The Customs Districts of Roman Egypt
Skip other details (including permanent urls, DOI, citation information)
This work is protected by copyright and may be linked to without seeking permission. Permission must be received for subsequent distribution in print or electronically. Please contact email@example.com for more information. :
For more information, read Michigan Publishing's access and usage policy.
In his Quadragesima Galliarum published at Rome in 2001, Jerôme France suggested three possible ways of setting up customs houses. First, they could be organized as a string of stations, as in the case of the quadragesima Galliarum. Second, that line could form a customs district by encircling an area, as in the case of the portorium of Asia. Finally, customs houses could be set up in specific areas where crossing was obligatory, as with the portoria of Syria and Judaea. Although France's aim was not to provide a complete overview of the Roman customs system Empire-wide, it is nevertheless unfortunate that he so quickly dismisses the case of Roman Egypt as atypical.
Nonetheless, when one tries to apply France's analysis of the customs structures present in Roman Egypt – taking into account the obvious geographical gap in our papyrological, epigraphical and literary documentation for this province – it is interesting to note that all the three types of organization mentioned above were present in the same provincial territory at the same time (roughly the two first centuries of our era).
Thus, the first type of organization is recognizable in the customs houses established along the Nile and its tributaries for collecting the internal customs dues levied in the Delta and the Thebaid.
The perfect example of an organization of the second type is represented by the Arsinoite nome which was encircled by stations where such dues were levied before entering and exiting the area thus defined.
Finally, the system established for the purpose of collecting the external customs dues levied both in the Red Sea ports and in the Mediterranean harbors of Alexandria necessitated a structure of the third type.
After a review of our evidence concerning the kind of structures present in this particular province, I would like to revisit briefly at the end of this paper the question of an Egyptian "specificity" in the Roman Empire.
The first part of this presentation, the review of the evidence at our disposal will be divided into Part A – which deals with what I call the "external customs dues," that is duties levied on goods entering or leaving the province of Egypt at its administrative borders – and into Part B – which deals with duties levied on goods moving inside the Egyptian territory or "internal customs dues."
I. Review of the evidence concerning the customs dues of Roman Egypt
Part A: The external customs dues
These different administrative and fiscal areas are geographically presented clockwise starting at the north of the country.
A.1. Alexandria and the Egyptian Mediterranean littoral
We have only scant indirect evidence for the existence of customs dues levied in this area, all of them linked, or probably linked, with the transport of goods to and from the city of Alexandria. Two of them are important here:
a. the Muziris Papyrus of the mid second century from which we learn that the imports from the eastern trade when bound for Alexandria paid in the customs warehouses of that city the customary τετάρτη, the 25% due levied on the value of such cargoes. As most of these goods coming from the east were raw materials, it has rightly been suggested that, once the τετάρτη had been discharged, a large part of them would have been sold in Alexandria itself, which had a great reputation as a manufacturing centre, instead of being directly shipped to Rome or other parts of the Mediterranean Sea. Of course, once these raw materials had been processed or, in some cases, reprocessed, customs dues were levied before exportation.
b. in this context, a passage of Strabo, where we learn that eastern goods paid customs dues a second time when leaving Egypt, confirms, if need be, the existence of such taxes at the northern border of Egypt.
What may have been the rate of this last customs duty is obviously purely conjectural, but it is quite reasonable to suggest that it could have been in the range of 2%, that is a quinquagesima – as in the customs district of Spain before the mid second century – or a πεντηκοστή, well attested, as we will see, as an internal customs dues in the Thebaid. The other option is a rate of 2.5%, a quadragesima or τεσσαρακοστή attested for instance both in the customs district of the Gauls and in the province of Asia.
Moreover, we cannot exclude that specific duties were not levied on certain types of merchandise, suffice here to mention a specific tax of 5% or εἰκοστή levied on the murex which is attested in the custom law of the province of Asia.
