Castle Hill, Almondbury: A Brief Guide to the Excavations 1939-1972 (1973) by W.J. Varley

The following is the OCR text of a book and will likely contain errors. This page is designed to be indexed by search engines. Click on a page number to view the book in your web browser.

Please note that the text is not in the Public Domain and should not be reproduced further without the express permission of the copyright holder or their estate.

Page 1

soa mr 3h) et


A brief guide to the excavations 1939 - 1972


Director of Excavations

Huddersfield 1973


Page 2







Phase One Stage I Stage Il Stage Ill Stage IV Stage V Stage VI Stage VII Stage VIII Stage IX

Phase Two

Phase Three Stage I Stage Il Stage Ill Stage IV

The First Occupation The First Interregnum The Univallate Enclosure The Second Interregnum The Open Settlement The Univallate Fortlet The Bivallate Fort The Multivallate Fort The Burning

The Great Interregnum

The Medieval Earthworks The Hunting Lodge The Slighting The Final Buildings




28 30 33 34

Page 3


Fig. 1. Plan of Castle Hill, Almondbury, showing the areas excavated.

Fig. 2A. Plan of the Univallate Enclosure in relation to the hut-floors of the First Settlement.

Fig. 2B. Plan of the Univallate Fortlet in relation to the hut-floors of the Open Settlement.

Fig. 3. Reconstruction of Section 35. Fig. 4. Reeonstruction of the defences of the Univallate Fortlet.

Fig. 5. Reconstruction of the north-western junction of the Bivallate Extension with its predecessors.

Fig. 6. Reconstruction of the gateway through the Univallate Fortlet. Fig. 7. Plan of the Multivallate Fort. Fig. 8. Reconstruction of the Inner Rampart of the Multivallate Fort. Fig. 9. Plan of Castle Hill, Almondbury (Phase Three, Stage 1). Plate 1a.Unburnt inner wall, inner rampart of Multivallate Fort (Site 30

Plate 1b. Burnt core of same rampart.

Page 5


Huddersfield is twice fortunate, first to have within its boundaries a remarkable hill vantage point with exceptional historic associations and secondly to have an expert archaeologist interested enough to pursue long research and exploration into the chequered history of the site.

The enthusiam, determination and dedication of Dr. Varley in the cause of research and truth is beyond.praise. For him the work, spread over the major part of his adult life,has been a labour of love, with increased knowledge as his only reward, apart from the thanks of the many who will read the story with pleasure, visit Castle Hill with added interest and will have their imagination stirred and stimulated by the light thrown on life in past centuries and sparks Struck by the excavation pick, so skilfully wielded by Dr. Varley.

It has been my privilege, as Chairman of the Estate Committee, in a small way to facilitate, collaborate with and encourage Dr . Varley in his last three seasons of excavation and I am delighted to have this opportunity to add my personal thanks to him for all he has done

to the benefit of the ever increasing number of people interested in

times past. CLIFFORD STEPHENSON, Alderman and Chairman of the Corporation’s Estate Committee who are responsible for the care Almondbury Common, and well being of Castle Hill and

1973. its Tower.

Page 6


Some places, like some people, thrust themselves into the pages of history as if by right. It is taken for granted that their fame will be adequately chronicled and if it is exaggerated in the telling, no matter. Folk-heroes expect to become folk-myths.

We who were born within sight of that familiar land-mark have always believed that Castle Hill was just such a place, though we may not have known why. Camden, writing in 1584, did his best to enlighten us. He drew attention to “ those manifest tokens of former grandeur, a rampart, some ruins of a Castle which was guarded with a triple strength of forts and bulwarks”. He then went on to list some of the events for which Castle Hill was famous, so far as he had heard, or read. Apparently, it had been the site of Camulodunum which appeared on the world map of Ptolemy in the second century A.D.; the site of a town built by Saxon kings and a cathedral built by Paulinus and dedicated to St. Alban. The burning, the effects of which he had clearly seen with his own eyes, he regarded as the consequence of the war waged by Penda, the pagan Mercian, supported by Cadwalladr the Briton, against Edwin of Northumbria. This catalogue has never been equalled since, although none of us has been exactly exempt from flights of fancy.

By the time of my schooldays, speculation had polarised around two possibilities. The prehistorians, led by Dr. T.W. Woodhead, saw Castle Hill as a Brigantian stronghold, possibly the very place where Caratacus, the ancient British hero, was betrayed by his infamous kins- woman, Cartimandua, the Queen of the Parisii of the East Riding and the Brigantes of less certain location. The medievalists, flourishing their documents, were quite certain that Castle Hill was the site of a motte-and-bailey castle built in the reign of Stephen. Mrs. Armitage, writing in the august pages of the Victoria History of the County of York, Vol. |, 1907, said so, very clearly.

