(This paper is an extended English version of the publication 'Een Romeinse Europoort aan de Schelde', Westerheem 2016, Vol. 65, August, page 185-195. The text of the publication in Dutch can be found here.
English version first published on this website 13 October 2016).



The Roman Scheldt harbour:
a gateway to the Low Countries


Antony Kropff  1

Since the altars for Nehalennia were recovered from the Scheldt some forty years ago, we know that a Roman maritime harbour was situated in the estuary of this river.
Recently, the Roman harbour of Voorburg-Arentsburg (Forum Hadriani) has also been acknowledged as a transhipment harbour for overseas trade, particularly on Britain.
Did the Dutch River Area encompass more than one maritime harbour?


We will discuss overseas transport from this area between ca. 150 and 250 AD. Which raw materials and products were shipped and where did transhipment from river barges onto seagoing ships take place?

Export from the Dutch River Area

Export from the region itself

Did the Dutch River area itself supply a surplus of products, available for export? According to a recent study, the Lower Rhine delta would have produced enough surplus to provision the army units in the area with -among other things- barley, emmer wheat and timber. About half of the army’s demand for cereals could be met. Cattle, wheat for bread and spelt had to be imported.1
The Lower Rhine delta would in our view not have contributed to export. In fact, the same applies to the eastern part of the Dutch River Area: any agricultural surplus here would undoubtedly have been used to furnish deficiencies in the western part of the area.

The region of the Scheldt evidently produced a surplus above the level of self-sufficiency. In this survey we will consider the part of the Scheldt region in the present day Dutch province of Zealand as a part of the Dutch River Area, although strictly speaking it was not: the river formed a part of the border between Germania Inferior and Gallia Belgica.
In the Civitas Menapiorum (for the most part in Gallia Belgica) to which Zealand belonged, we note the production of allec and salt.2  Koudekerke (Zealand) and Zeebrugge (West-Flanders) belonged to the production area of salt. Salt was not only produced for local use, as is illustrated by the Nehalennia altars and by inscriptions in Rimini commemorating a veteran centurion, while active probably responsible for the salt supply to the legionary fortress at Novaesum (Neuss) on the Rhine. The memorial stone was dedicated by two salinatores (salt producers) of the Morini and the Menapii respectively.3
In this area, with its limited hours of sunshine and wet summers, salinas could not be used for direct salt production. Salinas however could be used to increase the salt content of seawater to extract salt from the obtained brine by heating the fluid in terracotta containers in ovens (the briquetage technique, fig. 1).

briquetage

Fig. 1. Salt production with briquetage technique (photo: Comité des Salines)

Extensive salt production can be concluded from large amounts of (sherds of) salt containers and other briquetage material near production sites. Large-scale production is found in Zeebrugge, where the salinas covered 1.500 square meters in total, with ovens and briquetage material in some distance4 and in Leffinge, with its double row of 15 large salt ovens each.5
The four salt merchants dedicating an altar to Nehalennia and their colleagues in the trade probably exported salt to Britannia, among other destinations.6 In theory, the merchants could also have imported salt from Britannia, as Essex produced salt as well. However, this small-scale production ended in the second century A.D. An embargo on the export of British salt followed in the third century.7
We can conclude that salt exports from the Scheldt port are plausible, as is the export of allec. Large quantities of mussel- and cockle shells were found on sites also including an oven8 and the rim of a dolium found at Aardenburg (Zealand) carries a graffiti, indicating a content of 300 litres of allec.9

Menapian ham was an important export, as contemporary sources show a general preference for this product.10 The Nervian region produced a considerable surplus of wool and one of the woollen products can be found in the price edict of Diocletian: the Nervian hooded cloak, one of the most expensive cloaks in the edict.11 We can assume production of other woollen clothing and also of linens.
The Morini, also in the Scheldt area, were known for their export of geese, sought after even in Rome.12 Also, the price edict shows that goose feathers and –down were in demand as stuffing of mattresses and cushions.13

We can conclude that even if exports from the surrounding areas would have been absent, the Scheldt area itself produced enough goods to generate a lively trade which would have warranted a port and transhipment facilities. And what is more, later we will show that products from other areas were also shipped from the Scheldt harbour.

