Logo for Godmanchester Web site, UK  
The Causeway Godmanchester (c) Stuart Bond
Godmanchester Roman History - The Mansio
Max Cashback

Andrew Selkirk, Editor of Current Archaeology magazine has very kindly allowed us to reproduce an article on Godmanchester written by Michael Green and originally published in 1969. (Current Archaeology, number 16, September 1969 pp133-138)

If you like Time Team and our history in general, why not take time to read this authoritative text on the Roman History of the town, visit the Porch Museum in Godmanchester and also pop over to the Current Archaeology web site. The magazine is modern and very readable for anyone with a general interest in our history and we can thoroughly recommend subscribing to it!!

In addition to this detailed article on the Mansio, check-out Ben Robinson's review of 2000 years of habitation in Godmanchester. (Note: Images here are large as resolution is needed for legible text - but should be downloaded before you read that far!) I am keen to expand the historical information on this site, are you interested, then drop me a line.

Since the time of the 16th century antiquary, Leland, the site of the Roman settlement at Godmanchester, Huntingdonshire, has been known to be the centre of the modern town, roughly bounded by a ring of roads. The site has become increasingly built-up in recent years, but since 1949 excavations have regularly taken place with the ultimate aim of recovering something of the plan and history of the Roman town before the opportunity is lost.

Work began on the bath-house in Pinfold Lane as a result of discoveries made during drainage operations, and has subsequently extended over several acres on the west side of the town. In 1968 and again this year [1969] the work has been carried out under the auspices of the MOPBW. In 1956 the Roman town wall and probably part of the west gate were discovered, and were followed up by the sectioning of the defences in 1957 near the Causeway. Further work was carried out on the town wall by Mr Charles Green in 1959 and 1961 for the Ministry of Works, during the course of which the south gate was found. In 1964 and the following year excavations near the west gate uncovered a series of boundary ditches and huts sites. Just outside the town at Rectory Farm, Dr William Frend followed by Mr Roger Jacobi have been excavating a villa site.

The importance of Godmanchester during the Roman period was primarily geographical, since it controlled the river crossing of Ermine Street, the great trunk road to the north, and the junctions of two minor roads from Cambridge and Sandy. The choice of the site appears to have been determined by a low spur of gravel jutting north-westwards from the higher ground to the south. The gravel carried a dry route to within about half a mile of the river bend, where fording and later bridging was possible. The town was typical of a class of settlement which served the main roads as posting stations and administrative centres, and in many cases appear to have grown out of the civil settlements around early military forts. Neither its Roman name nor its municipal status are known for certain, but Godmanchester has in recent years been tentatively identified with the Drovigutum of the Ravenna Cosmography.


The earliest Roman occupation appears, as might be expected, to be of military character. Beneath the mansio a loft wide ditch with a box gutter has been traced for 60ft east-west. The dating of this feature is difficult since the clean fill of the 4ft deep ditch has produced no dating evidence. Following what must have been either a Claudian or Neronian fort, the early Roman settlement grew up along the Roman roads, although excavation indicates that occupation did not begin in earnest until the pacification of the area after the Boudiccan rebellion of AD 60.

The civil settlement began with the setting out of a regular grid of plots 139ft 5in square (135 Roman ft or 27 Roman paces square), laid out in alignment with the central section of Ermine Street through the town. The intersections of the plots were marked by massive posts (termini), 1ft 4in square. The areas of ground this marked out were somewhat larger than the Mediterranean actus, 116.05ft square (120 Roman ft square), but units of land measurement do not appear to have been uniform everywhere, particularly for subdivisions within the centuria. As the town was settled the plots were gradually developed, either singly or in groups, and the boundaries (limites intercisivi) marked out with fences or ditches.

Entrances appear to have been in the corners of the plots with appraoch roads running along the sides of the enclosures. In a larger settlement this might well have developed into a regular grid of insulae, although this does not appear to have happened in Godmanchester. In the areas excavated the plots appear to have functioned as farmyard enclosures as evidenced by corn-drying ovens and threshing floors. Around the perimeter of the compounds were found rectangular timber buildings varying in size from 17ft by 21ft to 40ft or more in length.The buildings were timber framed with walls of cob construction reinforced by stakes, and hard earth floors and thatch or tiled roofs.


In the centre of the town, but set some distance back from Ermine Street, some two or more plots of ground were cleared of buildings and levelled around hte year AD 120 preparatory to erecting a complex of masonry buildings. The size and character of these structures suggest it was a mansio or rest-house for official travellers using the imperial post service (curus publicus). Establishments of this type also had to provide facilities for changing relays of horses used for the carriages and gigs of the couriers, and for the storage of government goods in transit. In the later Roman empire mansiones tended to be used as police posts and collecting points for corn tax. Use of these inns was strictly controlled by travel warrants, since the service tended to be heavily overloaded abd enormously expensive. In the early Roman period the capital costs and running expenses of the service had to be met by the local government authorities, who also had the duty of organising the system. In the reign of emperor Hadrian (AD 117-138) the state took over their responsibilities.

This reorganisation of the postal system at a period when Godmanchester mansio was being built may be reflected in certain early structural alterations. A bath-house and the southern range of the inn were laid out and some of the foundation trenches dug. A considerable delay followed, in the course of which rubbish pits were dug amongst the foundations. When work was resumed, the plan of the bath-house, which haf formerly been of the 'Reihentyp', was turned inside out with the construction of a second entrance at the north end and the interior adapted to take two separate bath systems, rooms A-E, J-L and I-5 respectively. It is possible that, as a result of some sort of deal with the central government, the town was allowed to keep the southern end for its own use.

