Skip to Content.

Kent Archaeological Review extract

The Discovery of a Saxon 'Grubenhaus' at Keston.
by Brian Philp.

This highly interesting site was discovered in June, 1970, during the seasonal training excavation on the Roman villa and Iron Age sites in Lower Warbank Field, Keston. The whole site is due to be destroyed for Ringway 3.

The training scheme (Reference 1), which started in 1968, is run jointly by the West Kent Group and the London Borough of Bromley. The Roman cemetery, containing monumental tombs and excavated by the Group in 1967-8 (Reference 2), lies about 250 feet to the north. The discovery of the Saxon hut, at the extreme edge of the training-site, was the first positive archaeological evidence of Saxon settlement found in the parish of Keston. Other Saxon huts may await discovery nearby. The Group gratefully acknowledges the support of Mr A G Lockley Cook of Keston; the expert advice on some of the pottery kindly given by Dr J N L Myres and comments on the lead loom-weight supplied by Mrs M U Jones. The supervisors during 1970 were Miss E M Mynott, Mr G B Clewley and Mr D N Broadfoot and the plan has been prepared by Mr R Tedbury.


The hut was first detected as a large rectangular pit cut into the underlying chalk. In detail it measured 13 feet 3 inches east-west and 11 feet 8 inches north-south with its longer axis following the contour of the slope. The sides had been cut almost vertical and the corners at about 90 degrees. The base had been cut horizontally to form a terrace on the side of the hill so that on the uphill (north) side it lay 1 foot 4 inches beneath the surface of the solid chalk, whilst on the downhill side it was only about eight inches. It seems that originally it had been dug to a depth of about 1½ to 2 feet.

DRAWING: Plan of Saxon 'Grubenhaus'.

Drawing Caption: Figure 1. Plan of Saxon "Grubenhaus" hut at L.W.B. Keston. (Drawn by R Tedbury.)

Within the hut there were three large post-holes (see cast-west section) and a series of 29 small stake-holes all cut into the hard chalk floor. Externally, on the cast side, was a rough arc of smaller holes which must have held small posts, but not necessarily connected with the hut. The large post-holes on the west and east sides (A and B) were each 1 foot 8 inches in diameter, 1 foot 8 inches deep and had steep or vertical sides. Both were filled with compact chalk rubble which had been packed around a large upright post. No trace of the posts survived, but circular columns of black-brown loam, nine inches in diameter, clearly indicated the positions of the wooden uprights. Post-hole C was broadly similar, but was only 1 foot 3 inches in diameter and 1 foot 6 inches deep. It too, had contained a wooden post, nine-and-a-half inches in diameter packed around with chalk, but here the post appeared to lean slightly to the east.

The stake-holes in the floor were confined largely to the southern half of the hut. These were mostly circular, about 1-3 inches in diameter and between 2-9 inches deep and probably had been drilled with a metal tool. Four were rectangular in shape, about 2 inches by 2 inches and probably held posts driven into the floor. NO comprehensive pattern emerges, but several straight alignments can be seen and some form of light structure, or structures, seems to be implied.

The filling of the hut consisted of an even black-brown loam, partially disturbed by burrowing animals. There was no trace of any prepared floors, debris from collapsed walls or roof or of a hearth. A total of about 360 potsherds was recovered from the filling of which about 230 were Saxon, about 125 Roman and at least four were Iron Age. There were 90 pieces of Roman tile and about 360 bones. Of the Saxon pottery at least 12 vessels can be identified, of which only three are decorated, all being found within six inches of the hut floor. The fragmentary nature of all these vessels suggests they were deposited as domestic rubbish.

The primary layer also produced a number of other interesting objects. These included a bone comb (Figure 2, Number 1), a lead loom-weight (Figure 2, Number 2), a pair of bronze tweezers (Figure 2, Number 3), a weaving needle of bone and another of antler and two bone pins. Several of these items seem to relate to a weaving-process probably carried out in the immediate vicinity or more probably in the hut itself. A late-Roman bronze coin of Tetricus (AD 268-270) was found about 10 inches above the hut-floor and may have been washed into the hut sometime after it was abandoned.


