It has been suggested in the past that the Romano-British sites of the North-Kent Marshes, on mounds submerged in the alluvial silts of the Thames and Medway, have much in common with the mounds known as 'terpen' in the salt marshes of the Groningen and Friesland Provinces of Northern Holland. These Northern salt marshes were settled about 500 BC, by people managing to resist flooding with the raising of small artificial mounds of sods. These 'terpen' were raised when necessary and also extended after the Roman period for crop farming when the old marsh surface became wet.
The Dutch State Archaeological Service has surveyed and mapped these mounds which vary in height above the surrounding marsh from a slight rise in the level to some 8 meters above the marsh level at Hoogebeintorum village. They vary in size from large 'terpen' which sometimes support a village to smaller ones which may have had a small farmstead or barn on them. Some of these mounds can be dated to the Iron Age, showing a continuous occupation right through the Roman period (although the Romans never occupied the area, trade relationships existed with the Roman Provinces south of the Rhine), into Medieval times and on to the present day.
These 'terpen' sites have been mapped in great detail and given various productive status, although few of the sites have been excavated. Compare this with the mounds on the Thames and Medway marshes which range in date from the Roman period to the early Medieval period but do not enjoy any such protection and have not even the distinction of being recorded. There are no doubt problems in surveying the Roman mounds on the North Kent marshes which are partly submerged in the post-Roman silts and only exist as slight swellings in the marshes. I remember visiting some mounds of presumably Medieval date (known by the dialect word of "coterells" on Sheppey), possibly the result of salt-making and large enough to have supported a small village, near Great Bells Farm, Eastchurch, Sheppey, some thirty years ago. Possibly these mounds have long since been bulldozed away. However, even these coterells and any mounds with the odd looker's hut on it need recording as possibly their origins are medieval and much earlier.
Certainly, if field surveys are difficult, excavation of the sites is more so as the water-logged state of the marshes makes for severe excavation problems, while in the rare dry summer the marsh clays dry out into an intractable state, and excavation is only possible by keeping sites damp under polythene sheeting as the excavations at Cooling, some twenty-five years ago proved.
In May 1993, the opportunity arose to visit the 'Terp of Wijnaldum' (No. 73), near Harlingen, North Friesland, where the Archaeological Centre of the State University of Groningen was excavating the 'terp' which was possibly the home of a Medieval Friesian 'King'. The excavations at Wijnaldum were on a large 'terpen' site which had supported a modern farm that was being cleared and rebuilt after a fire. The excavations showed that the site had been occupied from Iron Age times through the Roman period and into the Medieval period. The stratification was extremely difficult to interpret as the top layers were eroded in part (possibly by farming operations) and complicated by dumping of deposits on the edges of the mound. The soil was a sandy clay which dried to a cement-like texture and could only be troweled while damp. This was achieved by keeping the soil covered by plastic sheeting.
An interesting feature was a 'Gruben Haus' which had collapsed into its pit, with later occupation on top. Also pointed out were the turf walls of a hut, which showed after sympathetic troweling, the individual turves which made up its construction. Traces of iron-working and traces of gold were also found, but salt-making was not thought to be one of the activities. The second-century pottery from the site was very crude, with fingertip impressions along the rim, reminiscent of Iron Age pottery.
Modestly, the Dutch said they knew very little about the archaeology of the 'terpen' sites, as the excavations at Wijnaldum were the first since the 1960s. Although Friesland abounds with 'terpen' sites and many raised 'platforms' could be seen in the marsh surrounding Wijnaldum, even the local church sits on a gentle swelling above the general marsh level.
It would have been interesting, although time did not permit, to have compared our Ordnance Datum with the Dutch NAP (New Amsterdam Level, Amsterdam being the zero for the European leveling network) and then compared our Thames and Medway inner marsh levels with the Friesian marshes, although I had the feeling that NAP would equate with our Average High Water, Mean Ordinary Tides.
I am indebted to Mr. S. W. Jager of the State Archaeological Service (Rijsdienst voor het Oudheidkundig Bodemonderzoek) and Mr. Piet van Wilgen who made the visit to the Friesian 'Terpen' mounds possible.
Alec Miles
