L’Alchimista antico is a timely survey of Greek alchemical treatises composed between the first and the ninth century CE. It follows on a new wave of research and editorial work inaugurated by the Alchimistes grecs collection in the early 1980s and continued by Matteo Martelli with his own edition of the Four Books attributed to Democritus. Unlike most other introductions to Greek alchemy, Martelli’s book brings textual evidence into clear focus and avoids making conjectures on the purported unity of ancient alchemical theories.

The first chapter deals with definitions. Martelli first reassesses two influential historical models. Working on a universal definition of alchemy, Joseph Needham hypothesized a distinction between aurifiction (the production of imitation gold) and aurifaction (the creation of gold). As Martelli points out, Needham’s neat dichotomy cannot explain how certain Greek alchemical recipes were described by expressions like chrusou poiēsis (“gold-making”) while referring to obvious processes of coloration or gilding (11–14). Martelli also points out that, contrary to what André-Jean Festugière’s famous definition of ancient alchemy implied, the theoretical baggage of the alchemical tradition cannot be exclusively attributed to “the Greeks,” as if “the Egyptians” had only contributed to the technical aspects of the art. Alchemical texts, as Martelli shows, sometimes mixed elements that we now associate with the ancient Jewish, Egyptian, Persian, and Greek cultural spheres (17, 45, 55–57, 63–73). Some material might have also come from the Near East. Expanding his earlier studies on the Four Books of pseudo-Democritus (which include the treatise more commonly known as the Physica et mystica), Martelli demonstrates that, despite an old tendency to see alchemy solely as a gold-making art, evidence shows that the alchemical corpus encompasses a large spectrum of techniques (36–41). Martelli also points out that ancient alchemy generally combined interest for practical and theoretical questions (42–43).

Chapter 2 looks at foundational alchemical authors by concentrating on the alchemical works attributed to Democritus (first century CE) and Zosimos of Panopolis (third/fourth century CE). Evidence taken from Zosimos as well as from inscriptions found in Egyptian temples show the “permeability of the boundary between the technical expertise of the artisans who worked in Egyptian temples and the later alchemical tradition” (55). This early form of alchemy, as Martelli argues, was indebted to the crafts of Egyptian artisans who worked on the preparation of statues in the “house of gold” of Egyptian temples (55). Turning to the fragments of the Four Books attributed to Democritus and the oldest alchemical manuscripts, P.Holm. and P.Leid. I 397 (third or fourth century CE), Martelli first demonstrates that Bolos of Mendes (third century BCE) is not the author of the alchemical books attributed to Democritus, as scholars have usually reported (59–60). Martelli also looks at the “marked syncretism of the ancient alchemical tradition” by analyzing foundational myths of the alchemical art, such as the meeting of Democritus and of the magos Ostanes in Memphis (70). The rest of the second chapter is devoted to Zosimos of Panopolis, his social milieu, and his use of enigmatic language. Zosimos advocated the use of “natural tinctures” (i.e., “processes based on the study of natural substances” [81]) and frowned upon the use of “timely tinctures,” which he attributed to his rivals. Timely tinctures did not work naturally but necessitated the use of sacrifice to demons. Still, according to Zosimos, success in this type of alchemy was dependent on the whim of the demons (79–81). Martelli argues that Zosimos’ rivals were Egyptian priests.1

Turning to Zosimos’ use of enigmatic language, Martelli argues that the vivid and violent dream-like narrative of the First Lesson on Excellence dramatizes alchemical operations (85). This interpretation is derived from Zosimos himself, who associated each vision with short expressions that can arguably be interpreted as technical operations. In the case of the first vision, Zosimos noted that it was a representation of the thesis hudatōn (“the setting of the waters”). Martelli reads this expression as the description of a technical procedure and evokes other interpretations without engaging them directly (86). While Martelli’s interpretation is plausible, it is not clear what Zosimos meant by thesis hudatōn. As far as I know, the expression is not discussed by Zosimos in his extant (Greek) treatises. The waters in question appear to have been the same “divine water” (theion hudōr) he mentioned elsewhere. This, however, is not entirely helpful considering that Zosimos ascribed different meanings to the notion of “divine water.”2 It is also likely that “thesis hudatōn” was just as polysemic as “theion hudōr.” This appears to preclude exclusive readings of the expression. While a technical reading is still plausible, other readings are also possible. In fact, several scholars have argued that the Lesson illustrated how alchemy could liberate the soul from the body.3 Could Zosimos have meant a connection between the work of “tinctures” and the work of immortalization?4 This seems to me to be a good reason why he would have used a third representation uniting these two, namely, the allegorical narrative itself.