It is at once clear that this hut belongs to the Saxon "Grubenhaus" (dug-out house) class as noted at more than a dozen British sites, mostly of the 5th-7th centuries having originated in Northern Europe. Its size, shape and depth are typical of Grubenhauser elsewhere, the major British parallels being at Mucking, Essex (Reference 3) and West Stowe, Suffolk (Reference 4). The posts set partly into the sides on the long axis are characteristic features of many of these huts and seem to have supported a ridged roof. The middle post is not a typical feature and may have been inserted at a later date, at an angle perhaps to help support the roof or the east post.,lt seems clear that this hut is the very first of this type to be found on a rural site in Kent though several Saxon huts found at Canterbury seem to have had sunk-floors. The presence of weaving needles, pins, weight and comb suggest that the hut itself was used for weaving. The series of stake holes on the south side of the hut may represent the actual outline of an upright wooden loom, or more likely a series of small looms. Many of the huts at Mucking had been used for weaving.

DRAWING: Some of the Saxon finds from Lower Warbank, Keston.

Drawing Caption: Figure 2. Some of the Saxon finds from Lower Warbank, Keston.

The domestic rubbish in the hut appears to have been dumped into it probably after the hut had gone out of use. The large proportion of Roman material, nearly 50 per cent of the dateable objects, must represent residual debris from the nearby Roman buildings. The animal meat-bones show a considerable preference for pigs, though here again a large element of Roman residual material may be present. If the bones do reflect the Saxon occupation of the site then a pastoral form of farming may be indicated with sheep providing both meat for eating and wool for weaving.

The provision of a precise date for this hut, based upon the material found in it, is at this early stage somewhat difficult. Some of the more significant pieces of pottery have been examined by Dr J N L Myres who has suggested that most may date from the 6th century AD. Dr Myres has also remarked that the stamped pieces look like 6th century Saxon pottery, rather than anything specifically Kentish, resembling material from Saxon sites along the Thames estuary.

As regards the lead ring, probably a loom-weight, somewhat similar ones have been found in Saxon huts of the 5th-7th centuries at Mucking. Some had previously been found in the adjacent Linford Quarry excavation by Mr K J Barton in 1955. Mrs M U Jones has very kindly supplied drawings of some of those found at Mucking and remarked that they were found with 5th century pottery and melted lead in one hut. Mrs Jones has also noted that lead rings have been found at Hanwell, Middlesex and at Ezinge terp, Holland.

In the absence of more exact dating criteria the Saxon hut at Keston may be dated provisionally to the 6th century AD.

The discovery of this Saxon hut at Keston is undoubtedly one of the most important discoveries so far made by the West Kent Group. Not only is the hut unique in Kent, but it also occurs right in the centre of a large Roman villa complex. Pottery and loom-weights found by the Group at Darenth in 1969 (publication pending) has proved Saxon occupation at another villa site in West Kent and Dr Myres has shown that Saxon pottery has also come from the Roman villa at Wingham near Canterbury. Thus, at least three important villa sites in Kent can now be shown to have been occupied at least sometime in the Pagan-Saxon period. These facts may finally explode the tiresome tradition that Saxon settlers always kept clear of Roman rural sites! Indeed similar evidence is now slowly being accumulated for other rural areas of Roman Britain.


  1. B J Philp, KAR Number 18 (1969), page 13.
  2. B J Philp, KAR Number 11 (1968), page 10.
  3. Mrs M U Jones, ANT. JNL., XLVIII (1968), page 210
  4. S. E. West, MED. ARCH- XII (1968), Page 161
  5. 1. N. L. Myres, ANTIQUITY, XVIII (1944), Page 52
Close Window
Accessed this page via search engine or bookmark? Full K A R Index here.