Transit

The Rhine and Meuse brought transit goods from the areas bordering these rivers to the Dutch River Area. Some of these exported products can be traced archaeologically in Britain: pottery, glass, terracotta and pipe clay figurines, wine in barrels from the Rhineland and other areas and lava querns from Mayen on the Middle Rhine.14
As far as the pottery is concerned, terra sigillata was imported from the Rhineland after 150 A.D., but in a limited quantity compared to the previous imports of sigillata from Central Gaul, where Lezoux had been an important production centre. Lower Rhineland colour coated ware and Trier black slipped ware (‘Moselkeramik’) were imported in significant quantities after 150 A.D.
The Rhenish wine in barrels came to Britannia in a limited quantity as is attested by the barrels dug in to serve as water wells.

From the villa-zone in Belgica, cereals were shipped by way of the Dutch River area. The wreck of the barge Woerden 1, found in the Lower Rhine, still contained residues of wheat mixed with weed seeds specific for that zone.15 We cannot ascertain whether the wheat from the villa-zone was also shipped to Britannia. To supply the Roman army units in Northern Britannia cereals and other foods were supplied, primarily from the south of the province. Continental import has been suggested, possibly by way of the Dutch River Area, but the latter has been called into question.16


Nehalennia and the Scheldt Harbour

Where was freight transhipped onto seagoing vessels? We know that an important harbour for maritime transport was situated in the Roman Scheldt, the present Oosterschelde (East-Scheldt) which is shown by many Nehalennia altars, recovered from the Scheldt since 1970.
The iconography and inscriptions of the altars show that the goddess Nehalennia bestowed protection to seafarers and sea trade. Her cult seems to have been limited to the Scheldt area, the location of the altars suggest temples at Domburg and Colijnsplaat in the present day Dutch province of Zealand.
A limited number of altars can be dated by means of an Imperial consulate in the inscription; these altars span the period of 188-227 A.D. This period is consistent with the first peak formation in the histogram (fig. 2) of the coins found at Domburg.17

histogram Domburg

Fig. 2. Histogram coin loss Domburg (graph by the author)

The votive altars (fig. 3) were dedicated predominantly by the merchants shipping from the harbour, to redeem a pledge to the goddess. The temples would have been associated with a harbour and a trade settlement, both of which have not yet been located.

Nehalennia altaar

Fig. 3. Nehalennia-altar (photo: RMO Leiden)

This association is also shown by the potsherds, collected methodically and without selection by Mrs Gerhardt on the Domburg beach during a period of eight years. Among the finds are 335 sherds from the Roman period, dating between 100 and 250 AD. Of these sherds, 29% is terra sigillata, in this context suggesting a settlement with a trade function.18
Most likely, the harbour was situated near Colijnsplaat, on the south bank of the Scheldt estuary. This type of setting, guarded by military bases (at Oostkapelle-Oranjezon we may assume a naval base, at Westerschouwen-Roompot an army base) was favoured by the Romans for a seaport location.19 Well-known examples of Roman estuary ports are Londinium (London) on the Thames and Ratiatum (Rezé) on the Loire.

The altar inscriptions show that merchants from the north-western part of the Empire shipped several products from the Scheldt port. We can identify merchants in allec, a type of fish sauce (three merchants, one of them from Trier), salt (four merchants, among them three from Cologne), pottery and terracotta figurines (one merchant, probably from Cologne, trading on Britannia) and wine (two merchants, one of them from Augst in Switzerland). Also, a number of altars were dedicated by merchants simply identifying themselves as negotiator; four of them stating trade on Britannia. Alongside the merchants’ altars, altars dedicated by three ship-owners and a ship-captain were found.
We already saw Trier, Cologne and Augst as domicile and the altars also mention Nijmegen, Dormagen, Ganuenta and the regions of Besançon and Rouen.20 Sixteen altars were dedicated by merchants belonging to the Civitas Tungrorum around present day Tongeren, judging by their names and the carboniferous limestone used for the altars, which was won near the Meuse between Namur and Huy. The town of Tongeren is situated on the river Jeker, discharging into the Meuse.
The altars demonstrate that freight from the Rhine- and Meuse area was also transhipped in the Scheldt harbour.

A maritime port in Voorburg?

