Fourth Report, 1972: Part V

 Click on image to enlarge

Today, with thanks to PWD for permission, the final part of the Fourth Report from 1972:

"The Period 3 Settlement

The Period 2 settlement was destroyed by fire, presumably at the same time as the burning of the Period 2 defences.  Before the laying of the Period 3b road surface a large roundhouse was built across the western end of House 2.  The post circle of this house was 11.6 m in diameter, but outside it, to the north, perhaps platform formed a shelf about 3 cm higher than the surrounding bedrock; this shelf, produced by traffic wear around the house, ran from 1.7 m to 2.4 m outside the circle; on the south side a burnt streak and clusters of burnt daub continued the line, which is taken to represent a wattled partition framed, like those of the Period 2 houses, on sleeper beams resting on the ground surface.  The porch lay on the west side, facing away from the entrance to the hillfort, and appears to have been supported by diagonal braces set in slots.

The total diameter of the roundhouse was about 14.8 m.  A number of post holes whose dating is uncertain lay within the circle.  Some may belong to partitions at the west end of House 2 of Period 2: postholes 522, 541 and 542 probably associated with the roundhouse, but none were substantial enough to have held up the roof.  No central support was possible, for a hearth lay in the middle of the house, and while roof supporting posts might have rested on the ground surface (producing a tower-like structure similar to that proposed for Little Woodbury roundhouse (Bersu, 1940) no evidence of this is likely to be forthcoming.  Like the roundhouses of similar size, and apparently similar construction, at Pimperne or Longbridge Deveril, the roof construction of the Crickley roundhouse is problematic.

To the south west of House 4 two small hearths were surrounded by clusters of small postholes.  Further holes, both small and large, lay to the south of the roundhouse and to the north of House 1.  For none are there adequate grounds to allow period attribution, and all are shown in figure 5 as of uncertain period.  The excellent preservation of the hearths might suggest ascription to the final phase of occupation, and this, more tentatively, might be indicated by the undue proximity of one of the structures to the front of House 5.  The post holes around hearth 801 could be seen as a trapezoidal building: heavy burning to the south of the hearth had reduced the bedrock to quicklime, an intense heat that suggests some industrial function from the building.  The internal post holes were perhaps to support a frame around hearth itself.  No clear pattern can be made of the post holes in other areas, and, as flimsy shelters, drying racks or similar two or more pos-structures, some may belong to the Period 2 and others too are shown for convenience, on the Period 3 plan (figure 6).

If the Period 3a defences are correctly interpreted as a temporary re-fortification of the hilltop after the Period 2 destruction (Dixon, 1970), the roundhouse may have been built during Period a 3a.  It was burnt down during the destruction of the 3b and  defences, and the site was then abandoned.

Discussion

The decorated sherds of fine fabric with incised linear ornament (illustrated and described in Dixon, 1971) which came from the info of the latest rampart (Period 3b) are now closely matched by pottery from the postholes of the Period 2 roundhouse.  Almost all the material from the Period 2 post holes, however, consists of sherds of coarse fabric with sagging shoulders and irregular rims; the only decorated sherd comes from the post-pipes of the five-post  structure between Houses 1 and 3, and presumably  relates to the period after the Period 2 settlement. 

This break in the ceramic tradition, whatever interpretation be put upon it, should therefore be placed between Period 2 and Period 3a, a period which saw the change from longhouse to roundhouse, and also from daub filled with grit to clean sifted daub in the wattling of these buildings.

Comparanda for the pottery should be treated with caution in view of the paucity of the material.  The closest parallels for the sherds of the latest period appear to come from the Upper Thames (references given in Dixon 1971), in context now ascribed to the initial stages of the Iron Age and dated not later than the sixth century BC (Harding, 1972).  The cultural context of the longhouses is at present quite uncertain.  Analysis of strata deposited by erosion between Period 2 and Period 3a indicated no considerable time-span (Dixon, 1970) but calibration of this in terms of years, decades or even centuries is hardly possible.