Was Alexandria the only harbor on the Mediterranean coast of Egypt where such duties were exacted? Although we have no evidence for the Roman period, I would suggest that the harbor of Pelousion would not be an unlikely suspect as customs dues were attested for its harbor in documents of the Ptolemaic period.
A.2. The Red Sea harbors
Three harbors on the Egyptian Red Sea coast will concern us in this section: Myos Hormos and Berenike which were the only two Egyptian harbors described as being active ports of trade at the time of Strabo at the end of the first century BC as well as at the time of the composition of the Periplus Maris Erythraei, that is in the middle of the first century AD, as well as Arsinoë-Kleopatris after its refoundation under the name of Klysma during the Trajanic period.
Through the Murziris Papyrus again we know that a due of 25%, a τετάρτη, was levied on goods imported from the east through the Red Sea harbors.
A.3. Syene/Elephantine and the southern Egyptian borders
Syene/Elephantine was the port of entry for goods coming from the African interior, notably from the kingdoms of Ethiophia, Meroe and Axum. The Ethiophians were controlling and performing the last steps of this trade, bringing into Egypt bullion, linen, ivory and spices by means of collapsible boats. It is thus clear that trade was still active in the southern part of Egypt and that it should have generated some substantial profits for the collectors of customs dues. Furthermore, as such taxes were levied in the periods prior to the Roman occupation of the country, it is difficult to imagine that the masters of the Mediterranean world would have discontinued such a practice.
After the conflict between Rome and Meroe in the 20s BC the southern border of Egypt was established at Hiera Sykaminos (modern Maharaqqa) where it seems to have stayed until AD 298 when Diocletian decided under pressure from both the Meroites and the Blemmyes to evacuate this southern area and to establish a new border at Syene.
Unfortunately we have no attestation in our documentation of a specific tax which could have been levied in connection with the trade from and to Nubia. As pointed out by P.J. Sijpesteijn, although we know that the receipts for the πεντηκοστή levied at the customs-house at Syene had been issued for both imported and exported goods, one never finds stated in them that these commodities came from or went beyond the southern border. In the absence of documents attesting such movements of goods it is better to leave the question open for the moment.
Finally, Sijpesteijn also suggested that a tax of 25% comparable to the one attested for the Red Sea harbors may have been levied on goods entering Egypt through its southern border. In the absence of further evidence, this is reasonable but still purely speculative.
A.4. The western desert
We have to wait until the fourth century AD to see something indicating the levy of customs dues in this part of the Roman province.
Two ostraka which have preserved for us lists of soldiers and civilians employed as "guards in charge of the gates at the Great Gate" (τηρηταὶ τῶν πυλῶν τῆς μεγάλης πύλης) at Kysis (modern Douch) have enabled their editor, Guy Wagner, to suggest the reconstruction of a quite developed system of customs-houses spread along Egypt's western border. It would be rather strange if a comparable system had not been already in place at an earlier period, especially as we know that a fortress had already been established at Kysis in the Ptolemaic period.
Part B: The internal customs dues
This time my presentation starts with the north of the country before roughly following the Nile upcountry.
B.1. The Delta
The only information at our disposal for this part of Egypt concerns the movement of goods between the Nile and Alexandria. From Schedia, a town about 30 km to the south-east of Alexandria, a canal left the Canopic branch of the Nile to enter the capital of Egypt through its eastern suburb of Iouliopolis, also named Nikopolis. Both of these towns placed at either end of this navigable facility also possessed a customs station or τελώνιον where merchandise moving up and down the canal was taxed.
Sijpesteijn suggested that, since the city of Alexandria did not juridically and administratively belong to Egypt, there would be a fair chance that taxes on goods leaving Alexandria through the canal were levied at Iouliopolis-Nikopolis, while the collectors at Schedia only took care of goods coming from the χώρα to the capital. In that case, Iouliopolis-Nikopolis would have been the Alexandrian customs-house and Schedia the Egyptian one and the geographical distance between these two towns would have played the role of a customs barrier separating two independent fiscal entities or districts and thus would be comparable to some customs dues levied when crossing the border of two ἐπιστρατηγίαι, a situation well attested as we will see for other parts of the country.
