Cosgrove Villa & Temple Excavation 1958-1959

Return to main Villa & Temple page

Wolverton and District Archaeological and History Society No. 3 January 1958

The name is Cufel’s ‘Graef’ and the ‘graef’ is interpreted in ‘Place-names of Northamptonshire’ as a pit, ditch or trench; it is feasible that the ‘fosse’ surrounding the Roman Villa was a marked feature when Cufel the Anglo-Saxon owner named it.

V.C.H. Northants records Roman coin finds at two points not exactly placed, and for some time our Society has closely watched soil disturbance. On the turnover of grass-land to plough in August 1957, R. Harris was requested to look over the ground from which he brought in pottery. Further examination with assistance of a probe proved masonry only a few inches below the surface near to the canal embankment. The field gave little away; it has apparently been enclosed park and meadow-land for centuries, protecting it from medieval ploughing which certainly would have exposed it.

When permission was granted and the land-owner expressed a wish for such action, the ground was marked out in grid and digging pro­ceeded without hitch. The first square disclosed pilao of the hypocaust and walls carefully faced on the correct alignment of the grid laid down which simplified the careful clearing of interior rubble from wall to wall. The excavation was restricted to such an area as could conveniently be covered, for the quality of the structure was such as to call for preservation.

The final exposure when digging ended was an oblong 24ft. by l8ft., leaving a centre balk for roof support which coincides exactly with the division of the two chambers. The northern limit cuts across the corridor leading from the furnace floored with blackened bricks. One side wall of brick is well preserved but the other is damaged and not so well defined. The entrance arch to the heater chamber has gone, but once inside we have on the right the brick flooring of the room above, supported by perfect corbelled pilao. The floor bricks are 23’ square and 3” in thickness. This, the’sudatorium’ or sweating room received considerable heat from the box flue tiles packed closely together along the walls of either side. All were smashed and came away with the rubble. The centre of the chamber received the impact of a heavy fall of immense stones, and there was possibly a cavity which could be entered, as the centre pilae were robbed, and a considerable litter of bone and pottery fragments thrown inside by later ‘squatters’! They had also made a rough fireplace from loose stones and over the entrance arch had trodden in the rubble, dirt, pottery and bones so hard that it was difficult to loosen. Falling stone had also severed the low brick arch through which fuel of lesser intensity was thrust into the next chamber.

Across the balk into the ‘tepidarium’ we are again fortunate for on the collapse of the building the floor apparently held and disin­tegrated gradually, leaving nine pilae at almost their original height, with possibly another row under the balk. Their capping bricks supported a floor of ‘opus signinum’ four to five inches in thickness - a strongly bound concrete of lime and brick with a mottled red surface finish capable of an attractive polish. The wall plasters are of warm pink, also red with broad and narrow white stripes and fragments of pale green. The bank of clay two feet high along the west wall of the ‘sudatorium’ does not carry through into the lesser heat. The S.W. corner has strong well-finished angle walls enclosing a square of bound rubble from which two brick steps lead down into the plunge bath. Here the excavation ends, but the portion exposed is seen to have smooth plain plaster sides, with regularly placed nail holes, from which a lining, presumably of lead, has been wrenched away. The bath was packed with a down-thrust of stones in which were large pieces of shaped ‘foreign’ stone resem­bling pumice.

The finds are not numerous; the hypocaust is not a place where one would expect them. Fragments of window and other glass, together with pieces of semi-molten lead suspected as window fittings have gone for examination to Dr. D. Harden of the London Museum. There is a floor tile which received a deep impression of a dog’s foot, numerous metal pieces, and a small silver earring. The pottery is being cleaned and some of it reassembled. Fabric pieces, together with finds, progressive photographs and E. Cockerill’s scale plans are finally to be housed at Cosgrove Hall for the convenience of visitors and students.

The exposed structure has been pointed and made secure, and covered with a permanent roof by Major. Hon. J. Fermor-Hesketh - an act of generosity for which he deserves the thanks of all who value this archaeological find, and of those who will eventually benefit from the study of it.

Digging team:- J. Berridge, J. Browning, E. Bailey, K. Tull, D. Gowland, J. Hawtin, W. Johnson, C. Nichols, K. Whiting, S. King, E. Cockerill, Mrs. King, R. Harris, Miss Escott.