The longhouses themselves are of a type seldom recognized in Britain.  At Park Brow in Sussex (Wolseley, 1925) a ten-poster longhouse was associated with Hallstadt pottery, and a very few others have been noticed in Wales, Scotland, and, conjecturally, on the Thames (Harding 1972, pp 34-5) But on the continent the longhouse type, many with evidence of byres as true longhouses is predominant in many areas from the Neolithic onwards.  The discovery of an extensive regularly planned settlement of longhouses at Crickley is thus an unexpected outcome of the 1972 season.  Although the dearth of insular parallels draws attention to origins on the continent, it is any reasonable to suppose that further settlements of this sort still await discovery, on new sites or perhaps even among reports of old excavations.

Future work

In 1973 we shall continue the excavation of the area behind the entrance, and hope to reveal more of both the longhouse and the roundhouse settlements.  The work requires volunteers and adequate financing to augment the generous support of the Gloucestershire College. Donations (to the Crickley Hill Fund) and offers of assistance should be sent to the Secretary of the excavations, Mr.  R. D. A.  Savage, Gloucestershire College of Art and Design,, Cheltenham Glos., GL52 3JG from whom further copies of these notes, and copies of the notes on previous seasons, may be obtained (1969, 15p 1970 and 1971, 25p all plus 5p for postage etc.).  Persons interested in taking part in future seasons of excavation in this series should get in touch with Mr Savage at the address given (telephone Cheltenham 32501).

References

Bersu, 1940: ‘Excavation at Little Woodbury’, Proc. Prehist.Soc. VI

Dixon, 1969, 1970, 1971: 1^st 2^nd and 3^rd interim reports (Gloucestershire College of Art and Design in Cheltenham)

Harding, 1972: The Iron Age in the Upper Thames Basin

Soudsky, 1969: ‘Etude de la maison Neolithique’, Slovenska Archaeologia XVII

Stanford, 1970: ‘Credenhill Camp…’ Arch. Journal CXVII

Waterbolk, 1964: ‘The Bronze Age Settlement of Elp’ Helinium 4

Wolseley et al, 1925: ‘Prehistoric and Roman Settlements on Park Brow’ Archaeologia LXXVI"

Fourth Report, 1972: part IV

Click on the images to enlarge.

Thank you again to PWD for permission to reproduce abstracts from the Fourth Report for 1972.

"Reconstruction of the Period 2 Houses

The Period 2 hollow way narrowed where it ran between Houses 1 and 2, and came no closer than 1.5m to the lines of posts.  The traffic was presumably here being funnelled by a barrier, and it thus seems likely that the outer walls of the houses lay the beyond the lines of the post holes, which thus formed internal supports of the roof structure.  Absence of post holes for such an outer wall creates no difficulties, for the problems, experimentally observed, during the course of our excavations, of cutting small post holes in the oolite might well induce even builders accustomed to earth-fast posts to rely on walling framed on sleeper beams.  In House 5 two post holes in this outer position can be seen as additional evidence for a framed exterior wall: House 5 had no other postholes in this position and the two outer holes perhaps supported a special structure -- the largest house having the most imposing facade.  Between House 1 and House 3 a small square structure with a central posthole, perhaps to support a raised floor, projected beyond the lines of postholes into the roadway.  As can be seen in figure 6, however, the proposed outer walls of Houses 1 and 3 aligned with the south wall of this square structure, which thus continued the line of the street frontage. 

Suggested reconstructions of the longhouses are shown in figure 7.  The simplest method would probably have been to fit horizontal timbers (aisle plates) between the aisle posts on the longitudinal axes of the houses; as at Stonehenge, a plain mortice-and-tenon joint would suffice to join aisle post to aisle plate.  Reconstruction 2 shows one such interpretation with the addition of a substantial outer turf wall to support the ends of a common rafter roof.  Considerable quantities of burned daub from wattling were found around the houses and in the postholes.  This suggests that the outer walls were merely wattled screens, but these could conveniently have been framed on sleeper beams and could, even without earth-fast posts, have supported considerable weight.  But it is noticeable that the post-settings are paired symmetrically, and this suggests an alternative more complex upper structure.  The aisle posts could have been held together by tie-beams across the width of the house and the roof supported on couples at about 10’ centres.  Such a structure - a principal rafter roof - is shown in Reconstruction 1 and in the perspective and axonometric drawings in figure 7. Further elaboration of the roof structure is too hypothetical, but could include bracing and crown posts on the tie-beams.