Excavations during 2007/2008 revealed a Roman harbour. The participating archaeologists proposed a function of transhipment harbour for maritime trade, particularly on Britannia. We will try to find out whether or not this Roman harbour on the Fossa Corbulonis between Rhine and Helinium could have functioned in a maritime context, just like the harbour on the Scheldt.
The Voorburg harbour (fig. 4) was constructed around 160 A.D. by dredging a natural gully and additionally by constructing a quay.21

Haven Voorburg

Fig. 4. Excavation harbour Voorburg-Arentsburg 2007-2008 (photo: Leiden University)

Supplying the army

The research team proposed that one of the functions of the harbour was to supply the small army posts in the coastal area. The pottery shows a military accent: plates (‘legionary ware’), colour coated pitchers and beakers typically used by soldiers. As these aspects of the pottery spectrum in the harbour resemble the characteristics of the pottery in the military posts in the area22 (Ockenburgh, The Hague-Scheveningseweg) and similarities can also be noted in the coin finds of the Voorburg harbour on the one hand and the military posts on the other23, this supply function is plausible. However, it doesn’t seem plausible that the sizeable harbour (circa 100 x 40 m.) would have been constructed just to supply a limited number of small military posts.

City wall?

We propose the hypotheses that the port was initially constructed circa 160 A.D. to harbour and unload ships carrying stones and other building material necessary to build the city wall, erected  around 170 A.D.24 Remarkably, this possible link was not considered in the excavation report and analyses. A relation between a harbour construction and the erection of a city wall has been demonstrated several times, for instance for Xanten and Cologne.25 In this context it is worth noting the discovery of building material found in the harbour, in particular a large tuff building block with dimensions (75 x 38 x 24 cm) known from the castellum of Albaniana (Alphen aan den Rijn) and a semi-manufactured column drum of 2,20 m. high.26

Transhipment port

In the report and analyses of the Voorburg port a function of transhipment port for maritime transport and trade was suggested, particularly to Britannia.27 Evidence mentioned is an open connection to the North Sea and similar pottery imports in the harbour of Voorburg and the London harbour of St Magnus House. Also, both these harbours were thought to be navigable by coasters, considering the operational depth of both ports.
However, first of all the operational periods of both harbours do not match. The Voorburg port, constructed around 160 A.D., silted up from circa 230 onward, notwithstanding a second dredging campaign around 210 A.D.28
The earliest quay construction in the harbour of London-St Magnus House which can be related to commercial harbour activity was built between 225 and 245 A.D.29 This limited convergence of operational periods would have confined connectivity of the two harbours.

Coasters

Near the London harbour, the Blackfriars I was found. Ships of this class, the report on the Voorburg excavation suggests, could also have moored in the latter port. This is questionable. The Blackfriars I was 6,5 m. wide and a hydrostatical analyses showed that it drew 1,5 m. of water under full freight.30 The depth of the Voorburg-Arentsburg harbour was probably -depending on tide and water level- between 0,5 m. and 1,5 m.31 Taking into account a keel clearance of 20 cm.32 we can conclude that a ship like the Blackfriars I could only moor in Voorburg under optimal conditions and even than with a limited freight. These conditions do not suggest a maritime function of the harbour. Also, ships of the Blackfriars I class probably could not pass through the northern part of the Fossa Corbulonis, the connection of the Voorburg harbour to the Rhine. The width of 12-15 m. and the average depth of 1,4 m. of the Fossa are no impediment, but near Leidschendam a tapering to circa 4,5 m. has been detected, too narrow for a Blackfriars I type ship. The narrowing in the Fossa was probably the entrance to a portage or a rolling bridge on the spot, necessary on account of the dissimilar water levels in the Rhine and the Helinium estuary.33 Hypothetically, the Voorburg harbour might have been accessible through the southern part of the canal, giving access to the Helinium. However, this hypothesis cannot be verified because data about this part of the Fossa are not available.

Similar import pottery?

Similarity in pottery imports found in the harbours of London-St. Magnus House and Voorburg-Arentsburg were also thought to support the maritime function of the Voorburg port. And yet, the ensemble of terra sigillata, an important continental pottery import, shows a difference.
In London, the share of the terra sigillata from Central Gaul (primarily from Lezoux) is notable. This terra sigillata dominates in the provinces of Britannia and Gallia but is relatively scarce in the Rhine provinces.34 In the Voorburg harbour, the share is only 3,3%.35 In Voorburg, the Eastern Gaul terra sigillata from Trier has a large share. Trier sigillata has also been found in the London harbour, but this type of pottery would not have been shipped from Voorburg because the production period is dissimilar. Of the decorated sigillata from workshops in Trier found in the Voorburg harbour, more than 70% was produced in the second century, in part possibly in the first decade of the third century.36 A large share of this type of sigillata in the London harbour dates from the third century. The stamped terra sigillata from Trier which was found in the London harbour also shows a later accent than this type of pottery found in Voorburg.37
The terra sigillata as such provokes more questions: in the London harbour, this fine ware encompasses no less than 63% of the total pottery spectrum.38 The harbour of Voorburg-Arentsburg has a sigillata share of a mere 6%39, quite normal for a rural settlement but too low for a trading town. Goedereede-Oude Oostdijk40, a trading post, yields a larger share of terra sigillata and the same applies to the potsherds from the Domburg beach.41

