In the ancient world, magic generally was based on a theory of sympathetic relationships in the world. Human beings, animals, demons, metals, stones, and other materials had sympathetic or antipathetic connections with each other. In the Classical world, this led to the idea of homoiopatheia or similia similibus curantor; an understanding that any harm could be cured by the same or similar substance that caused it. For example, magical dolls could be created to mimic the source of evil, which then could be bound or pierced with needles to stop the power from escaping. Stones were worn to have a therapeutic effect on the body, or alternatively, could serve as apotropaic devices against dangerous influences, such as an attack by a demon. Jews had similar beliefs and practices. From archaeological discoveries, we know that ancient Jews used all kinds of magical objects in their daily lives.
The evil eye was widely feared in Jewish society since one never knew when and where it would strike. It belonged to a specific realm of magic that did not require special incantations or chants to invoke its power. Any person or supernatural could look at someone and bewitch or charm him or her. A specific category of supernatural creatures that was especially well connected to the evil eye were demons: they were viewed as able to transfer their evil intentions onto people, places, and things through the power of their eyes. According to rabbinic texts, the best way to combat the evil eye was knowledge of the Torah. In addition to citing Scripture, there were also spells that could be used to counteract magic. Since removing the evil eye was difficult and sometimes even impossible, however, focus was mostly placed protecting against it. For example, benedictions could be uttered regularly to ward off the eye. Reciting the words of Gen 49:22 provided strong protection against the evil eye. Over time, other methods and practices were developed as security measures. One of the most famous methods was the placement of incantation bowls (also known as Aramaic magic bowls or demon bowls) under the corners or thresholds of buildings. These magic bowls were used mostly in Upper Mesopotamia and Syria between the 6th and 8th centuries CE and acted as traps to snare demons trying to enter the building through the ground. The clay bowls were inscribed on the inside with scriptural quotes and other incantations, mostly in Jewish Babylonian Aramaic, spiraling towards the middle, with often a depiction of a demon in the center. The bowls were buried upside down, and to date around 2000 examples have been found in excavations.
Another apotropaic device was amulets (kame’a or kami’a), which are referred to throughout the Talmud. bt. pesachim 111b tells of a rabbi writing an amulet to protect a city-ruler against a demon and another rabbi writing an amulet to protect against 60 demons, indicating that rabbis were involved in the manufacture of these devices. The Talmud even states that a person may carry an amulet in public on Shabbat provided it was made by an “expert” (m. Shabbat 6:2; t. Shabbat 4:5, 9, 10; bt. Shabbat 53a, 61a). Amulets could come in various shapes and materials, but the most common ones were pieces of paper, parchment, or metal inscribed with various formulae in Aramaic. A specific example is the amulet with the Seal of Solomon to restrain the evil eye, discovered in various archaeological excavations. The use of inscriptions to ward off the evil eye stemmed from a belief in the holiness and power of words. The text of the Priestly Blessing (Num 6:24–26) was considered effective against the evil eye. Permutations and combinations of the letters of the different names of God were frequently used; names of angels were also common. The simplest amulets were inscribed with the name of God on a piece of parchment or metal, usually made of silver. They were worn close to the person, as a piece of jewelry or sown into the clothing. Amulets have also been found in ancient synagogues in Palestine, mostly in the form of inscribed thin metal plaques (called lamellae) rolled up or folded, indicating a connection between synagogues and the supernatural. At the synagogue of Ma’on (Nirim), for example, archaeologists discovered 19 amulets of thin copper or bronze in the apse area. Together with the amulets were other small finds, including bone and ivory objects, iron nails, fragments of pottery and lamps, and five coins. The excavators suggest that the finds might have been stored in the Torah shrine or a wooden box inside the shrine. Some of the amulets were wrapped in cloth, suggesting they were treated with care. One of the rolls still had the remains of a thread adhering to it, indicating according to Rahmani