Wolverton and District Archaeological and History Society No. 4 January 1959

The Roman Villa

The Ministry of Works have called for a detailed publication of our work on the Cosgrove Roman Villa, due to them as custodians of a scheduled Ancient Monument. This will be prepared for inclusion in the 1959 Report of the Northampton Natural History and Field Club (Archaeological Section) So that here only a condensed commen­tary is necessary.

We have first to thank Major Hon. J. Fermor-Hesketh for kindly providing us with a substantial hut, divided into a tool store and office; for his generous gift of light wheelbarrows and the heavier tools, and his readiness to offer whatever may be required for the preservation of the structure.

The Bath House hypocaust is now uncovered, and work extended south to ascertain its relation to the main building. Further north we cannot go because of the embanked terrace which takes the canal along a slightly higher level. The terms north or south are of convenience; the precise alignment is S.E.-  N.W., and the refer­ences hereafter are as the plan inside the back cover.

The building is a well constructed unit, with strong walls of local limestone carefully faced, containing a sufficient variety of pilae and internal brickwork to make it an attractive survival of Roman construction; as Mr. Alan Warhurst puts it -‘the most inter­esting piece of Roman structure between . Albans and Leicester’.

The overall outside measurement of the hypocaust is 34 x 13 ft. divided into the usual compartments:-

(a) The Stokehole (Praefurnium) at the north end is 9 ft. x 6 ft. with the furnace recessed on the west side. We must assume that it was a cellar with no superstructure, possibly, but not proven, with a stone barrel-roof, which, later in time became the receptacle for the stones and rubbish around. So much of it obviously did not ‘belong’ and only nearing the soot layer did we find the genuine ‘fall’ - the shaped stones of the furnace dome or arch, broken metal fittings, innumerable nails, and , fallen forward from the next room above, roofing, wall plaster and window glass. The floor is of marl on rammed gravel, and the back of the furnace gravel cement.

The entrance from the outside of the north-east corner is not quite what we expected. True, we see a path approach from the east wing, post sockets and incisions on the wall for the wail for the pivot-shoes of the door, but no steps - only a slope down beside the outer wall. The narrow heat charnel to the under-floors is walled on one side with bricks, the other with a lower brick course and rubble above, and a floor of square bricks. The soot layer, six or eight inches in the stoke-hole gradually diminished to two inches at the furthest end.

(b)  A tiny service alcove in the narrow waist of the corridor seems to have provided us with much or our translucent glass, molten lead, and deep red wall plaster with a white stripe, and the roofing in the furnace debris. Trodden or crushed into a soft floor, were many small bones, and a concentration of coarse pottery, some of it since reassembled by Mr. and Mrs. King. We name it an alcove, as pottery was on the cross wall, which would imply that it was part of the wider room which ends at the internal pilasters, which had a floor of large bricks. We see two of them two feet square and three inches thick ‘in situ’ on perfect corbelled pilae, with the others so placed for the bricks to extend from wall to wall. These also formed the floor of the

(c)  Caldarium, which, if commencing at the pilasters, is 8 ft. x 5 ft. This chamber received the maximum heat from the jacket of flue tiles packed tight along the outer walls. We hoped for complete specimens to retain in position but they were crushed almost to the floor. Along the west wall is a bank of clay up to two feet in height, inexplicable unless as assisting heat conduction. This room was entered from the west by a porched doorway, and provided large pieces of pinkish plaster, some of it shaped as for a window embrasure or the turn-in of the doorway. A double cavity wall of bricks with spaced apertures and damaged centre arches leads under the

(d)  Tepidarium, the chamber of lesser heat. We see in their places the thin stone slats in front of the apertures acting as checks in the passage of heat. The flue tiles were less closely spaced along the walls, and the flooring is of ‘opus signinum’- a 4” thick mix of brick chips, gravel and cement, with an overall red finish capable of an attractive polish. Fifteen columns of brick pilae at almost their original height, which with their capstones supported this floor give the impression that this end of the hypo­caust suffered no subsequent interference after subsidence of the roof. A lateral stone wall ends the heated basement, and we now move up on to floor level to the

(e)  Frigidarium. This, the starting point of our excavation is only a few inches below top soil, and is equally divided between a bound rubble floor and the splash bath four feet square. Two brick steps lead down to the bath, its plaster lining showing the impress of regularly placed nails, the original lining presumably having been of lead. The lead robbers damaged the end of the lower step by wrenching away the ‘intake’, but the outlet pipe underneath was too much for them and is still there. We found the bath carefully packed with very large stones, which if thrown or fallen in would have damaged the plaster, on which was scarcely a scratch.