House 1 and House 7 (see figure 3) preserved evidence of centrally placed hearth, and other houses may have been similarly heated.  The hearth in House 1 survived merely as the circle of scorched bedrock, and the erosion in other areas could easily have removed the evidence of fire. 

The hollow way continued as far as the end of the second bay of houses one and two.  No further trace of it could be found, and it is therefore possible that the entrances to these longhouses were lateral (see cover drawing).  At the east end of House 4 an additional posthole, 811, may have been for a gable door.  Apart from a streak of burning which might indicate that a porch stood on the north side of bay five in House 5, there was no other evidence for door positions, and there is so far very little trace of the subdivisions that are found in comparable structures on the continent of Europe (e.g. Waterbolk, 1964)."

Fourth Report 1972, part III: The Interior of the Hillfort 1972

(click on the images to enlarge)

Thank you as ever to PWD for permission to post these abstracts from his 1972 Fourth Report. I'm afraid the Albergo Parker is not equipped with an A3 scanner so each A4 page has to be scanned and posted separately:

"An area 80 m long and up to 40 m broad was stripped behind the entrance passage (cuttings BI - BXII).  Stratigraphy in this area was slight, in many places less than 5 cm between turf and bedrock, but a number of the 180 postholes uncovered here can be placed in relative sequence: near the centre of the excavated area postholes 472 cut into and was later than a large posthole; further southwards posthole 529 - 534 were cut into a layer of small packed stones which sealed the tops of posthole's 544 - 547 and 516.  Postholes 529 - 534 formed part of an almost perfect post circle, whose postholes must be seen from their regularity as contemporary; it does not therefore follow that the postholes stratified earlier in the sequences were all of the same date, but this conclusion is indicated by the consistency of their layout: almost all formed pairs from 2 1/2 metres to 3.3 m apart, with similar spacing between the pairs.  

The structures represented by both these groups of postholes may be tied in with the phasing of the rampart.  Within the inturned entrance of the timber-laced rampart (Period 2) the excavations in 1970 exposed a hollow way, produced by traffic wear.  This hollowing continued on the axis of the passage to run between the alignments produced by the pairs of (earlier) post holes.  During Period 3b a cobbled road was laid against the entrance bastions and could be followed running into the interior of the hillfort.  These cobbles overlay the structures identified in figure 5 as "House 2" and "House 6", but curved to avoid the post circle, which must therefore have been in existence in Period 3b.  In the case of the rows of postholes to the west of the post circle (Houses 3, 4 and 5) no evidence of phasing was uncovered, but their alignment corresponded quite closely with that of Houses 1, 2 and 6, and all are thus assumed to form part of one settlement, whose plan consisted of houses set on either side of the road away from the Period 2 entrance.  None of the settings were completely regular.  House 2 and House 5 in particular contained pairs more widely spaced the other postholes in the group and House 3 and House 4 each had a pair of post holes stepped southwards of the general alignment. Post pipe packings survived in many of these post holes, and showed that the post might be placed eccentrically in the hole to compensate the irregularities in the spacing, but even this was not always sufficient to produce a straight row of posts.  Each could thus be argued that some or all of the "houses" shown in figure 5 were in fact separate contiguous buildings, perhaps of four or six posts as identified on a number of Iron Age sites (Stanford, 1970).  But between each "house" lay gaps greater than the spacing of pairs within each of the "houses": these gaps, together with the close correspondence in overall alignment within each "house" group in contrast with slight variations in direction between groups, make such a suggestion implausible.  Indeed, it has recently been argued that such variations in direction and construction within a single structure may indicate variations of function within the building - between house and byre, or between single and double storeyed portions of the house (Soudsky, 1969).  Arguments as to whether all the longhouses attributed to Period 2 were in fact strictly contemporary cannot be satisfactorily resolved: none of the buildings overlapped, and the suggested plan of the settlement (figure 6) reveals a regularity in the layout which very strongly indicates that all the longhouses were designed to form a single coherent settlement."