To summarize, the terra sigillata from the Voorburg harbour on the one hand and the sigillata of the London harbour on the other do not show enough similarity to support trade connections between the two ports.
Taking all aspects into account, we can conclude that Voorburg-Arentsburg in all probability did not function as a maritime harbour.

Rivers as trade routes

Voorburg-Arentsburg did not function as a transhipment harbour for the Rhine and Meuse. Still, it seems to be obvious that export products from the Rhine and the Meuse area would have been shipped from ports in or near the mouths of these rivers. But can any maritime ports be detected or assumed?

Rhine

In the Dutch River Area many ships from the Roman period were found. Especially the present day Leidsche Rijn and the Oude Rijn (forming the lower course of the Roman Rhine) are actual ship cemeteries. We now know of nineteen ships, among which are eleven large flat-bottomed barges, mainly from De Meern, Woerden and Zwammerdam; eight of these ships were in operation during the second half of the second century A.D.42 These flat-bottomed ships were suited for bulk transport. The ship De Meern 4 had carried basalt, the Zwammerdam 2 still contained residues of schist and in the plank seams of the Zwammerdam 4 brick waste was found. On the ship from Druten on the Waal stone grit and parts of tegulae were found.
As quite a number of ships date from the second half of the second century, we may assume a connection between this bulk transports and the fact that many castella were partly rebuilt in stone during this period.43 Stone was brought in from the Eiffel and roofing-tiles from the Holdeurn near Nijmegen. We can conclude that these ships functioned in a military context. This is also illustrated by the way these ships were disposed of: sunk deliberately to form a bulkhead or sheeting of an embankment or quay near a castellum.
Two ships deserve special attention. The ship De Meern 1 (fig.5) had not been sunk deliberately but was wrecked during an accident. Tools and militaria found in the ship suggest it had carried a team of carpenters, or even soldiers, maintaining army installations on the river.45

De Meern 1 opgraving

Fig. 5. Excavation ship De Meern 1 (photo: Ronald Correljé)

We mentioned the residues of grain aboard the Woerden 145 which seems to suggest a civil trade transport, but this ship was also deliberately sunk to add strength to the river embankment near the Woerden castellum.
The transformation of the Rhine castella from wood to stone during the second half of the second century would have caused a shortage of shipping capacity. The construction of the stone castellum Woerden phase IV alone would have required 180 shiploads of basalt and tuff and an additional 20 shiploads of roof-tiling.46

In short, no civil trade vessels were found in the lower course of the Roman Rhine. This part of the river was mainly used as a military infrastructure, to carry surplus and building material for the castella and other military installations on the Rhine limes. A transhipment port for maritime transports has not been found. The question arises whether this part of the Lower Rhine was appropriate for a maritime harbour. The Rhine did not end in an estuary large enough to accommodate the required cover by a naval base and army posts. Also, it has been suggested that the mouth of the Rhine was susceptible to silting, as was in fact the Helinium.47

kaartje sites en rivieren

Fig. 6. Map of discussed sites, rivers, kanals, etc (source: www.romeinsekust.nl)

The freight transports between the Rhineland and Britain would probably have followed the Rhine until the fork of the Waal near the present day Dutch town of Spijk, and subsequently would have reached the Scheldt harbour via the Waal and watercourses we will discuss below.

Regarding the Rhine, the period of 150-250 A.D. did not yield any historical, epigraphical or archaeological evidence to suggest transhipment of goods from the mouth of that river.

Meuse / Helinium

Excavations have provided data on towns and settlements on the Meuse, for instance on Maastricht and Cuijk. Still, the Meuse presents a knowledge lacuna.48 Chemical disintegration in the province of Noord-Brabant and dredging for sand and pebbles in Middle-Limburg have destroyed archaeological remains in the subsoil. The lower part of the Meuse, the Helinium, has suffered from erosion, harbour construction and from dredging.