that it originally was worn around the body, probably as a necklace, or, according to Naveh and Shaked, suspended from the Ark or the wall behind the Ark. Three have been published so far under the auspices of the Israel Museum: one amulet seems to ask to relieve a woman named Natrun of headaches, the second was written to protect a mother and child, and the last one was made to protect a girl/woman named Asther from evil spirits, including the evil eye. At Meroth, a bronze amulet measuring 4.8 X 13.8 centimeters was found below the threshold of the easternmost entrance in the north wall of the Phase II synagogue. It is dated to the 7th century, when the new north wall of the synagogue was constructed. The amulet has 26 lines of texts written in a mixture of Aramaic and Hebrew, with the supplication of a man named Yosi ben Zenobia, who asks God for control over the community (“so may the people of this town be suppressed and broken and fallen before Yosi son of Zenobia”). At the small synagogue of Bar’am, archaeologists found a rolled-up bronze amulet under the corner of a large stone in the western stylobate wall, inscribed with an Aramaic blessing and a protection against the evil eye. One bronze lamella was also found in the fill of Building 300 directly below the Horvat Kanaf synagogue. This amulet was inscribed with a song of praise, followed by a healing prayer. While its exact finding spot is unknown, it could have been connected to the synagogue; the stratigraphy of the archaeological site is, as we have seen, rather complex, making it possible that the lamella was originally placed somewhere in the synagogue building. Finally, at Korazin a bronze amulet was found in the fill between the stones underneath the threshold of the eastern entrance in the south wall. The context of these amulets (under the threshold, hidden in a wall, kept inside the Torah shrine) shows that they were purposely installed in the building. For the amulets hidden under the thresholds, or walls, it is unclear if they were installed with approval of the synagogue builders or if there were snuck in. Nevertheless, the synagogue was apparently seen as an appropriate space to hold magical items; a powerful place indeed.
So far, we explored two categories of objects that were hidden underneath the floors and thresholds, and inside the foundations and walls of ancient synagogues to protect the building or individuals against the evil eye: magic bowls and amulets. To this assemblage, we may add another group of strange objects: bones. Under the threshold of the entrance to the synagogue at Dura-Europos in Syria, the remains of two human finger bones were discovered. The bones were found in a cavity under the doorpost, gouged out of the rubble bedding upon which the sill was set, sealed by a metal plate. Since the bones were carefully placed in a purposely made socket, it is clear they were put there intentionally. Unfortunately, this is the only example we have of human bones deliberately deposited in an ancient synagogue. We do have one Talmudic passage that might explain its function:
Why do they go to the cemetery? With regard to this there is a difference of opinion between R. Levi b. Haman and R. Hanina. One says: [To signify thereby], we are as the dead before Thee; and the other says: In order that the dead should intercede for mercy on our behalf (bt. Ta ‘anith 16a)
This passage indicates that the dead can act as intermediaries on behalf of the living. Could this function also include protection against evil spirits or the evil eye?
The last category of magical “objects” from ancient synagogues are magical decorations, graffiti, and other motifs added to the structure. In the Dura-Europos synagogue, for example, two of the 234 ceiling tiles display the “much suffering eye,” which was used throughout the Roman Empire as a protective symbol against the evil eye. In several other synagogues, such as Meroth, Qasrin, and ‘En-Nashut, entrances and other strategic locations were decorated with the so-called “Hercules knot” (a wreath consists of stylized leaves ending in a bound ribbon), a symbol that was considered to have apotropaic properties. Last, graffiti scratched into the walls or written on the surfaces by visitors, mostly around the door openings, have also been found in multiple synagogues. The most famous example is perhaps Dura-Europos, where hundreds of vernacular drawings and writings have been discovered, written in Aramaic, Hebrew, Greek, Middle Persian, Parian, and Pahlavi script. According to Karen Stern, many of these can be interpreted as wishes made by the inscribers for passersby to see and read, perhaps even out loud before a human and/or divine audience: they were magical “graffiti-petitions.”
