalized on cycles of analysis and synthesis. Van Helmont made glass out of sand and salt of tartar, for example, employing an excess of the latter to produce the soluble silicate that we nowadays call “water glass” (potassium or sodium silicate). Then he reacted the water glass with nitric acid and found that a sand-like product was released. He concluded that the grains of sand used in the original glass had been present all along and that glass itself, despite its uniform appearance, was in fact not a scholastic perfect mixture but, rather, was composed of bonded particles at the microstructural level. Hence we find Van Helmont, like Sennert, employing analysis and synthesis in the interest of disproving Thomistic mixture theory.

Unlike Sennert, however, Van Helmont added an important consideration to the demonstrative use of analysis and synthesis. In short, Van Helmont weighed the ingredients that went into the glass and then reweighed the final products. He claimed, as a result, that the sand that went into the glass had the same weight as the sand that was released by analysis with acid. This emphasis on weight is entirely characteristic of Van Helmont's approach throughout his work, not just in the famous willow tree experiment that led to the incorrect—though widely believed—conclusion that the tree was composed solely of water. Indeed, Van Helmont's chymistry led him to an explicit recognition of the principle of mass balance. Although the principle ex nihilo nihil fit (“nothing comes from nothing”) had been widely accepted since the Presocratics, Aristotelian physics did not view preservation of matter as entailing preservation of mass. Moreover, Aristotle had accepted that elemental fire was endowed with levity rather than gravity; and since fire could be transmuted into heavy earth by mere replacement of its heat by cold, it was in principle quite possible to begin a reaction with ingredients having practically no weight at all and to end it with products that were drastically heavier (or vice versa). Coupling argument together with experiment, Van Helmont came to the conclusion that this view was untenable and summarized his own position thus: “Nothing comes into being from nothing. Hence weight comes from another body weighing just as much.” Although mass balance had been implicitly assumed by metallurgists and alchemists for many centuries, it had existed in a subterranean and uneasy relationship with the dominant Aristotelian physics of the Middle Ages. It was only in the mid-seventeenth century that Van Helmont explicitly enunciated the identity of input and output weights and used this as a tool for undermining the traditional theory of the four elements. Again, alchemy provided the interplay of philosophical questioning and laboratory experiment that distinguished this field from either pure technology on the one hand or unalloyed scholastic disputation on the other.11

CHYMICAL ATOMISM AND THE NEGATIVE-EMPIRICAL PRINCIPLE

The explicit enunciation of mass balance in seventeenth-century chymistry was accompanied by another chymical development of equal historical significance. Already in the Summa perfectionis of Geber one finds a hierarchical theory of matter according to which corpuscles of the four elements combine to form larger particles of the principles sulfur and mercury, which in turn combine to give particles of the metals proper. As Geber says, at both levels of composition corpuscles are bonded together in a fortissima compositio—literally, a “very strong composition” or “very strong juxtaposition.” Sennert, building on this concept in the early seventeenth century, speaks of “atoms of their own genus.” Like Geber, Sennert accepts the theory of the four elements and argues that fire, air, water, and earth bind together to form prima mixta—“first mixeds”—meaning the alchemical principles mercury and sulfur along with the Paracelsian principle salt. These atomic principles in turn combine (while retaining their robust identity, of course) to form metals and other materials. The same hierarchical theory, minus the four elements, reappears in Boyle's work. Adopting Sennert's term “prima mixta,” Boyle says that these primary mixts are made up of smaller particles of uniform catholic matter called prima naturalia (first natural things). The larger prima mixta can themselves combine to form still larger corpuscles that Boyle calls decomposita, and these can in turn combine to form still larger particles. Variants of this terminology would be adopted in the Stahlian chemistry of the eighteenth century, where they would attain a prominent place. But how did chymists justify their belief in atoms or semipermanent corpuscles at any level?

