Pagham Excavations

Saxon Churches in Sussex
Home Up

horizontal rule


by D. J. Freke, B.A., M.A.

 A substantial part of the nave and crossing of the church of St. Thomas the Martyr at Pagham was excavated prior to the installation of central heating. Two periods of stone building pre-dating the present standing structure were uncovered, and a fragment of a tenth-century ring-headed cross was found in a grave filling.


In 1976, Mr. F. Aldsworth, Archaeological Adviser on the Chichester Diocesan Arts Council, informed the Sussex Archaeological Field Unit that underfloor central heating ducts were to be installed in the parish church at Pagham. This would entail considerable disturbance of the archaeological layers under the floor of a church which was considered to have a possible late Saxon wall still visible,1 and which is in an area of proven middle and late Saxon activity. With the kind permission of the incumbent, the Reverend J. Maynard, and the essential co-operation of the contractors, Betteridge and Sons, excavations under each half of the nave and crossing were undertaken in two ten-day periods in October and November, 1976.2 The work was directed by the author, assisted by Gilian Craddock, B.A., Ian Blair, students from the Institute of Archaeology, London, and local voluntary help. Only the areas which were to be disturbed by the ducting and the re-laying of the floor were excavated. None of the graves detected were emptied, so Myres' suggestion that there might be an extensive pagan Saxon cemetery in the area was not checked.3 In this report the term `stage' is used to describe the accumulated standing fabric at a particular time. The term `phase' is used to denote a period of building activity. They may describe the same thing, for example, Stage 1 is also Phase 1.




The earliest evidence of occupation in the area is the Roman material discovered in the excavations at Becket's Barn in 1974.4 Middle Saxon pottery was found in the same excavations. The sixth-seventh century pot found in the churchyard in 19545 is further evidence for middle Saxon settlement in the area. Pagham is the subject of a transaction dated c. 680 in which Caedwalla gave the area to Bishop Wilfred, the missionary to the South Saxons, although the charter recording this transaction is probably not as early as the seventh century.6 A church in Pagham is first mentioned in Domesday Book. Fleming considered that this referred to the chapel of St. Andrew7 whose ruins are in the grounds of `Little Welbourne' west of the church, but there is no evidence to suggest that this structure is any earlier than the early thirteenth century.8 The excavations reported here showed fairly conclusively that the Domesday church and probably at least one earlier church were on the site of the present parish church of St. Thomas the Martyr.



The removal of the pews from the south side of the nave and crossing made available for excavation a sub-floor area aproximately 14.8 m by 2 m with a 2 m by 0.3 m extension eastwards and a 3 m by 0.2 m spur across the south aisle. A similar trench 10.6 m by 2.9 m with a spur 3 m by 0.65 m became available after the removal of the pews from the northern side of the nave and crossing. It was excavated after the heating ducts and flooring had been installed in the southern half.

A sub-floor space had been excavated under the pews when they were installed in 1837.9 The restoration carried out at this date also raised the floor level approximately 16 cm.10 This was confirmed by two slabs of polished Paludina limestone, one trapped under the refacings of each of the piers west of the crossing, 16 cm beneath the present flagged floor. Shallow brick sleeper walls had been built down the length of each trench to carry the joists of the wooden floor, and their foundation trenches had penetrated earlier layers and features (Fig. 4).

Under general layers of soft debris (Layer 1) over the whole of both trenches, there were intermittent layers of hard mortar (Layers 2 and 42). The date and function of these layers are conjectural, but as the Paludina limestone slabs (see above) were bedded on them, they probably represent the bedding of a flagged floor of which the limestone slabs are the sole survivors. The layer in the northern trench (Layer 42) had a hard, almost polished surface which may indicate that there were no slabs in this area. Only the area of surfaced mortar is shown on the plan. This surface was not found in the southern trench, but it may have been lost when the sub-floor space was excavated in 1837. The mortar varied in thickness, being up to 12 cm thick in places. It overlay a layer of dirty clay (Layer 41 a) above the natural silty clay. No other lower floors were detected in the area of the trenches, although the fact that the mortar covered the earlier demolished walls at the east end of the trench (Fig. 4) implied that Layers 2 and 42 could not have been the original floor.

The graves cut through this mortar were not emptied. They consisted of three types: 1, brick vaults; 2, stones over an earth filled grave; and 3, earth filled. Twelve of the latter were recognised although more may have escaped detection as the fillings were virtually identical to the surrounding earth. There was one brick vault under the unexcavated central aisle which projected into the southern trench and three stone covered graves were also found. A fragment of a tenth-century ring headed cross was found amongst the stones covering the one in the northern trench (report by D. Tweddle below).