This attractive hypothesis is unfortunately first undermined by the text of Strabo himself who precisely indicated that both goods exported from Alexandria to the country as well as those leaving the Egyptian χώρα for the capital were taxed at Schedia. If one wants to support Sijpesteijn's position, two explanations only are possible in my view: either we have to admit that Strabo decided to summarize a more complex situation or that the administrative and fiscal division between Alexandria and the rest of Egypt was not yet in place at the time of his writing, i.e. around 20 BC. As it happens we do not have to choose between these two options as Sijpesteijn's position has simply been proved untenable by Bernard Boyaval. This author rightly points out that nothing indeed prevented the Roman authorities from having tolls levied on the same cargo on both sides of the canal, in particular when it was probably passing from one customs district to another. Such a decision would have unnecessarily hampered the profits that the tax collectors were expecting to reap from their business.
B.2. The Arsinoite nome
Three main different customs dues were levied during the Principate in the Arsinoite nome: the ρʹ καὶ νʹ which was a nomarchic tax peculiar to this nome; the ἴχνους ἐρημοφυλακία also called simply ἴχνους or even ἐρημοφυλακία which was also collected outside of the Fayum; finally, the λιμὴν Μέμφεως which was levied at the borders between the epistrategies of the Heptanomia and that or those of Lower Egypt. Each of these three taxes did enjoy its own tariff (γνώμων) and rate different from those of the two others. This is confirmed by P.Oxy.Hels. 16 (after AD 134) for both the ρʹ καὶ νʹ and the ἐρημοφυλακία taxes, thus making the existence of a similar γνώμων for the λιμὴν Μέμφεως more than simply probable.
Let us start with the ρʹ καὶ νʹ νομ(αρχίας) Ἀρσινο(ΐτου), the "tax of 1 and 2%." This 3% tax which is currently attested in our documentation from at least AD 9, if not already from the Ptolemaic period, to 211 in nearly all the villages for which customs receipts are attested was within the competence of the nomarch. One can securely admit with Sijpesteijn that originally we were dealing with two taxes which at a certain point in time, maybe soon after AD 133, started to be collected together systematically.
The tax was levied on an ad valorem basis based on the nature and the quantity of the products transported.
Looking at a map we can see that the geographical area along which this tax was collected follows the physical pattern described by France as a "customs district."
The customs duty known as the ἴχνους ἐρημοφυλακία, the "tax for the protection of the desert route," more commonly known under its abbreviated form of ἐρημοφυλακία (or sometimes simply described as ἴχνους) and in a couple of receipts equated to an ἐρημοτελωνία tax is attested in our documentation from AD 41 to 212 and was, like the ρʹ καὶ νʹ, levied all around the Arsinoite. Naturally, and as one would expect from the name of the tax itself, these levies were made at villages situated at the ends of the trails crossing the desert surrounding that nome.
This duty appears to have been levied at a fixed rate according to special categories such as animals, agricultural products and so on. As far as we can tell from the meagre information provided by our documentation on this point, the rate for the ἐρημοφυλακία seems to have been higher than the one exacted from the λιμὴν Μέμφεως, both taxes being levied at a lower rate than the 3% tax.
The sums levied were certainly used for the upkeep of the desert routes and of the ἐρημοφύλακες, the police force whose duty it was to patrol and enforce the security on these trails, to help travellers to find their way both through the desert and sometimes surely also to the customs-houses, and finally to fight against smuggling activities which should have been rife in such an open space.