Objects found during the construction of the Benelux harbour in Rotterdam and during trailing hopper dredging in the New Waterway (‘Nieuwe Waterweg’) suggest a military context for the banks of the Helinium.49
No traces of a harbour were found. Further research into the possible function of the Meuse as a trade- and transport route is currently being considered.50

The Scheldt, a Roman Euro-port?

We will now return to the transhipment port near Colijnsplaat on the Scheldt. Products from the Scheldt area were shipped, but the Nehalennia altars also indicate transhipment of trade products from the Rhine and the Meuse.


From Rhine and Meuse to the Scheldt

In what way were goods from the Rhineland and Meuse area transported to the Scheldt harbour? Flat-bottomed ships could not navigate from the mouth of the Rhine or the Helinium to the Scheldt by coasting via the North Sea, as this type of ship could not withstand swell and waves.51
An in part man-made canal (the Fossa Corbulonis) connected Rhine and Helinium. A waterway between Helinium and the Scheldt has been taken for granted in the past, as precursor of the medieval Striene.52 A possible Roman predecessor of the Striene has later explicitly been rejected.53 The official paleo-geographical map of the Low Countries for circa 100 A.D. does not show an inland watercourse between Helinium and the Scheldt.54 However, this rejection of an inland waterway needs to be reconsidered. The barge Woerden 1 transported grain mixed with weed seeds from the Belgian löss area. The most plausible route for these transports is from the Scheldt to the Helinium, and then by way of the Fossa Corbulonis to the Rhine.55 Because this ship could not use a sea route, this suggest an inland connection.
In the ships De Meern 1, 4 and 6 oak-wood from Flanders was used. We can assume that these barges were built in the Scheldt region. North Sea navigation being ruled out, the origin and find spot of the ships also strongly suggests an inland waterway between Scheldt and Helinium.56
The last but not least indication is a segment of a man-made canal, possibly Roman, found on a north/south trajectory crossing the Hoekse Waard, which the medieval Striene has also flown through.57

Gateway to Gallia Belgica

The mouth of the Scheldt is, in addition to the guarded estuary, an ideal location for a maritime harbour. The river formed part of the border between Germania Inferior, which has the character of a military territory in the northernmost part, and Gallia Belgica, which province gave scope to a civil culture and a productive civil economy. The Scheldt was not burdened by a military infrastructure, had a less peripheral course than the Rhine and traversed in the lower course a more extensive area with a surplus of products. Two of these products, ham and cloaks, are recorded in the price edict of Diocletian (fig. 7) while products from the Dutch River Area are absent in the text, for as far as products with a geographical label are concerned.

Prijsedict

Fig. 7. Part of price-edict Diocletianus (Pergamon-museum Berlin, photo: Wikipedia)

Other products from the Scheldt area, not specifically mentioned in the text with a Belgica origin but epigraphically or archaeologically verifiable are salt and fish sauce, as discussed above.

We can assume that the Nehalennia temples in Domburg and Colijnsplaat formed a part of a cluster of trade-related facilities like a harbour with wharfs, warehouses and offices of merchant- and shipping agencies.58 The importance and conceivable size of this trade cluster can be deducted from the large number (312) of recorded altars.
Sea transport from present day The Netherlands started out from the Scheldt port, but whether the Channel and the North Sea were crossed nonstop or by way of the port of Boulogne cannot be determined at present.59

Conclusion

During the period 150-250 A.D., export goods produced in or transported through the Dutch River Area were shipped from a port in the Scheldt estuary. This port transhipped products from the Scheldt area itself, like salt, fish sauce, ham and other conserved meat products, wool and woollen clothing. Some of these products were traded all over the Roman Empire. No other region in the Low Countries produced an equally substantial surplus of products suitable for export.
The Scheldt port also handled products shipped by merchants from the Rhine- and Meuse area, as is testified by their Nehalennia altars in Domburg and Colijnsplaat on the Scheldt.
The lower course of the Roman Rhine (the present day Leidsche Rijn and Oude Rijn) was most likely not used intensively for trade transport with an overseas destination. Probable causes were the intensive use of this part of the river as a military infrastructure and the associated substantial army demand for shipping facilities from circa 150 A.D. onwards in relation to an ambitious building campaign: the Rhine castella were rebuilt in stone.
Also, a suitable location for a transhipment port in the mouth of the Rhine would have been lacking. Exports from the Rhine- and possibly the Meuse area were also shipped from the Scheldt harbour. We can now assume that a canal or watercourse connected the Scheldt with the Helinium while the Fossa Corbulonis connected the Helinium and the Lower Rhine (Oude Rijn), allowing transport of goods from Rhine and Meuse to the Scheldt port.