Fourth Report 1972, part II : the Neolithic Enclosures

Thanks again to PWD for permission to post these extracts from his 1972 Fourth Report - click on the images to enlarge:

"The Neolithic Enclosures (Figures 2 – 4)

Two cuttings, CIII and CIV, were placed adjacent to the areas excavated in 1971; they confirmed the phasing established in 1971.

The two period bank identified in CII (Dixon, 1971) continued across cutting CIII (see figure 2).  In its latest phase, now called 1d, the bank was surmounted by a fence, the burnt traces of which were visible without a break as far as the north section of CIII. both phases of bank contained quantities of Neolithic pottery and worked flint, of types comparable to those found in 1971.

In the flat area behind the bank two hearths were uncovered.  These, and another in CII, are perhaps to be associated with post holes which may have supported screens, for no enclosed structures here can yet be identified.  At the southwest corner of CIII a pit contained only heavily burnt slabs similar to those of the hearths, which resemble in their construction the hearths, possibly of Period 3, found near the roundhouse and the hearth in CIV.  The design, however, is rudimentary, and in the absence of dating evidence any period ascription would be unwise. It is thus still unclear whether any of the structures behind the banks can be attributed to a Neolithic occupation.

The Period 1d ditch in CI continued without interruption across CIV (Ditch 304/603).  Some 45 m of this ditch have now been examined; the material from it and its bank is Neolithic throughout, including a fragment of a polished stone axe, but there is by now some doubt that it is in fact segmental, and the description "causewayed enclosure" seemed inappropriate for phase 1d.  Below the 1d bank ditch 853 (of Period 1b) formed a further section of ditch segment 384; at the northern side of the cutting the 1d bank has not yet been removed, and may conceal a causeway, but it is still possible that 853/384 is the same ditch segment as Ditch 319 which was exposed in cutting CI; thus the causeways uncovered in 1971 remain the only causeways across the two inner ditches.  From the primary silt of 853 came two decorated sherds kindly identified by Mr Humphrey Case as similar to material from the causewayed enclosure at Abingdon.

Towards the eastern end of CIV two segments of the ditch line (ditches 612 and 699) continued the outer ring of ditches excavated in 1971.  The broad causeway between 612 and 699 was matched by gap in the bank on their west sides. Through Across this area ran a structural weakness in the hill (a "gull"), filled with eroded limestone which had the consistency of concreted sand.  Unlike the normal oolite laminae elsewhere on the site, the material in the gull retained the impressions of stake holes, but the interpretation of the structures represented here and in the corresponding part of CIII was complicated not only by the continuing leaching of limestone into the gull but also by the activity of tree roots and rabbits which had taken advantage of the softer rock.  At least two phases were distinguishable (see figure 4).  The latest consisted of a hearth (628) set in a hard-trodden floor (629) of gull material mixed with earth and charcoal dust.  To the south and east the floor merged into a layer of clean yellow dissolved limestone; to the north it had been eroded by activity in the gull.  As a consequence it could not be established which postholes within the gull had been sealed by the floor.

Postholes 879, 615, and 880 can be associated with the hearth and floor, and a further small posthole between 615 and 880 was only doubtfully sealed;  within the gull postholes 640 and 621 resembled the former postholes in size; to this group posthole 860 seems an obvious addition, but this was larger and shallower than the others, similar in fact to postholes 861 and 858; posthole 858 predated 640 and it is possible 861 and 858 were earlier postholes in the group, replaced during a renewal of the structure by the adjacent postholes 640 and 621.  In either case the resultant plan would be a small boat-shaped house with a slightly eccentric hearth.  Daub was found in a hollow, 877, and some of the stake holes may have supported wattled screen wall.