Before considering coins as possible means of protection against the evil eye, we need to complete our overview of apotropaic devices by mentioning some objects that so far have not been found or recognized in ancient synagogues. Among these are mirror plaques, which are clay or stone tablets into which one or more mirrors were set. The plaques were round or shaped like a bird, fish, temple, or menorah, and have been found mostly in graves. Small holes indicate that they might have been hung around the house. The plaques likely were used to reflect the evil eye, reflecting back the harmful gaze. Up until now, however, no remains of these have been found in ancient synagogues, possibly because they were meant to be seen by visitors and could not be hidden in walls or other parts of the building. Magical gems (often placed in rings) have also been discovered at Mediterranean sites dating from the 1st century to the Byzantine period. They are made of semi-precious stones such as jasper, steatite, hematite, and carnelian, and were engraved with various images including humans, mythological creatures, animals, or floral motifs. The poor execution of the images might indicate that the gems were worn on the body not for their esthetic beauty but for their magical function. Short inscriptions on some of these gems confirm this: for example, a black jasper gem found in Caesarea depicts a man harvesting wheat, with the inscription “for the hips,” indicating some sort of wish or medical remedy. However, these inscriptions have been found only in Greek, not Hebrew, indicating that the Jewish population might have refrained from using figurative devices. They have also not (yet) been found in ancient synagogues, perhaps indicating that they were solely meant to be worn as jewelry on the body and not placed inside buildings. Finally, from rabbinic sources we know of other apotropaic devices made out of perishable materials such as eggs or a cow’s afterbirth. Obviously, these are not preserved in archaeological contexts.
 Thompson 1971, pp. 142–174; Ulmer 1994, p. 133; Bar–Ilan 2002.
 Ulmer 1994, p. 134. The same could be done by wearing coins as amulets, see below.
 Ulmer 1994, pp. 137–138.
 For examples in rabbinic literatures, see Ulmer 1994, pp. 153–154. One demon in particular was feared, a monster called Ketev Meriri, who stared with one eye, and was often depicted as having many eyes.
 Ulmer 1994, p. 140.
 b. Pesahim 111a.
 m. Aggadah I 191; Pesiqta Rabbati 5:10.
 See, for example, the work done by Naveh and Shaked 1985; Shanks 2007; Bamberger 2015, 2020; Gideon 2019 for more information on the content of the magic bowl writings. I would like to thank Avigail Manekin Bamberger for meeting with me and further discussing the ideas in this project.
 Unfortunately, there is widespread trade in illegally excavated incantation bowls, with thousands more known to exist on the black market.
 Chronologically, all amulets found in controlled excavations from Late Antiquity date to the 5th and 6th century CE. However, we know what the tradition is much older, as indicated by the Ketef Hinnom amulets of the 7th–6th century BCE, as well as Iron Age II amulets from Beersheba, Megiddo, and Lachish (Schmidt 2016, pp. 123–144).
 This also shows that rabbis were involved in the creation of amulets. The same can be said about Aramaic magic bowls, which often contain Talmudic passages or even the names of rabbis. See, for example, Swartz 2018, pp. 34–38.
 Elliot 2016, p. 133 and following. The literature on Solomon’s role in Judeo–Christian magic is vast: see for bibliography Russell 1995, p. 39.
 For example, some Fayyum mummy portraits in Egypt depict the deceased with a cylindrical metal capsule around the neck. Similar capsules to those containing rolled metal amulets have been found in archaeological excavations (Thompson 1982; Bohak 2008, p. 150, Rowan 2009, p. 4).
 Kotansky 1994; Fine 1997, pp. 73–75, 145–146; Bohak 2008, pp. 318–322; Eshel and Leiman 2010, p. 189; Hachlili 2013, pp. 537–538; Stern 2016; 2018; 2021.
 Rahmani 1960, pp. 14–16.
 Naveh and Shaked, 1985, pp. 90–101.