Recently, a harbour was discovered on the Fossa Corbulonis in Voorburg-Arentsburg (Forum Hadriani). This port was hypothesized to function as a transhipment harbour for maritime trade and transport, particularly to Britannia. This hypothesis should be rejected: the presented evidence is unsatisfactory.

The possible role of the Helinium (the lower part of Meuse and Waal) in the overseas trade is less clear, but the Nehalennia altars of traders from the Tungri found in the Scheldt indicate that the harbour at Colijnsplaat also attracted merchants from the Meuse area.
We can conclude that regarding overseas trade, the Scheldt held a special position among the rivers in the Low Countries during the period of 150-250 A.D.


(The author wishes to thank Aagje Feldbrugge, Curator of the Archaeological Collection of the Royal Zeeuwsch Genootschap der Wetenschappen for her contributions).

References

1. Antony Kropff obtained his MA in archaeology at Leiden University and specialized in ancient numismatics. He has published on third century coin circulation, coin hoards, barbarous radiates, Forum Hadriani and the Lower Rhine limes.
1. Van Dinter et al. 2014, 32.
2. Van den Broeke 1996a, 199.
3. De Clerq and Van Dierendonck 2008, 22.
4. Van den Broeke 1996b, 55, note 20.
5. Thoen 2000, 141.
6. De Clerq and Van Dierendonck 2008, 22.
7. Fawn et al., 1990; Saile 2015, 204.
8. Van Beek 1983, 4.
9. De Clerq and van Dierendonck 2008, 23-24.
10. Martialis, Epigrammata XIII, 54; Lauffer 1971, 104, 4.8.
11. Lauffer 1971, 156. 19.44.
12. Plinius, Naturalis Historia X, XXVII.
13. Lauffer 1971, 151, 18.1.a, 260.
14. Morris 2010, 53-81, 102-107.
15. Brouwers, Jansma and Manders 2013, 23.
16. Jones 1996, 213, 214 note 82; Van der Veen 1992, 154-155.
17. Boersma 1967, 70.
18. Besuijen 2009; Van de Vrie 1987, 15-16.
19. Dhaeze 2011, 111-113, 285-286; De Clerq and Van Dierendonck 2008, 29.
20. Stuart and Bogaers 2001, 32-38, 47-48.
21. Driessen and Besselsen 2014, 20-21, 46, 107, 152.
22. Ibidem, 158-159, 469-471.
23. Ibidem, 598.
24. Bink and Franzen et al. 2009, 57.
25. Schäfer 2014, 125;
Wawrzinek 2014, 191 note 2.
26. Driessen and Besselsen 2014, 552-553.
27. Driessen and Besselsen 2014, 159-161.
28. Ibidem, 163-164.
29. Miller, Schofield and Rhodes 1986, 63-64.
30. http://www2.rgzm.de/navis/home/..%5Cships%5Cship020%5CShip020.htm (Retrieved December 17, 2015).
31. Driessen and Besselsen 2014, 161.
32. http://informatie.binnenvaart.nl/algemeen/veiligheid/97-afladen-op-waterstand (Retrieved December 17, 2015).
33. De Kort and Raczynski-Henk 2014, 55.
34. Morris 2010,  60-62, 103.
35. Driessen and Besselsen 2014, table II-1.6.1, p. 474.
36. Ibidem, table II-1.6.3, p. 479.
37. Miller, Schofield and Rhodes 1986, 139, 142-145, fig. 81, 82.
38. Ibidem, 98, fig. 74.
39. Driessen and Besselsen 2014, 328.
40. De Bruin et al. 2012, 117, 136-137.
41. Besuijen 2009, 95-98; De Vrie 1987, 15-16.
42. Brouwers, Jansma and Manders 2013, 14, 16, 22-26.
43. Norde 2006, 55.
44. Van Holk 2006, 37.
45. Haalebos 1997.
46. Blom and Vos 2008, 419.
47. De Bruin et al. 2012, 140-141.
48. Van Enckevort et al. 2005, 4, 19, 29.
49. Bogaers 1974; Haalebos 1974.
50. Stoepker 2006, 127.
51. Jansma, Haneca and Kosian 2014, 493
52. De Clerq and van Dierendonck 2008, 8.
53. Vos and Van Heeringen 1997, 66-67.
54. Vos and De Vries 2013.
55. Bazelmans 2006, 4.
56. Jansma, Haneca and Kosian 2014, 491-493; Brouwers, Jansma and Manders 2013, 19.
57. Kranendonk et al. 2006, 54-57.
58. Besuijen 2008, 25.
59. Trimpe Burger 1997, 28-29.