Floor 629 merged eastwards into clean material ultimately derived from the gull.  This stratum overlay the edge of ditch 699, and a laid stone platform, 622, perhaps paving outside the house, similarly overlay the edge of bank 697 but neither ditch nor bank can confidently be placed in the overall phasing of the site; the absence of the causeway in 1d ditch 304/603 to the west perhaps would make otiose a gap here in the outer ring unless the latter pre-dates Period 1d, and ascription of the outer ring to 1b, or even as a single ring to the so far unenclosed period 1a and 1c, may be strengthened by consideration of the extreme erosion of the outer bank, which survived to a height of only 10 cm.  But with no general agreement about the function of causewayed enclosures the argument is weak.  The house itself could well belong to the hillfort occupation, and may be compared with the structures to the south of House 4, and perhaps should be associated with occupation debris, including a shard from a pot with a rounded shoulder decorated with finger tipping, in the upper levels of ditch 699.

Seal below floor 629 two large post holes, 878 and the complex around 884 and 879, can be linked with the ditches.  They sit midway between the bank terminals and were substantial enough to serve as supports for double or triple posts of a small gate; to the north postholes 894, 895 and 898 would then form a fence, but no corresponding postholes were found to the south, beyond the line of the gull.

The group of postholes cut pass into the sides of ditch 603 and part into the infill certainly belonged to the hillfort occupation of the site.  No posthole was found as a pair to posthole 654 but it should have lain entirely within the ditch and could easily have been overlooked in the rubble infill.  Spacing of posts exactly corresponded with that of the Period 2 longhouses and the house and its hearth, centrally placed at the top of the ditch infill, should be attributed to Period 2.  Beyond the crest of the bank in CIII four postholes of similar size and spacing to those of House 7 cut through the 1d bank, and perhaps formed a slightly trapezoidal structure of Period 2.  Immediately to the east of House 7 lay a palisade trench (617) from which the only find was an intact Neolithic arrowhead.  It lay parallel to the Period 1d ditch; equally, therefore, it aligned with the Period 2 house, and its date is uncertain.  A similar slot (682) to the east of ditch 612 belonged to the phase of the outer ditches.

Thus the inner bank excavations, in addition to producing problems about the interpretation of the Neolithic enclosures, have shown that the longhouse settlement, and perhaps the Period 3 settlement, continued at least 130 m within the rampart on the line of the hillfort entrance."

The Fourth Report 1972:

By kind permission of PWD, extracts from the 1972 report. Figure 1 above shows the area excavated during the 1972 season. An aerial photograph taken during the season I posted earlier here:

"Summary

In the fourth year of excavations the area behind the hillfort entrance was stripped.  Here were found traces of six long aisled houses, arranged either side of the road from the entrance and associated with the first defences.  During the second period of fortification a large roundhouse was built in the same area.  Cuttings across the inner bank exposed more of the Neolithic enclosure, and showed that this part of the site was also occupied during the period of the hillfort.  Evidence from the pottery found so far suggests that the site was abandoned before the fifth century BC.

The work was again made possible by the generous support of the Gloucestershire College of Art and Design, in Cheltenham, and we are thanks to the owner of the land, County Councillor Tom Morris, for permitting us to excavate.  Muir-Hill Ltd, of Gloucester, lessees of the quarry, kindly allowed us access through it to the fort, and were most helpful to us.

We are most grateful to the County Valuer's Department of the Gloucestershire County Council for permitting us to live in the County Training Centre, Ullenwood, during the excavation, and to Mr and Mrs Marcinkiewicz for their constant kindness to us during our stay there.  We would like to take this opportunity of expressing our gratitude to the local and other firms who have helped us in various ways, including Chelhire Limited, its proprietor Mr E E Jasper, and Mr John Kear, whose special skill as a JCB operator has relieved that of many worries.  Bowring's of Cheltenham, printers of the text of this and other papers issued by us; Holton Studios in their careful preparation of the litho plates and illustrations; Fred Stephens Ltd; J. Jones and Son; Sharp and Fisher (Builders Merchants) Ltd; Central Motors, Gloucester; A.C. Hands Ltd; G. A. Willetts Ltd; Swanbrook Transport. All these have made our work considerably easier.