Built centrally against the end wall on the rubble floor is the latrine, with the well constructed drain sweeping away in a curve and joined by the outlet on its way under the apsidal room towards lower ground.

(f) The apsidal room is the terminal corridor from the west wing of the house. Having no faced walling it was clearly a timber framed building on a floor of angle-pitched stones and rubble; Its surface lost in the field cultivation, is  only indicated by minute brick chips bound with plaster seen in between the stones. The timbers left a dark stain along the south edge of the platform, and nails were spaced at intervals. A threshold of wood, with sockets for door, porch posts and Steps down, give egress to a small court or garden. A cutting shows no yard layer, so garden is assumed. On the lower step level and in front of it was a small semi-circle of oyster shells and round stones - children at play, but in what era?

 Room No. 1 has squared walls, dovetailed from the turn of the apex. Here again the room floor has gone, but the pitched base was packed too tight for plough ejection. Only the corner is within our outside grid, but inside the wall is a large post-hole for a stout upright supporting the roof. The opposite side of the little court suggests a verandah. It has a rubble floor, very much timber stained, but a necessary wide balk for working purposes obscures its lines for the time being.

The curious incidence of our 20 ft. grid falling exactly on the Roman building plan had a parallot.at Lullingstone, causing there as here a mixture of confusion and amusement; each of our sub-divided squares of 8 ft. balks having walls or room division under them, nothing making sense until their removal. Otherwise the excavation was singularly uncomplicated, but so far we have not cut down and across the foundations, our object this season having been simply to trace the building layout within the enclosure. Specimens of fabric components, pottery, metal and glass are housed in the museum, kindly placed at our disposal by Major Fermor~Hesketh within the Cosgrove Hall grounds. There were few small finds and only one surface coin - a silver Elizabeth 1, but this is not a part of the house where finds would be likely.

If or when we extend on each side to the wings, the question will immediately be solved as to whether the house is quadrangular with the bath house in the centre courtyard, in which case the north side is under the canal bank.

Crossing the canal where the pasture is level with it a grid was laid out to locate the water source, which decided the selection of the villa site. We have so far uncovered the ‘nymphaeum’ the supply cistern for the Villa (placed similarly to Chedworth), the head of a small stream which once ran alongside the house, but was severed and diverted by the canal works into their drainage. There had been some interference around it, and they had led a stone drain into it by breaking into the parapet. How odd that the contemporary and industrious Baker who reported the coin finds in this canal operation did not see and recognise its Roman origin! The spring on clearance become clean and lively, and can be heard falling down into a constructed channel covered with capstones not yet removed. At that point the weather deterioration made further work impossible.

C. W. Green

Drawings from scale plan by E G Cockerill referring to Commantary Page 9
Scale 10 feet to 1 inch.


Pilae  Pillars of brick (or stone) 20-40cm sq. up to 0.8m high supporting the floor of a hypocaust.

Tepidarium Warm bathroom heated by hypocaust

Hypocaust A space under the floor of an ancient roman building where heat from a furnace was accumulated to heat a room or bath

Sudatorium A hot air room used for sweat baths

Opus signinum A term for a kind of roman hydraulic concrete partly composed of crushed brick that was used for covering walls and floors.

Praefurnium A stokehole

Caldarium  Hot room (moist heat) usually contains or is adjacent to a hot plunge bath which produces a hot humid atmosphere.

Frigidarium An unheated room in a bath house containing a cold bath and sometimes a labrum (cold shower).

Apsidal  Of or relating to an apse

Nymphaeum Sanctuary consecrated to water nymphs also served as reservoirs and assembly chambers for weddings.