By 1939, negotiations led to the decision that the moment of truth should be expedited by a programme of excavation planned to occupy three seasons. The first ended summarily with the sounding of the first air-raid siren to announce the beginning of the Second World War. We resumed work in 1946 and 1947. We then rested, partly by accident, partly by design, until it became opportune to try to finish the job in 1969, 1970 and 1972.

Page 7

At long last, as the man in charge since the task began, I feel that I can now call ‘Enough’. Naturally, there is always more which could be done and better done, but as archaeology has still a lot more to learn, it would be sensible to wait until the new learning comes along after the customary respectful interval.

Such has been the measure of support and interest I have received from the public, directly, face-to face, on the spot, that I can do no less than write a brief summary of our results, intended for them. It is fitting that it should appear under the auspices of the Tolson Memoria! Museum whose successive Directors, Dr. Woodhead, Dr. Grainger and Mr. Aubrook, have always given me every possible assistance throughout this enter- prise. So too, I hasten to add, have the officials and staff of the Corpora- tion of Huddersfield, despite all their other preoccupations.

Acknowledgements are also due to Mr. J.L. Brooke for his kindness in providing the sketch of Castle Hill on the title page.

Finally, of course, we are all indebted to the trustees and guardians of the site, the Estate and Property Management Committee of the Corporation of Huddersfield and especially to their present Chairman, Alderman Clifford Stephenson.

This summary will, of course, be followed by a full, more technical report wherein, I hope, will be found the evidence for statements made here.

Page 8


What the visitor to Castle Hill sees today is no more than, as Camden said, “ the tokens of former grandeur, ramparts and some ruins of a castle”. The principal building of that castle stood in Camden’s time where the Jubilee Tower now stands. All that now remains of the other buildings are two sides of a former hall and a well-head which we have recently restored.

The rampart he referred to is the grass-covered earthen bank running round the edge of the hill-top some 875 feet above sea level, which falls into three sections. At the south-west end, facing Farnley Tyas, there is the Inner Ward, three acres in extent, which once contained the Castle keep, the seat of authority. This was separated from the Middle Ward by a deep, dry moat. The Middle Ward now contains the Castle Hill Hotel, its outbuildings and the decayed remnants of a Victorian bowling green on which some of us remember the game being played.

The Middle Ward is separated from an Outer Ward by another trans- verse ditch, part of which is now occupied by the track leading to the hotel. The main entrance of the Outer Ward lies at the north-east corner of the hill top (nearest to Almondbury) and is occupied in part by the modern steps which lead down to the road (Ashes Lane).

All three wards are now surrounded by a second rampart, running roughly parallel with the first. Beyond this again are the ruins of third and fourth banks, visible only between the road up to the hotel and the north-east entrance. This latter is prolonged by a hollow pathway which leads down to the bottom of the hill, albeit as a dog-legged track. To the right as you proceed downhill, and perched on a flattish shelf well below the top of the hill, there is a rectangular enclosure, surrounded by its own bank and filled-up ditch which we have called the ” Annexe”, for that is what it was, an annexe tacked on to the main fort above. Finally, round the bottom of the hill, and only visible in stretches, there runs a low, wide bank on each side of a filled-up ditch. Just inside the angle between the road to Farnley Tyas and LumbLane in the south-west corner of the hill, there is a slight hollow which is all that is left of the only entrance into the earthworks on this side. All these features belong to what we call the Visible Earthworks.

We now know that they all belong to the former “motte-and bailey” castle built on behalf of the de Laci family, Lords of the Honour of


Page 9

Pontefract and the Manor of Almondbury, and Earls of Lincoln, in the twelfth century A.D.

If that were all there were to say or prove, the task would be simple. But it most assuredly is not. Underneath and unseen until we uncovered them are a whole series of earlier defences and occupations which belong to prehistory, i.e. before the Roman Conquest, which is represented locally by the Roman fort at Slack. It is their existence which has taken so long to demonstrate, and although they do not, in fact, belong to anything that could properly be called a Brigantian stronghold, they do prove that the prehistorians were correct in thinking that Castle Hill had been occupied and defended long before the Romans landed.

As you will have noticed, what we now know is that both schools of thought about Castle Hil! were correct. The medievalists were right in attributing what they sawand wrote about to the medieval period. The prehistorians were right in their contention that Castle Hill had been defended and occupied long, long ago. I hope they are all content now.

The prehistorians, counting myself among them, never expected that early history to be so protracted as it was, or so far away in time, and none of us anticipated that it would be so difficult to unravel.