The success of the excavation depends on volunteers and adequate financing.  We should be most grateful to any contributions to supplement the generous help from the Gloucestershire College of Art and Design.  Any donations should be made out to the Crickley Hill Excavation Fund, and sent to the Secretary of the excavations, Mr R D A Savage, Gloucestershire College of Art and Design, Pittville, Cheltenham, Gloucestershire GL52 3JG, from whom further copies of these notes, and copies of the note on previous seasons, may be obtained (1969, 15p 1970 and 1971, 25p all plus 5p for postage etc).  Persons interested in taking part in future seasons of excavation in this series should get in touch with Mr Savage at the address given (telephone Cheltenham 32501).

Crickley Hill

Fourth Report 1972

The fourth season of excavations at Crickley Hill took place between the 30th June and 13th August, 1972.  During this period work was carried on by an average of 80 volunteers a week, and I am most grateful for their efforts both to them and to my site supervisors, Clive Anderson, Simon Bruton, Terry Courtney, Janet Dixon, Chris Gingell, Mike Hall, and Alice Pandrich, whose initials appear on the drawings for which they were responsible.

The Site

Crickley Hill is a spur of the Cotswold escarpment, about 4 miles south of Cheltenham and a mile to the north of Birdlip.  The flat-topped hill is bounded on two sides by cliff edges, and forms a roughly triangular area, whose third side is cut off by a rampart about 300 m long.  The excavations of 1969 on the area of the entrance through this rampart revealed four main periods of occupation: the rampart itself belonged to the second and third of these periods, and was accordingly identified as a Period 2 and Period 3, the latter subdivided because of reconstruction into Period 3a and Period 3b.  Period 4 described intermittent occupation until modern times after the abandonment of the hillfort as a fortification; structures below the first rampart were ascribed to Period 1.  The investigations of 1971 in the eroded bank about 120 m within the rampart compelled the formulation of a separate division into periods of both this area and the interior of the hillfort.  One result of the 1972 season has been to allow correlation of the periods across the site, and in the following account a revised system is adopted.

Phasing of the Site (1972)

Neolithic          

1a        Occupation, revealed by post holes, before the building of the first Neolithic bank

1b        The first enclosure (ditches 309, 311, 319 and 853 = 384)

1c        Occupation, revealed by post holes and a fragment of walling in the filled in Period 1b ditches.

1d        The second enclosure (ditch 304/603).

1e        Occupation, revealed by post holes below the first hillfort rampart.  This period may be identical with any of the preceding periods or maybe a separate occupation between the end of the second enclosure and the founding of the hillfort.

Hillfort

2          The first rampart, built with timbers lacing together the drystone walling, and an inturned entrance passage defended by two gates.

3a        After destruction by fire, the entrance was repaired and the front of the rampart rebuilt.

3b        The entrance was reconstructed to produce a much more formidable defence.

4          And the fort was again burnt, and abandoned; subsequent occupation involved no significant building.

...  this scheme is only interim, and may well be complicated by further digging."

There was a little bit more digging to be done... but only more than twenty seasons' worth...

To be continued/

The 1972 report ...

Truffling about in cyberspace a couple of weeks ago I found a secondhand bookseller offering a copy of "Crickley Hill Fourth Report 1972" for sale for a modest sum, so I snapped it up. It was described as an "ex-library copy": it turns out to be ex-Gloucestershire College of Arts and Tech Arts Library from the stamp on the reverse of the cover shown above. I'm not sure they should really be disposing of their copy, but I'm glad to give it a safe home. I suppose it's too much to expect a library to contain books and reports these days ... sounds of early-middle-aged grumbling, huffing, moaning, sighing ...