Was the ἐρημοφυλακία a tax specific to the Arsinoite Nome? The obvious answer is no, for in essence this levy was exacted for the crossing of a precisely determined geographical space, namely the desert surrounding the Fayum. As such the tax was naturally levied at both ends of this area, i.e., on one hand at the villages set up at the entrance to the Arsinoite, on the other at those located at the borders of the nomes surrounding the Fayum.
The last of these customs dues attested in the Arsinoite Nome, the "tax for the harbour of Memphis" or λιμὴν Μέμφεως, was levied on goods crossing the border separating the epistrategies of the Heptanomia in Upper Egypt from that or those of Lower Egypt from at least AD 42 to 216.
Not surprisingly, the λιμὴν Μέμφεως was levied at Memphis and at the villages of both the divisions of Themistos and Herakleides where trade routes were leaving the Fayum to go to the north of the country; as such the tax was naturally not levied at customs-houses set up along the southern part of the Arsinoite (division of Polemon) in villages such as Tebtunis or Kaine. Since the tax was levied on both imports and exports without them having to transit through the town of Memphis, we are faced with what France describes as a cordon douanier, a customs barrier consisting of a string of stations, comparable on a smaller scale for its shape and functions to the quadragesima Galliarum, the customs duty of 2.5% levied on merchandise leaving and entering the Gauls.
Originally the profits coming from that tax were probably used to maintain and improve the port facilities of Memphis. At the time for which this duty is attested, it seems to have been levied at a fixed rate, the products liable to it being, as with the two other taxes, grouped into specific categories. A rate of 2% for this tax is, I think, a good hypothesis.
To conclude this section on the Arsinoite nome, let me add that a detail which strikes the mind is the fact that there is no attestation for the collection of customs dues in communities located further inside this ring of stations surrounding the Arsinoite nome. It seems therefore difficult to avoid suggesting that the Arsinoite nome formed its own customs district where dues had to be paid at stations encircling that area before moving freely inside the whole nome.
B.3. The Thebaid
Shortly after the Roman conquest of the country, Strabo mentions that in this area the traveler would find a place named Ἑρμοπολιτικὴ Φυλακή (the "Hermopolite Guardpost/Garrison") which he described as "a kind of customs-house for goods coming down from the Thebaid" (17.1.41). The Greek geographer mentions in the same passage that after leaving Ἑρμοπολιτικὴ Φυλακή one reached another Φυλακή/"Garrison-Guardpost" that of the Thebaid (Θηβαϊκὴ Φυλακή).
The editors of SB XVIII 13775 have rightly in my view suggested identifying these two customs stations as the ones placed on each border of the administrative limit separating the epistrategy of the Heptanomia and those of Lower Egypt.
What could have been the rate of a customs tax levied at this point? I would suggest a 2% levy on imported and exported goods as such a rate is the only one attested for a series of harbors of the Thebaid.
B.4. From the Nile to the Red Sea
A series of documents recently discovered at Berenike are transit bills or passes for various goods, mainly but not exclusively wine of different origins, destined for local consumption at Berenike, for the supply of crews aboard ships or for exportation outside Egypt throughout this harbor facility.
Although there is no mention of any customs taxes in these documents, the simple existence of such transit bills clearly indicates that goods exported from Egypt via Berenike to the African and Indian coasts were not only controlled, but presumably also subjected to levies. In this case customs taxes were probably paid at their point of departure when leaving the Nile Valley, that is mainly at Koptos but also, at least in the early imperial period, at Apollinopolis Magna (modern Edfu).
What could have been the rate of such a customs due is unknown, but a rate of 2.5% of the value of the goods transported as been suggested by one commentator. Another possibility is a tax of 2%, a rate more customary in that particular area.
II. The Egyptian "specificity"
A. The problem of the rates attested in Egypt
A customs due of 25% levied on the value of the merchandise transported was not specific to Egypt in the Roman period here under consideration. Thus, a tax of the same rate is also attested in the harbor of Leuke Kome in the mid first century (the problem of Leuke Kome being under Roman or Nabataean control at that time is not important for my purpose here). As a similar tax was also exacted at the border between the province of Syria and the Parthian Empire in the second half of the second century, this has led some commentators to conclude with some reason that such an amount was collected on imported commodities all along the eastern border of the Roman Empire.