In chapter 3, Martelli surveys the most important works of alchemy from the fourth to the seventh century CE, while paying attention to the connections between alchemy and natural philosophy. Proclos and Aeneas of Gaza, Martelli argues, described transmutation processes recalling those found in pseudepigraphic treatises that would have been attributed to Synesios of Cyrene (fourth to fifth century CE) and Olympiodoros “the Younger” (sixth century CE)—two scholars who studied in Alexandria like Proclos and Aeneas (89). The later alchemical tradition (sixth to eleventh century CE) is introduced with the story of Isthmeos, described in Byzantine chronicles as a cheimeutēs and a swindler. Martelli deduces from this story that the term cheimeutēs, recalling that of cheimeia (“alchemy”), carried negative connotations outside of the alchemical corpus in the sixth century CE (96–97). Evidence suggests that extant alchemical anthologies originated in compilatory work started a century later during the reign of Heraclius (97–100). One of the most important ancient alchemical authors, Stephanos of Alexandria, dedicated one of his Lessons on Gold-Making to Heraclius, further suggesting that the emperor patronized works of alchemy. Stephanos advocated the use of clear language, strove to integrate alchemy within natural philosophy, and described transmutation using the concept of exhalations found in Aristotle’s Meteorologica.

Turning toward material culture in the last chapter, Martelli notes that alchemical texts do not mention “alchemical laboratories” (119–128). Following an observation made earlier by Robert Halleux, he remarks that the names of substances and of tools are mentioned as belonging to certain trades, which would imply that the authors of the texts in question did not belong to these trades (e.g., “take a quarter of tested gold and cast it in a goldsmith’s crucible”; 127–31). Alchemists, Martelli argues, were defined by the choice of the right tool for the right occasion and by a tendency for systematizing the work and the implements of certain crafts (131–37). The last part of the book analyses texts describing two types of distillation apparatus as well as the mysterious kērotakis.

As with all surveys, omissions are just as important as assertions. It is noteworthy that Martelli neither mentions nor proposes the hypotheses that alchemical authors necessarily pursued the goal of transmutation or that their different enterprises necessarily implied a form of philosophical inquiry into the nature of matter. I take Martelli’s omission as indicative of a positive change in the historiography of ancient alchemy. It is also worth noting that Martelli has generally avoided engaging with evidence pointing to overlaps between what we would call religious and technological concerns.

This last kind of omission notwithstanding, L’Alchimista antico is a reliable introduction to ancient Greek alchemy. Meticulously researched, up-to-date, and mostly free from conjectures, it is the best introduction to ancient Greek alchemy currently available.

Olivier Dufault



See Matteo Martelli, “Alchemy, Medicine and Religion: Zosimus of Panopolis and the Egyptian Priest,” Religion in the Roman Empire 3 (2017): 202–20.


See Matteo Martelli, “‘Divine Water’ in the Alchemical Writings of Pseudo-Democritus,” Ambix 56 (2009): 9–10, and Michèle Mertens, ed., Les alchimistes grecs, 4.1, Zosime de Panopolis: Mémoires authentiques (Paris: Belles Lettres, 1995), 21.


See Garth Fowden, The Egyptian Hermes (Princeton: Princeton University Press, 1986), 120–26; Mark J. Edwards, “The Vessel of Zosimus the Alchemist,” Zeitschrift für Papyrologie und Epigraphik 90 (1992): 55–64; Kyle Fraser, “Baptized in Gnōsis: The Spiritual Alchemy of Zosimos of Panopolis,” Dionysius 25 (2007): 33–54; Sergio Knipe, “Sacrifice and Self-Transformation in the Alchemical Writings of Zosimus of Panopolis,” in Unclassical Traditions, ed. Christopher Kelly et al. (Cambridge: Cambridge University Press, 2011), 2:59–69. See also Shannon Grimes, Becoming Gold: Zosimos of Panopolis and the Alchemical Arts in Roman Egypt (Auckland: Rubedo, 2018), 127–53, and Olivier Dufault, “Lessons from the Body,” in Lovers of the Soul, Lovers of the Body, ed. Ilaria L. E. Ramelli and Svetla Slaveva-Griffin (Cambridge, Mass.: Harvard University Press, forthcoming).


This metaphor can be found in the Coptic Gospel of Philip (NHC II.3.61). See Régine Charron and Louis Painchaud, “‘God is a Dyer’: The Background and Significance of a Puzzling Motif in the Coptic Gospel According to Philip (CG II, 3),” Le Muséon 114 (2001): 41–50.