Two masonry phases were found under the mortar layers at the east end of the trench. Lack of time allowed only three sections to be cut through these walls (Fig. 3 for locations). The smaller, and presumably earlier, structure (Fig. 5, stage 1) had walls approximately 90 cm thick, constructed with a double skin of facing stones filled with beach pebbles and mortar. Some of the facing stones came from the Mixon Reef, now under the sea off Selsey Bill, but accessible in the medieval period. The mortar was made with beach material (report by S. Hamilton below) and in situ it was so hard that it could not easily be removed without power tools, despite the fact that only one course of masonry survived. The foundations consisted of a foundation trench approximately 30 cm wider on the inside of the wall and 40 cm wider on the outside of the wall. It was `U' shaped, about 1 m deeper than the lowest masonry course, and was filled with various forms of hard-core: coarse gravel and beach pebbles in clay under the southern arm, and larger flints, pebbles and clay under the west and north portions. This structure is the first evidence of any building on the site. No traces of an earlier timber building were found, although if one had existed on the same plan as the masonry structures, any post-holes or beam-slots would have been obliterated. No medieval post holes or beam-slots were found.


The second masonry structure (Fig. 5, Stage 2), encased the north and south sides of the earlier one, but extended further to the west. Its 1 m thickness was only completely exposed in the narrow spur trench across the south aisle. Its construction was similar to that of the earlier structure, except that its 1 m deep foundation trench was filled with mortared pebbles and flints. Only one course of facing stones survived. No west wall of this phase was found in the length of our trenches.

It is possible that the quoins visible on the outside of the south wall of the chancel are a surviving upstanding portion of this stage. The spur trench across the north aisle revealed no trace of this phase of building (Fig. 4b), so there is a possibility that this stage incorporated a north aisle or north porch. Figure 5, Stage 2, shows a conjectural north aisle, but there is no evidence in 1976 to decide which of the two possibilities is more likely. This second phase has been assigned to the eleventh century, following Fisher and Poole, on the basis of the quoins and quasi herring-bone work in the south wall of the chancel.

The main structure of the church as it is now seen under its nineteenth century renovations, is thirteenth century. The sections of the second phase, visible in our trenches, were demolished to provide foundations for the present crossing piers and south nave arcade. Possibly the north aisle outer wall is contemporary with Phase 2, or rebuilt upon the same foundations.



Stage 1

The date of Stage 1 is impossible to fix. The fragment of tenth-century cross was discovered, re-used with other stone fragments, in a grave which must post-date the Phase 2 reconstruction, as it was cut through the mortar which covers both Phase 1 and 2 masonry. The cross itself does not prove the existence of a tenth-century phase of building. The presence of St. Wilfred in the vicinity in the seventh century is tantalising, but a church at Pagham is not mentioned until the Domesday Survey. The fragments of a late Saxon pot (Fig. 6, No. 1) and a medieval pot (Fig. 6, No. 2) came from disturbed areas at the east end of the southern trench. They were unstratified. The most that can be said is that the earliest phase is very likely to be Saxon on three grounds, none of them conclusive:

1. It is the earliest structure found on the site in an area of known middle and late Saxon occupation.

2. It is very narrow (4.6 m). out of 43 measured churches considered by Fisher to be Saxon or on Saxon foundations, only four are narrower (Chithurst, Poling, Selham and West Stoke) and two others equal (Coombes and Sompting).

3. It pre-dates Phase 2, which may incorporate features considered to be Saxon.

The length of the nave and the shape of the chancel of Phase 1 cannot be estimated. The churches mentioned above have nave lengths varying from 7.5 m to 14.6 m.

Stage 2

The west end of Stage 2 probably underlies the remodelled present west end. The east end of this stage is conjectural. The quoins in the south wall of the chancel may indicate the east end of the nave or the chancel. The former would give an atypical (for Sussex) ratio of width to length for the nave of 1:4, whereas the latter, assuming the Stage 1 west end served as the foundation for a chancel arch, gives a ratio of approximately 1:2, which is much more likely. The possibility of a north aisle or porch is discussed above. It may be this Stage 2 structure in which St. Anselm consecrated the Bishop elect of London in 1105.11



Descriptions of layers in Figure 4

1. Soft rubble, mortar, and wood shavings.

2. H and packed surface of mortar. Some brown earth (Floor).

4 and 43. Wall, single course of dressed facing stones, with beach pebble and very hard mortar infill (Phase 1).

11. Soft brown earth.

14 and 44. Wall, single course of dressed facing stones, with beach pebble and very had mortar infill (Phase 2).