On the other hand, rates of 2, maybe 2.5, and 3% attested for the internal (and maybe also some external) customs dues levied in Roman Egypt are perfectly in keeping with most of those exacted in the other Roman provinces.
B. The multiplicity of these customs dues
Although our information is much more scanty for the Ptolemaic period than for the Roman one, a great diversity of customs dues levied at different rates is also clearly attested (whether these different taxes were exacted at the same time or whether some were replacing others as time went on is still an unsolvable problem). It is therefore possible to consider that the situation we face in Roman time was partly a legacy of the previous period.
The situation is far more clear in the other provinces and customs districts of the Roman world due to a comparative lack of documents (after all we do not possess customs receipts outside Egypt, although such documents must have existed) and this is especially annoying in the case of the Greek-speaking part of the Empire where a legacy of their Hellenistic past is to be suspected. Nevertheless, some of the documents we do possess for this part of the Roman Empire clearly suggest that a complete uniformity of customs taxes was never planned nor implemented Empire-wide and that specific duties levied on certain types of goods (I have earlier mentioned the 5% tax levied on the murex in the province of Asia) still co-existed with the more general 2 or/and 2.5% levies.
Be that as it may, we need also to remember that such a multiplicity of customs dues remained present in many pre-industrial European societies. Why should the Romans have done a better job of unifying their customs system when more modern societies did not feel this urge until quite recently?
C. The specificity or non-specificity of the documentation for the province of Egypt
If there is an Egyptian specificity, it seems mainly due to the haphazard survival of our primary documentation. Customs receipts and registers were a bureaucratic and fiscal necessity and they must have existed and been produced in every other part of the Roman Empire, although they have not reached us. The chance survival of such documents only in Egypt is, in this case, due to the environmental specificity of this country.
J. France, Quadragesima Galliarum. L'organisation douanière des provinces alpestres, gauloises et germaniques de l'Empire romain, Ier siècle avant J.-C. – IIIe siècle après J.-C. Collection de l'École française de Rome, 278 (Rome 2001) 346.
Although I have not provided specific maps with this presentation, I would suggest the interested reader to refer to those published in R.S. Bagnall and D.W. Rathbone (eds.), Egypt from Alexander to the Early Christians: An Archaeological and Historical Guide (Los Angeles 2004), a work nowadays easily accessible.
Strabo 17.1.45, with Young, op.cit. (above, n. 4) 40; Myos Hormos is the only harbor mentioned by Strabo (2.5.12 and 16.4.24) in connection with the trade with Africa and India. Periplus Maris Erythraei 1 (thereafter PME); for the probable date of this work, see L. Casson, The Periplus Maris Erythraei, Text with Introduction, Translation and Commentary (Princeton 1989) 6; and Young, op.cit. (above, n. 4) 6.
On this trade, cf. M.G. Raschke, "New Studies in Roman Commerce with the East," ANRW 2.9.2 (1978) 647 and n. 898–899; and Young, op.cit. (above, n. 4) 37 quoting Philostr. VA 6.2, and Juvenal 11.124, for a list of the products crossing the border at this point; for the boats used by the Ethiopians, see Plin. NH 5.10.59.
R. Morkot, "The Darb el-Arbain, the Kharga Oasis and its Forts, and Other Desert Routes," in D.M. Bailey (ed.), Archaeological Research in Roman Egypt. The Proceedings of the Seventeeth Classical Colloquium of the Department of Greek and Roman Antiquities, British Museum, Held on 1–4 December, 1993. JRA Suppl., 19 (Ann Arbor 1996) 83; For the deployment of Roman troops in Nubia, see now V.A. Maxfield, "The Deployment of the Roman auxilia in Upper Egypt and the Eastern Desert during the Principate," in G. Alföldy et al. (eds.), Kaiser, Heer und Gesellschaft in der römischen Kaiserzeit. Gedenkschrift für Eric Birley. Heidelberger althistorische Beiträge und epigraphische Studien, 31 (Stuttgart 2000) 410–418.
