As we have seen, deposits of coins stored in genizot, charity hoards, treasuries, and emergency hoards are found in synagogues constructed and used ranging from the 4th to 8th century CE, with a more-or-less equal spread throughout these centuries. However, trying to pinpoint when the phenomenon of the magico-religious coin phenomenon started and ended is more difficult. To date the origin of the practice, we must take into account three factors vis-à-vis coins: first, the date of minting, second, the date of the (first and possibly second) deposition, and third, the possibility of residuality.[574] “Residual” can refer to a coin that remained in circulation long after it was minted and even after it had lost its monetary value. It can also refer to the re-deposition of coins in a secondary context (coins that were once already lost or mixed in with debris, but now disturbed again), which is the definition I use here.
As for the date of minting, tithing deposits usually contain mostly coins of the final quarter of the fourth century, few coins of the fifth century, and rarely some from the beginning of the sixth century CE.[575] The following table gives an overview of the terminus post quem of each deposit, in chronological order:[576]
| Capernaum (deposit 5) | Unknown |
| Capernaum (deposit 6) | Unknown |
| Bar’am (deposit 1) | 341-346 |
| Capernaum (deposit 4) | 341-346 |
| Ostia (deposit 1) | 337-347 |
| Capernaum (deposit 3) | 383-387 |
| Bar’am (deposit 2) | 367-395 |
| Dabiyye (deposit 2) | 395-408 |
| Korazin (deposit 2) | 395-408 |
| ‘En Gedi (deposit 3) | 395-408 |
| Dabiyye (deposit 1) | 402-408 |
| ‘En Nashut (deposit 1) | 395-450 |
| Bar’am (deposit 3) | 425-450 |
| ‘En Nashut (deposit 2) | 337-455 |
| Capernaum (deposit 1) | 457-475 |
| Horvat Kanaf (deposit 1) | 409-498 |
| Meroth (deposit 1) | 507-512 |
| Horvat Kanaf (deposit 2) | 512-518 |
| Qasrin (deposit 1) | 498-518 |
| Deir ‘Aziz (deposit 3) | 527-565 |
| Deir ‘Aziz (deposit 4) | 527-565 |
| Horvat Kur (deposit 1) | 527-565 |
| Capernaum (deposit 7) | 475-575 |
| Sardis (deposit 1) | 565-578 |
| Korazin (deposit 1) | 602-610 |
| Horvat Sumaqa (deposit 1) | 610-613 |
| Sardis (deposit 2) | 612-613 |
| Sardis (deposit 3) | 615-616 |
| Capernaum (deposit 2) | 650-700 |
| Korazin (deposit 3) | 683-750 |
The terminus post quem of each building is the following, in chronological order:
| Ostia | 337-347 |
| ‘En Gedi | 395-408 |
| Dabiyye | 402-408 |
| Bar’am | 425-450 |
| ‘En Nashut | 337-455 |
| Meroth | 507-512 |
| Qasrin | 498-518 |
| Horvat Kanaf | 512-518 |
| Deir ‘Aziz | 527-565 |
| Horvat Kur | 527-565 |
| Horvat Sumaqa | 610-613 |
| Sardis | 615-616 |
| Capernaum | 650-700 |
| Korazin | 683-750 |
If we follow this table, the earliest example of a magico-religious deposit can be found in the Diaspora synagogue at Ostia. This would mean that the phenomenon started outside of Palestine in the middle of the 4th century.[577] However, since the many Ostia excavation reports are confusing, often contradicting themselves on matters of architecture or dating of the specific features surrounding this deposit, this early date might be incorrect.[578] In Israel/Palestine, the earliest examples are found at Dabiyye and ‘En Gedi, with a date of the early 5th century. However, all of this is based on the date of the minting of the coins. As Bijovsky has indicated, coins from the 4th century were in circulation for a long time. Fifth-century coins, on the other hand, are rarer and give a better indication of the residuality of the coins.[579] Thus, another methodology needs to be followed to better identify the inception of the magico-religious floor deposit.
When we look at the construction date of the (floors) of the synagogues in which floor deposits have been found in combination with an assessment of the residuality of 4th and 5th century coins, we get a different picture. A total of 30,675 coins were found in floor deposits. Of these, 5,802 could be dated, with most of them belonging to the second quarter of the 4th century to the fourth quarter of the 5th century (see https://www.ancientsynagoguecoins.com/deposits-categorized-as-magico-religious-deposits/) . After this, there is a sharp drop-off at the beginning of the 6th century.[580] However, since 4th and 5th century coins remained in circulation much longer than 6th century ones, we need to take into account the latest coins in each deposit. Based on this assessment, I am placing the beginning of this phenomenon in the second half of the 5th century.
To better understand what happened in this period that could have sparked this phenomenon, we need to take a deeper look at the geographical area in which the custom was performed.
___________________________
[574] Based on Lockyear 2012, p. 195.
[575] However, we need to keep in mind that large number of coins from any given hoard found in Palestine are illegible and these mostly include fourth-fifth century coins which cannot be precisely dated because of poor minting and preservation. Therefore, our deposits surely include many more fourth-fifth century issues which cannot be identified (Bijovsky 2012, pp. 75–77).
[576] Here I dropped the later, clearly intrusive coins and coins that only have a general Late Roman date. One will also note that the end dates are different from Ahipaz and Leibner 2021, due to fact that they only used the published coin reports, while I corrected these publications using the IAA database.
[577] If indeed true, perhaps the long distance to Jerusalem made the urge to somehow “get rid” of these coins in a different way more palpable?
[578] See appendix, case-study X.
[579] Or at least, fewer of them have been identified. See Bijovsky 2012, pp. 75–77.
[580] This holds true, even if we remove the large number of coins from Capernaum.