 For the full text, see Naveh 1985, pp. 282–367 (Hebrew), Ilan 1989, pp. 29–30, Ilan 1995, pp. 270–272, and Stern 2021, pp. 227–229. This is the only amulet that contains a curse instead of a blessing, which puts it in the category of defexiones; curse or binding tablets that have been found throughout the Hellenistic and Roman–Byzantine world, mostly written in Greek. Another example of a curse in a synagogue context can be found at ‘En Gedi, where an 18–line mosaic inscription curses, among other things, people who reveal the secret of the town to the Gentiles (Levine 1981a).
 Naveh 2001, pp. 179–180; Aviam 2004, p. 159; Bohak 2008, p. 319; Stern 2021, p. 227.
 Naveh and Shaked 1998, p. 51; Stern 2021, pp. 226–227.
 Several more amulets have recently been found in ancient synagogues, but are not yet published (Stern 2016, p. 225).
 This might document the individual’s effort to take advantage of the potency of select movable and architectural features inside the synagogue, to deposit messages for their own gain (Stern 2021, p. 242).
 To be clear, amulets have also found in other contexts, for example in houses (at Horvat Kanaf, which Hachlili 2013, p. 537 mistakenly associates with the synagogue at Kanaf, and at Khirbet Wadi Hamam (Leiman and Leibner 2016; Leiman 2018)), in commercial settings (Sepphoris), and in tombs (Tiberias, Samaria, Emmaus). I am focusing here, however, only on the ones found in synagogues to analyze a possible connection between the function(s) of amulets and coin deposits in synagogues.
 Kraeling 1956, p. 19; Magness 2012b. Unfortunately, the bones are now lost as they were among the materials left in the excavation house at Dura, which was destroyed in the fighting during the 1948 Arab–Israeli war.
 The doorpost is the same space where most of the “prayer” graffiti in the synagogue of Dura–Europos were found, indicating the importance of thresholds and doorways to ancient Jews living in the city (Stern 2012, see below).
 That we know of. It might well be that more bones were hidden underneath ancient synagogues, but: A. most archaeologists do not lift the thresholds or walls of their buildings to look for objects underneath it, B. bones cannot be detected by metal detectors, or other ground penetrating radars to go look for them, C. human bones could have been found but discarded by archaeologists as just animal bones or unimportant materials, D. this could have been a local, Syrian phenomenon. Since no other ancient synagogues have been excavated in Syria, we do not know if there were similar examples in other cities.
 If so, this could have been in analogy to the Christian cult, where the bones of saints are often kept as relics in churches to intercede with God on behalf of the community (Magness 2012, p. 236). Since the inhabitants of Dura–Europos were a mixed pagan, Christian, and Jewish community, it would be no surprise if different groups influenced each other’s rituals. Bohak 2008, pp. 193–194 refers to a handful of human skulls inscribed in Aramaic from Late Antique Mesopotamia, but these have never been analyzed. Exceptions are a skull with a love–inducing spell written on it, currently housed at the Vorderasiatische Museum in Berlin (see image and interpretation Saar 2017, pp. 126–128), and four other skulls in the same museum with incantations and drawings similar to the Aramaic magic bowl writings (Levine 2006). These skulls show that the use of human body parts in Jewish magic might not have been as rare as previously thought, at least in the Diaspora.
 Kraeling 1956, pp. 48–49; Goodenough 1953–1968, Vol. II, pp. 238–241; Bohak 2008, p. 322.
 See for example, Nicgorski 2013, Stern 2016. This knot might also be connected to the use of knots against kashfaniyot or witches. In bt. Shabbath 66b we read: “Three [knots] arrest [illness], five cure, seven are efficacious even against keshafim.”
 Noy and Bloedhorn 2004; Stern 2012; 2021, pp. 233–239.
 Stern 2021, pp. 235–236.
 Rahmani 1964; Fischer and Saar 2007; Saar 2010, p. 19.
 Bohak 2008, pp.158–165.
 Saar 2010, pp. 17–18.
 This is, of course, highly speculative and there may be many other reasons for why similar gems do not bear Hebrew or Aramaic inscriptions. It could, for example, indicate that the Jewish population was influenced by Greek culture and did not mind having their magical devices written in Greek, the lingua franca of the time.