32. Medium hard, dark brown earth.

37a. Hard packed mortar. 37h. Brown sandy clay.

37c. Brown clay and flint nodules. 37d. Gravel and shell. 41. Soft brown earth. 41 a. Soft brown clay.

42. Hard packed surface of mortar. Some brown earth (Floor). 46. Medium, hard, dark brown earth. 47. Hard brown earth.

48. Hard grey earth.

50a. Flint nodules and clay. 50b. Flint nodules.

55. Oak and elm timbers (Identification by C. Cartwright). 58. Pebbles and sandy clay.

59. Pebbles and clay.

60. Mortar fragments and brown earth. 61. Dark brown clayey earth.

Stage 3

The tower is dated to the thirteenth century by Poole12 after Fleming,13 but is considered to be Norman by Pevesner.14 On architectural grounds, particularly the way the north aisle arcade butts up against the tower arch, it is considered by this writer to be earlier than the thirteenth-century arcade. It may be twelfth century, but certainly pre-dates the present aisle arcade.


Stage 4

The present shape of the church was arrived at in the thirteenth century.

Later work included internal and external renovation in 1837 and the insertion of the nineteenth-century west end.


THE FINDS (Fig. 6)

There were very few finds. Two large fragments of pottery and a spindle whorl were found in a disturbed area in the southern trench.

No. 1. Rim of late Saxon cooking pot, medium flint filler, black reduced ware, external surface patchy buff-grey. Layer 2.

No. 2. Twelfth-century (?) cooking pot, fine flint filler, black reduced ware. Layers 33 and 35. No. 3. Fired clay spindle whorl, medium flint filled, grey. 30 gm. Layer 33. No. 4. Tenth-century cross (Plate 11). See report by D. Tweddle, below.



Part of an arm and the ring of a ring-headed cross. The inner end of the arm is roughly broken, the break rising from right to left. The outer end is convex, the curve being continuous with that of the outer edges of the short surviving portions of the ring, each of which terminates in a rough break. The sides of the arm are concave, the curves being continuous with those of the inner edges of the ring. The front and rear faces are framed by narrow, plain, raised mouldings of indeterminate section, the front face being decorated in relief with a flaccid, disorganised interlace, the similar decoration of the rear face being largely defaced. Along the edges of the outer end of the arm damaged narrow, plain frames confine a panel of four-strand plait, separated by narrow, undecorated zones from further fragmentary interlace.

The fragment, of Corallian limestone, is a maximum of 12 cm high, 23.5 cm wide, and 10 cm deep, and was orginally c. 35 cm in diameter. It was found in the fill of a grave, feature 53.



The stone is unusual as it is the first fragment of an Anglo-Saxon cross head to be discovered in Sussex, and one of the few to survive from southern England. These can be classified as follows: Free-armed:

Reculver (Kent);15; Amesbury (Wilts);16 Bath (Somerset),17 2; Colyton (Devon);18 Glastonbury (Somerset).19

Ring:               Bath;20 Glastonbury Tor;21 South Leigh (Oxon);22 Deerhurst (Glbs);23 Bradford-on­Avon (Wilts).24



Circle:          All Hallows-by-the-Tower (London).25

Hybrid         Amesbury.26


Disc:            Abingdon (Berks);27 Rodbourne Cheney (Wilts);28 St. John, Walbrook (London).29 Indeterminate: Puddleton (Dorset);30 Cattistock (Dorset).31

The arms of a free-armed head are not linked, whereas those of a ring-head are linked by straight or curved bars, which on a circle head are continuous. The disc-head has a cross marked on each face of an unpierced disc.

Of the southern English ring-heads those at Bath and Bradford-on-Avon probably belonged to large standing crosses, but the small diameters of those from Deerhurst (c. 25 cm), Glastonbury Tor (38 cm), and the closely related example from South Leigh (c. 40 cm), suggest that they belonged to small memorial crosses similar in form to those of the East Midlands.32 These originally stood about I m high, and their heads ranged from c. 35-45 cm in diameter. The small size of the Pagham head suggests that it belonged to a similar small memorial cross.

The suggested function of the Pagham fragment is supported by the fact that no large standing cross, complete or fragmentary, survives in Sussex,33 although there is monumental sculpture. Around Midhurst; at Chithurst,34 Stedham,35 and Cocking,36 is a group of grave slabs made of the local greensand. These are probably of eleventh-century date, since the example from Cocking was recovered from the foundations of the chancel, dated to c. 1080,37 and those at Stedham from the foundations of the eleventh or twelfth-century nave,38 and at Steyning are two similar slabs, one of which came from the foundations of the mid-twelfth-century nave .39 At Stedham also a single round-headed headstone survives of the four which were discovered built into the walls of the twelfth-century nave.40 On the basis of this material it is possible to suggest that in the late pre­Conquest period in Sussex there were active local traditions of the manufacture of grave slabs and markers in stone. It is against such a background that the Pagham fragment should be viewed, and it may, therefore, represent a pre-Conquest cemetery associated with the late Saxon church, the evidence for which is assessed above.