A careful examination of alchemical theory and practice reveals that already in the High Middle Ages (if not earlier) alchemists were employing an approach that modern historians of chemistry have dubbed the “negative-empirical principle.” According to this analytical ideal, a substance is viewed as elementary if it cannot be decomposed by the tools of the chymist. Undoubtedly the most famous enunciation of the negative-empirical principle lies in Antoine-Laurent Lavoisier's definition of an element as “the final limit that analysis reaches.” By strictly limiting all discussion of elements to the products of chemical analysis, Lavoisier was able radically to reframe the category of materials that attained elementary status. Historians have also located the negative-empirical approach in Boyle's famous definition of elements as “those primitive and simple Bodies of which the mixt ones are said to be composed, and into which they are ultimately resolved.”12 Given Boyle's recently discovered reliance on earlier chymical atomists such as Sennert, it should come as no surprise that the negative-empirical principle is actually widespread in earlier alchemical sources. An operational atomism based on the resistance of materials to laboratory operations such as sublimation and calcination can already be found in the Summa perfectionis of Geber. When knowledge of the newly discovered mineral acids came into circulation, probably a few years after the Summa's composition, these strikingly powerful analytical agents provided chymists with the means to identify “atoms” that could withstand dissolution into their components. This is the sense of Sennert's use of the term “atom”—namely, a particle of matter that cannot easily be divided or decomposed into other substances, if it can be divided at all; the same usage can be found in the work of the young Boyle.

The operational atomism of many medieval alchemists and their early modern heirs de-emphasized several of the features that ancient atomism had stressed. Although the size of the particles remained important for determining features such as volatility, solubility, and the ability to penetrate the pores of other bodies, few medieval alchemists or their early modern successors were concerned with the shape or motion of their corpuscles. Exceptions can be found, certainly, such as the late seventeenth-century chymist Nicolas Lemery and of course Boyle himself, but for the most part the tradition of chymical atomism was content to think in terms of semipermanent corpuscles of undefined shape. This feature, though it may seem to lend an air of vagueness to texts of the period, actually testifies to the empirical commitments of the alchemical tradition. Working with their hands among furnaces and stills, medieval and early modern alchemists tried to argue from the appearances rather than importing unnecessary speculation into the invisible realm beneath the senses. In this respect, they were very much the forebears of nineteenth-century chemical atomists, whose own version of the atomic theory seldom descended into the microworld beyond the claim of combining units of fixed weight having a marked degree of substantial robustness.13

Conclusion

As I have stressed throughout this essay, alchemy was normally characterized by the interaction of head and hand. To be sure, there are numerous late antique, medieval, and early modern practicae (process- and recipe-books) that represent a less philosophically oriented alchemy than what one finds in the Geber corpus or in the work of Sennert. Even in the more technological genre of the practicae, however, one frequently finds a tacit integration of theory and practice, as when the principles sulfur and mercury are employed to lead base metals to perfection. Additionally, alchemical recipes and processes often represent the final stages of working out known chemical properties and interactions. An excellent case of this may be seen in the laboratory notebooks of the early American chymist George Starkey, who explicitly devised “conjectural processes” based on his Helmontian theory and personal laboratory experience. Starkey would go on to test his proposed operations by carrying them out; once tested, he would revise or reject them in accordance with their success or failure.14

The widespread integration of theory and practice that one finds in medieval and early modern alchemy testifies to the discipline's intermediary position between the purely verbal disputations that occupied many university arts masters and theologians on the one hand and the realm of metallurgical assayers and apothecaries on the other. It is important not to engage in a simplistic reification of speculation and operation as though they were watertight categories. The De mineralibus of the thirteenth-century theologian and natural philosopher Albertus Magnus, for example, provided a philosophically sophisticated discussion of alchemy that in turn served as a major inspiration for alchemists themselves, probably including the author of the Summa perfectionis. At the same time, Albertus's impressive study of minerals and metals was still being read by those with an interest in mining and mineralogy well into the early modern period.15 This surprising gift of medieval scholasticism found its inspiration in the work of Aristotle, whose individual treatises received long and detailed commentaries by Albertus.

In conclusion, then, alchemy has a great deal to teach historians of science. The lessons range from the nature and influence of Aristotelian natural philosophy and scholasticism in general to the particulars of medieval and early modern matter theory and extend even to the origins of modern chemistry. The recent historiography of alchemy has revealed an experimental side to medieval scholasticism that urgently needs to be examined in other fields as well, such as optics and medicine. The presence in medieval and early modern alchemy of an operational atomism based on laboratory procedures such as analysis and synthesis derails attempts to impose a spurious uniformity on matter theory before the emergence of the mechanical philosophy. And the explicit enunciation and implementation of doctrines such as mass balance and the negative-empirical principle in an alchemical context provide an important and previously unsuspected link with later chemistry. Many more lessons can be drawn, of course, as the other essays in this Focus section reveal. Perhaps because of its very dismissal by the early generation of professionalized historians of the Scientific Revolution, the subject still promises many surprises.