As eventually completed the mansio complex was of considerable size, with an overall length of some 300ft. The inn was entered from the north by a gravel service road, which passed through the main gate (1) into a stable yard (3). On either side were two blocks which were probably stables (6 and 9), each holding six or more horses and provided with tack rooms (5 and 8) flanking the yard. Fodder and straw were probably stored in lofts over the stables. Carts and carriages were probably parked overnight in the stable yard, which perhaps explains the off-centre position of the gateway.

Wooden bollards lined the sides of the yard to prevent vehicles from damaging the walls. Extra vehicle accommodation was provided by a shed (2) outside the main gate. At the end of the stable yard was a corridor (4) linking the return wings of the north block. Here the large ground floor rooms (7 and 10) were unpaved and, although there is no definite evidence, it is possible that they were used for storage accommodation by the goods service. There was probably access from the corridor into the inner court, but there was also an external corridor (11) on the east side giving direct access to the stables from one of the bedroom staircases.

The inner court (13) had ranges of bedrooms (14, 16-19, 20 and 22-24 on the ground floor) with the main reception rooms (28-31) along the south front. Each bedroom wing was served by a separate staircase (15 and 21) at its northern end. Only the dinning room (28) and its adjoining kitchen (26) can be identified with any confidence amongst the public rooms. Outside the back door of the kitchen were found large numbers of rubbish pits full of domestic material, whose study is giving a detailed picture of the local pottery industry and catering arrangements of the establishment. During the later 3rd century a timber building (R-10) was erected in the kitchen midden area, and garbage was diverted to the abandoned bath-house leat. Projecting from the southern bloack were two small rooms (27 and 32) which were later additions, and whose character suggests that they were granaries.

South of the inn lay the bath-house. The southern section of the building was served by a gravel road from Ermine Street. From a colonaded portico (L) th bather entered a big changing room (A) with attached lavatory (J2) and lobby (J1). A cold room (B) with a plunge bath (C) opened out of the changing room, and beyond them was a hot room (D) heated by a hypercaust with the heat carried over the vaulted roof by hollow voussoir blocks. The arrangement of the northern section was similar, but with cold rooms (1-3) compressed into a smaller space. An interesting discover was that the changing room (2) at this end had a boarded oak floor whose impression still survived. The hot rooms (4 and 5) at the northern end were larger and had a hot plunge bath added at some later period. The stoke-holes for both sets of baths were probably in the area lettered K. Water for the baths was conveyed to the building by means of an open leat. This also served as a drain for the waste from the cold bath (C) which was emptied into the leat below the main collecting point. There was also a stone lined well for drinking water at the south corner of the inn.

In its hey-day the Inn may have looked like this. Courtesy "Cambridgeshire, Huntingdon and Peterborough Life.

The inn and bath-house were substantially built with 2ft 6in wide masonry walls plastered internally and externally, and a half-timbered construction above the ground floor. The imprint of te timbers has been found on the back of various sections of plaster. The decorative schemes were simple in character with broad bands of yellow, green, pink and red on a white ground, and panels picked out in black. The external surfaces were painted to imitate marble, but Hopton wood stone and Purbeck marble were used for lining the cold plunge bath. All the corridors had brick tessellated floors, and most of the pavements in the principal rooms were of mosaic apparently carried out by a West Country firm. Some of the firms engaged on the building probably had government contracts, including the heating engineers whose flue tiles (Lowther's Group 5, die 16) have been found on the sites of comparable public buildings. Glazed windows only seem to occur in the bath-house, where their positions have been traced by the clusters of broken glass round the outside of the building. The roofs were of tiled and Collyweston slate.


During the 2nd and 3rd centuries various minor additions and alterations were made to the structure of the mansio. During the 3rd century Godmanchester was given town walls, but these did not prevent a major disaster overtaking the place in AD 296. The mansio was gutted by a fierce fire, which taken with other factors suggests the work of raiders, possibly Saxons. The bones of a disarticulated human skeleton found on the Roman ground surface near the kitchens had been gnawed by dogs or wolves. A hoard of jewellery and coins was hurriedly buried in a sump (pit R 4) behind the bath-building and never reclaimed. The hoard represents the contents of a woman's jewel box and included a gold pendant, silver and bronze finger rings, a glass necklace, intaglios, bone hair pins and some 60 coins, mostly radiates.

The archaeological evidence indicates that the site lay derelict until the later 4th century. The floors and walls were then systematically dismantled and the building materials removed for use elsewhere. The northern end of the bath-house was rebuilt at this time in a very shoddy manner, and together with the site of the mansio, was enclosed in a ditched compound. The dating evidence for this work suggests that it took place at part of the Theodosion reorganisation of the province in AD369. The pottery of this phase of occupation is of great interest since it includes sub-Roman and early Saxon forms. Both coinage and pottery indicate that occupation continued until the end of the Roamn period, during the course of which the building was remodelled and the boundary ditch re-cut. If indeed this group of late 4th century structures represent an official establishment, it seems unlikely it would have survived the collapse of Roman rule in the early 5th century.

© 2000 Current Archaeology - Permission granted to Stuart Bond to reproduce.
GMC small logo