The Pagham fragment, however, may be slightly earlier than such eleventh-century monumental sculpture. Based upon their distribution, Collingwood suggested that the wheel-head (comprising both the ring and circle heads) was a type associated with the area of Scandinavian settlement, originating in the Isle of Man, and spreading thence to the mainland in the early tenth century.41 To Collingwood's map, however, must be added the more recently discovered southern English examples, and when the sheer bulk of those which have survived in Northumbria is weighed against the meagre survivals in southern England, the existence of these must seriously weaken Collingwood's argument for the association of the type solely with the areas of Scandinavian settlement. Moreover, the discovery at Deerhurst of a fragment ornamented with ninth-century seed-pod decoration, also in an area well away from the Scandinavian settlements, suggests that this type of head originated in the ninth-century, and strengthens the suggestion that its origin should be sought in a non-Scandinavian context.

Despite these reservations, Collingwood's assertion of a principally tenth/eleventh-century date for this type of head remains unshaken. The Pagham head should probably be placed in the tenth-century since there is a tendency for later examples to be more angular, with square-ended arms.42 The Pagham head should, therefore, be placed typologically earlier than the examples of this type from Glastonbury Tor and South Leigh, which are probably of eleventh-century date. The flaccid interlace ornament is of little help in close dating, but would support a broadly tenth/eleventh-century date.



ANALYSIS OF THE MORTAR (by Sue Hamilton, B.Sc.)

A sample of mortar, from the Stage I structure, weighing 143 g was disaggregated to extract molluscan inclusions for identification. The disaggregated sample was then treated with 10% hydrochloric acid. When all the mortar was dissolved (17.6% by weight) the remaining inclusions were wet sieved through a nest of sieves. The particle size distribution of these and the shells is represented by a cumulative graph (Fig. 7) and is as follows:

Sieve intervals                     % by weight

63 µm                                                      9.47

212 µm                               26.82

500µm                                  1.36

1 mm                                         19.21

2 mm                                         39.83

6 mm                                           3.31

Total:                                  100

This size distribution and the presence of multi­coloured inclusions is suggestive of beach shingle. Under the microscope, the mollusca comprised fragmented marine gastropods and bivalves. One intact shell was identified by M. Bell, B.Sc., namely Gibbula cineraria (L). This is a marine gastropod found in clean, sheltered habitat in pools or among seaweeds and common on all rocky shores.

The inclusions tend to confirm the marine origin of the shingle used in the mortar. An obvious local source is the extensive shingle bar across the mouth of Pagham Harbour.


Author: D. J. Freke, Rescue Archaeology Unit, University of Liverpool.





1     A. E. Fisher, The Saxon Churches of Sussex (1970) 161-2. H. Poole, Victoria County History 4 (1953) 23 1-2.

2     The interim report in P. L. Drewett (ed.), 'Rescue Archaeology in Sussex' The Bulletin of the Institute of Archaeology 14, (1977) 50-1, was written when only the southern half had been excavated. This report supercedes that one, especially the plan p.51.

3     J. N. L. Myres in A. H. Collins, 'Saxon cinerary urn from Pagham churchyard,' Sussex Notes and Queries 14, nos. 7 and 8 (1955), 123-5.

4     V. L. Gregory, 'Excavations at Becket's Barn, Pagham, West Sussex, 1974' Sussex Archaeological Collections 114 (1976), 207-217.

5     B. Cunliffe, 'Some late Saxon stamped pottery from Southern England' in V. I. Evison, H. Hodges, J. G. Hurst, (eds.), Medieval Pottery from Excavations (1974), 127-135 and A. H. Collins, 'Saxon cinerary urn from Pagham churchyard'. Sussex Notes and Queries 14 nos. 7 and 8 (1955), 123-5.