Literature

Bazelmans, J., 2006: Het schip van Vleuten, een vreemde eend in de bijt, in: R. Oosting en J. van den Akker, (eds), Rivierscheepvaart. Inleidingen gehouden tijdens het negende Glavimans Symposion Vleuten-De Meern, 16 mei 2003, Amersfoort, 3-5.

Beek, B.L. van, 1983: Salinatores and Sigillata: the coastal areas of North Holland and Flanders and their economic differences in the 1st century AD, Helinium 23, 3-12.

Besuijen, G., 2008: Rodanum. A study of the Roman settlement at Aardenburg and its metal finds, Leiden.

Besuijen, G., 2009: De collectie Gerhardt. Romeins en middeleeuws aardewerk van het strand van Domburg, in: Zeeland 18.3, 95-98.

Bink, M., P.F.J. Franzen et al., 2009: Forum Hadriani-Voorburg, Definitief Archeologisch Onderzoek. (BAAC rapport A-05.0125), ’s-Hertogenbosch.

Blom, E. en W.K. Vos, 2008. Woerden- Hoochwoert. De opgravingen van 2002-2004 in het Romeinse castellum Laurium, de vicus en van het schip de ‘Woerden 7’, Amersfoort.

Boersma, J.S., 1967: The Roman coins from the province of Zeeland, Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek 17, Amersfoort, 65-99.

Bogaers, J.E., 1974: Romeinse militairen aan het Helinium, Westerheem 23, 70-78.

Broeke, P.W. van den, 1996a: Southern sea salt in the Low Countries.
A reconnaissance into the Land of the Morini, in: M. Lodewijckx (ed), Archaeological and historical aspects of West-European societies, Leuven.

Broeke, P.W. van den, 1996b: Turfwinning en zoutwinning langs de Noordzeekust, Tijdschrift voor Waterstaatsgeschiedenis 5, 48-59. 

Brouwers, W., E. Jansma en M. Manders, 2013: Romeinse scheepsresten in Nederland, Archeobrief 17:4, 13-27.

Bruin, J. de, G.P.A. Besuijen, H.A.R Siemons en R.J. van Zoolingen, 2012: Goedereede-Oude Oostdijk, Een havenplaats uit de Romeinse tijd, Leiden.

Clerq, W. de en R. van Dierendonck, 2008: Extrema Galliarum. Zeeland en Noordwest-Vlaanderen in het Imperium Romanum, Zeeuws Tijdschrift 58:3-4, 5-34. 

Dhaeze, W., 2011: De Romeinse kustverdediging langs de Noordzee en het Kanaal van 120 tot 410 na Chr, Gent.

Dinter, M. van, L.I. Kooistra, M.K. Dütting, P. van Rijn en C. Cavallo, 2014: Could the local population of the Lower Rhine Delta supply the Roman Army? (Part 2), Journal of Archaeology in the Low Countries, V:1, 5-50.

Driessen, M. en E. Besselsen (eds), 2014: Voorburg-Arentsburg. Een Romeinse havenstad tussen Rijn en Maas (Themata 7), Amsterdam.

Enckevort, H. van, Groot, T. de, Hiddink, H., Vos, W., 2005: De Romeinse tijd in het Midden-Nederlandse Rivierengebied en het Zuidnederlands dekzand en lössgebied, NOaA hoofdstuk 18 (versie 1.0), (www.noaa.nl).

Fawn, A.J., K. A. Evans, I. McMaster and G. M. R. Davies, 1990: The Red Hills of Essex: Salt Making in Antiquity (Colchester Archaeological Group, Annual Bulletin 30), Colchester. 