P.Customs, p. 5. The Syene customs receipts are: WO II 35 (= P.Customs III 7; AD 89); SB V 7526 (= P.Customs III 8; AD 89); WO II 43 (= P.Customs III 10; AD 95/6); SB V 7579 (= P.Customs III 11; AD 100) and 7580 (= P.Customs III 19; AD 128); WO II 150 (= P.Customs III 20; AD 129) and O.Stras. 250 (= P.Customs III 22; AD 138–161); to which we can now add: P.Brook. 64 (AD 129); O.Eleph.DAIK 56 (AD 106/7–114); 57 (AD 117–138); 58 (AD 144); 60 (I–II c. AD) and 61 (I–II c. AD).
O.Douch IV 384 and 385 with the commentaries of their editor, and G. Wagner, "Le camp romain de Doush (Oasis de Khargeh-Egypt)," in Studien zu den Militärgrenzen Roms, III. 13. Internationaler Limeskongress, Aalen 1983. Vorträge. Forschungen und Berichte zur Vor- und Frühgeschichte in Baden-Würtemberg, 20 (Stuttgart 1986) 671–672.
Schedia: Strabo 17.1.16; SB VI 9365, 11 (AD 259) with R. Bogaert, "Les opérations des banques de l'Égypte romaine," AncSoc 30 (2000) 201 and n. 228; SB VI 9210, 3 (II–III c. AD); P.Fay. 104.21 (late III c. AD). Iouliopolis-Nikopolis: SB VI 9210.4, and P.Fay. 104.12–13. For the interchange of "Juliopolis" and Nicopolis" in first-century AD usage, see A.E. Hanson, "Juliopolis, Nicopolis, and the Roman Camp," ZPE 37 (1980) 241–254.
17.1.16: τὸ τελώνιον τῶν ἄνωθεν καταγομένων καὶ ἀναγομένων; a point which has not escaped the attention of the editor of P.Diog. 13 and 14, pp. 109–110. Note also that Strabo does not mention in his text the customs-house at Iouliopolis-Nikopolis.
On these three taxes, see P.Customs pp. 19–25. On the division of Egypt into three or four epistrategies in Roman time, see J.D. Thomas, The Epistrategos in Ptolemaic and Roman Egypt. Part II: The Roman Epistrategos. Pap.Col. 6 (Opladen 1982) 15–39.
The tax is attested in the following villages (all located along the borders of the Arsinoite Nome): Dionysias, Sokno-paiou Nesos, Philopator/Theogenes (between Soknopaiou Nesos and Karanis), Karanis, Bakchias, Philadelphia, Tebtunis and Kaine (division of Polemon, probably on the fringe of the desert).
The tax is attested in the following villages: Dionysias, Soknopaiou Nesos, Philopator/Theogenes, Karanis, Bakchias, Philadelphia and Anoubias (south of the division of Themistos?), as well as in the town of Memphis.
A πεντηκοστή is attested in the following towns and villages: Thebes, Memnoneia, Hermonthis, Syene and Toyt, cf. P.Customs, pp. 192–194 (List III) to which one must now add P.Brook. 64 (AD 129), O.Eleph.DAIK 56 (AD 106/7–114), 57 (AD 117–138), 58 (AD 144), 60 and 61 (both of the I–II c.).
D.W. Rathbone, "Koptos the 'emporion': Economy and Society, I–III AD" in Autour de Coptos. Actes du colloque organisé au musée des Beaux-Arts de Lyon (17–18 mars 2000). Topoi Suppl., 3 (Lyon 2002) 184 and n. 22.