6     E. Barker. 'Sussex Anglo Saxon charters' Sussex Archaeological Collections 86 (1947), 50-58.

7     L. Fleming, History of Pagham (3 vols.) (1949), 6, 585-6.

8     Victoria County History 4 (1953), 227, note 11.

9       I. Nairn and N. Pevsner, The Buildings of Sussex (1973), 289.

10     L. Fleming, ibid (1949), 596.

11     L. Fleming, ibid (1949), 15.

12     H. Poole, ibid (1953), 231.

13     L. Fleming, ibid (1949), 593.

14     I. Nairn and N. Pevsner, ibid (1973), 289.

15     C. R. Peers. 'Reculver: its Saxon Church and Cross', Archaeologia 77 (1927), 253, Fig. 9.

16     Unpublished.

17     J. O. Westwood, 'Archaeological Intelligence: Saxon and Early Norman', Archaeological Journal 3 (1846), 356, Figs. p.356.

16     T. D. Kendrick, Late Saxon and Viking Art (1949), pl. XXXIV.

19     F. B. Bond, 'Glastonbury Abbey', Proceedings of the Somerset Archaeological and Natural History Society 59 (1913), 62-3, Fig. 1.

20     T. D. Kendrick, 'Fragment of a Cross head from Bath', Antiquaries Journal 21 (1941), 75-6, pl. XVIII.Ring heads of a non Anglo-Saxon type from Cornwall and Devonshire are excluded from this survey.

21     P. Rahtz, `Excavations on Glastonbury Tor, Somerset, 1964-6', Archaeological Journal 127 (1971), 48-9, Fig. 21.

22     Unpublished.

23     P. Rahtz, Excavations at St. Mary's Church, Deerhurst, 1971-73 (1976), 26, Fig. 14.11, pl. Xl.

24     A. Burder, `Excursions: Bradford-on-Avon and Great Chalfield', Proceedings of the Somerset Archaeological and Natural History Society, Bath and District Branch (1904-8), 180.

25     E. Okasha, `An Anglo-Saxon Inscription from All Hallows, Barking-by-the-Tower, London', Medieval Archaeology I I (1967),249-51, pl. XXXIXA and B.

26     A. R. Dufty, 'Amesbury Church', Archaeological Journal 104 (1947), 178, Fig. 7.

27      E. A. G. Lamborn, 'A Note on a Fragment of an Anglo-Saxon Wheel Cross found at Abingdon', Berkshire Archaeological Journal us. 39 (1935), 88-9, Figs. 1, 3, and 4.

28     C. E. Ponting, `Notes on the Churches of Rodbourne Cheney, Lydiard Millicent, Stapleford, Wylye, Wishford, Steeple Langford, and Little Langford', Wiltshire Archaeological Magazine 35 (1907­8, 372, pl. opp. p.372.

29     W. R. Lethaby, London Before the Conquest (1902), 170-1, Fig. 33.

30     Royal Commission on Historical Monuments, Dorset III pt.i (1970), xlviii, pt.ii (1970), 225. Hereafter cited as RCHM.

31     RCHM, Dorset 1 (1952), xxvi, 76, pl. 6.

32     Fully discussed in C. Fox, 'Anglo-Saxon Monumental Sculpture in the Cambridge District', Proceedings and Communications of the Cambridge Antiquarian Society 23 (1920-1), 15-45.

33     There are several fragments of interlace built into the base of the war memorial at Selsey, near Pagham. It has been suggested that these formed part of a large standing cross, but they have been trimmed for re-use, and their original function is, therefore, a matter for speculation. E. Heron-Allen, Selsey Bill: Historical and Prehistoric (1911), 102-3, pls. XXI-XXII.

34     P. M. Johnston, 'Chithurst Church', S.A.C. 55 (1912), 106, pl. V1.

35     J. E. Butler, `The Antiquities of Stedham Church', S.A.C. 4 (1851), 19-21, Figs. 1 and 5.

36     P. M. Johnston, `Cocking and its Church', Archaeological Journal 78 (1921), 182, Fig. 1.

37     P. M. Johnston, Ibid (1921), 182.

38     J. E. Butler, Ibid (1851), 19-20.

39     M. H. Bloxham, `Notes on Places Visited on the Annual Meeting, 14th August, 1863. The Churches'. S.A.C. 16 (1864), 238.

40     J. E. Butler, Ibid (1851), 20, Fig. 2.

41     W. G. Collingwood, Northumbrian Crosses of the Pre-Norman Age (1927), 137-9, Fig. 153.

42     W. G. Collingwood, Ibid (1927), 142.

From Sussex Archaeological Collections 118, 1980
Reproduced by courtesy of the Sussex Archaeological Society (SAS). 
SAS grants this licence for the stated purpose in respect of such rights as SAS may have over the articles, 
but those rights may not include the author's copyright in the words and/or images.

horizontal rule

All presented material is for Private Study Only
Copyright © 2002-2010 Martin B Snow. All rights reserved.
Revised: February 20, 2010 .