Haalebos, J.K., 1974: Romeinse rommeltjes uit Rozenburg, Westerheem 23, 78-82.

Haalebos, J.K., 1997: Ein Römisches Getreideschiff in Woerden (NL), in: Jahrbuch des Römisch Germanischen Zentralmuseums, Vol 43, Mainz, 475-509.  

Holk, A.F.L van, 2006: Een Romeinse aak bij Vleuten-De Meern, in: R. Oosting en J. van den Akker, (eds), Rivierscheepvaart. Inleidingen gehouden tijdens het negende Glavimans Symposion Vleuten-De Meern, 16 mei 2003, Amersfoort, 25-38.

Jansma, E., K. Haneca and M. Kosian, 2014: A dendrochronological reassessment of three Roman boats from Utrecht (The Netherlands): evidence of inland navigation between the lower-Scheldt region in Gallia Belgica and the limes of Germania Inferior, in: Journal of Archaeological Science 50, 484-496.  

Jones, M.E., 1996: The end of Roman Britain, Ithaca.

Kort, J.W. de and Y. Raczynski-Henk, 2014: The Fossa Corbulonis between the Rhine and Meuse estuaries in the Western Netherlands, Water History vol. 6-1, 51-71.

Kranendonk, P., P. van der Kroft, J.J. Lanzing en B.H.F.M. Meijlink, 2006: Witte vlekken ingekleurd. Archeologie in het tracé van de HSL-Zuid, Amersfoort, 54-57.

Kropff, A. 2016. Een Romeinse Europoort aan de Schelde. Westerheem 65, 185-195.

Lauffer, S., 1971: Diokletians Preisedikt. Berlin.

Miller, L., J. Schofield and M. Rhodes, 1986: The Roman Quay at St Magnus House, London.

Morris, F.M., 2010. North Sea and Channel Connectivity during the Late Iron Age and Roman Period (175/150 BC – AD 409. (BAR International Series 2157), Oxford.

Norde, E., 2006: Verstening en verscheping; het ontstaan van het Zwammerdammer type, in: R. Oosting en J. van den Akker, (eds), Rivierscheepvaart. Inleidingen gehouden tijdens het negende Glavimans Symposion Vleuten-De Meern, 16 mei 2003, Amersfoort, 39-58.

Saile, T., 2015: Competing on unequal terms: saltworks at the turn of the Christian Era, in R. Brigand and O. Weller (eds.), Archaeology of Salt, approaching an invisible past. Leiden, 199-210.

Schäfer, A., 2014: Köln: Römischer Hafen und Rheinseitige Stadtbefestigungen. Zur Rolle des Römischen Heeres als Bauträger, in: H. Kennecke (ed), Der Rhein als europäische Verkehrsachse. Die Römerzeit. (Bonner Beiträge zu Vor- und Frühgeschichtlichen Archäologie), Bonn, 117-143.

Stoepker, H. (ed), 2006: Synthese en evaluatie van het inventariserend archeologisch onderzoek in de Maaswerken 1998-2005, Maastricht.

Stuart, P. en J.E. Bogaers, 2001: Nehalennia. Römische Steindenkmäler aus der Oosterschelde bei Colijnsplaat. Leiden.

Thoen, H., 2000: Zoutwinning: de teloorgang van een antieke industrie langs de Vlaamse kust, Vlaanderen 49, 139-142. 

Trimpe Burger, J.A., 1997: De Romeinen in Zeeland, Goes.

Veen M. van der, 1992: Crop Husbandry Regimes.
An Archaeobotanical Study of Farming in Northern England: 1000 BC - AD 500, Sheffield.

Vos, P.C. and R.M. van Heeringen, 1997: Holocene geology and occupation history of the province of Zeeland (SW Netherlands), in: Fischer, M.M. (ed), Holocene evolution of Zeeland (SW Netherlands). Mededelingen NIT G-TNO no. 59, Haarlem, 5-109.

Vos, P. & S. de Vries, 2013: 2e generatie palaeogeografische kaarten van Nederland (versie 2.0), Deltares, Utrecht. (Retrieved December 17, 2015 from
www.archeologieinnederland.nl ).  

Vrie, D. van de, 1987: Scherven op het strand. De verspoelde Romeinse nederzettingen in het deltagebied. (Unpublished MA thesis, University Utrecht).

Wawrzinek, C., 2014: In portum navigare: Römische Häfen an Flüssen und Seen, Berlin.