And if, alas, we have to say good-bye to many of our former illusions, e.g. Cartimandua, the Brigantes, the Saxon kings, the cathedral, Paulinus and Penda, we still have left the most ancient historical monument in this area of Yorkshire. You may prefer the fairy stories. Like Earl Baldwin before me, I am impressed not so much by the ‘diversity of testimony as by the many-sidedness of truth’. I trust that none of us who have toiled so long are vain or stupid enough to think we have seen all the truth, but I do hope we have unveiled at least some of the main outlines.

Page 10

Phase I Stage Event Date One I The First Occupation 2151 b.c.* aT The First Interregnum W The Univallate Enclosure IV The Second Interregnum V The Open Settlement vi The Univallate Fortlet 590 b.c.* Vu The Bivallate Fort 555 b.c.* Vit The Multivallate Fort 463 b.c.* IX The Burning 431 B.C. Two The Great Interregnum 431 B.C. - 1147 A.D. Three I The Medieval Earthwork 1147-1260 A.D. II The Hunting Lodge The Demolition After 1260 before 1340 A.D. IV The Final Building After 1340 A.D.

A table to show the principal events in the history of the Castle Hill earthworks.

Note: all the dates are quoted in calendar years. The dates marked with an asterisk have been determined by the radiocarbon technique and are based on the international convention that the half-life of the radio- active carbon 14 isotope is 5688 years reckoned backwards from 1950 A.D.

The level of radioactivity on which radiocarbon dating is based fluctuates slightly; therefore dates determined by this method may not always correspond exactly with true dates. To indicate the difference between the two kinds of dates archaeologists now use capital letters B.C. or A.D. for historically true dates and lower case letters b.c. or a.d. for radiocarbon dates.

Page 11


The areas excavated to date are shown in plan on Fig. 1. Collated from the results from all sources, the table opposite sets out the succession of structural and occupation periods, established by relative stratigraphy and dated, wherever possible, by one or other of the techniques now current for absolute dating. Where sufficient evidence from excavation data permits, I have attempted to project the structures of various periods in three-dimensional scale-drawings. The evidence in full will appear in my final report in a separate publication.

It will be seen that the table is arranged in chronological order. The brief description which follows is arranged in historical order, beginning with the first known defences. All the pre-Roman structures are regarded as belonging to one major phase of structural activity, labelled Phase One. This is succeeded by what I have called the Great Interregnum when, to the best of our knowledge, the site lay empty (Phase Two). This was followed by the medieval occupation in Phase Three, after which we enter the final post-medieval Phase, which has not been the subject of our investigation. The first and third phases are both complex, the first more so. The major structural or occupation events within the first and third phases have been regarded as stages, numbered in order of their occurrence.

Phase One, Stage |: The First Occupation

The first occupation of Castle Hill of which we have any proof was neither known nor suspected before we began work, simply because it is the deepest buried. The evidence has stared us in the face for several seasons, but it was only in 1972 that, thanks to scientists at Harwell, we have managed to put a date to it.

The signs are identical at several places. They consist of a couple of inches of charcoal, burnt bone, and decayed humus, resting on the undisturbed land surface at several places. Each exposure covers an area of about 8 yds. x 10 yds. underneath the lowest rampart, protruding front and rear therefrom, and separated from the rampart base by a few inches of sterile soil. These exposures are discontinuous, separated from one another by considerable gaps, but they lie at the very edge of the hill top. We have not found any inside any part of the

Page 12

interior. In one area only, so far, (Site 27 in the north-west corner) have we found any structures which could possibly be contemporary with the charcoal spreads. Here, the kerb of a hut, defined by stones let into the surface on which the floors rest, could therefore have been put there when the floors were in use As huts of this time were almost certainly built of inflammable thatch or wattle-and-daub, which catch fire easily, it is possible that the charcoal spreads represent the debris resulting from such fires. Alternatively, we know from elsewhere that in some instances the occupants of such huts carried out their cooking outside, presumably to diminish the risk of fire, as do some African tribes today.

The really important discovery, based on samples of ash from one of the floors taken from Site 40, is that they can be given a conventional radiocarbon date of 2151 b.c.* However much future research modifies that date, the fact remains that the occupation belongs to the third millennium B.C., called the Neolithic Occupation, in Britain. This was the period during which immigrants into Britain brought knowledge of how to make implements by polishing flint or stone, how to fashion needles and combs of bone and how to make pottery from loca! materials. They lived not merely by hunting, fishing and collecting wild grasses and berries. They could cultivate wheat and barley, and they had domestic animals such as sheep, goats, cattle, pigs and dogs, but not horses. They had distinctive ways of building family tombs and in some parts of Britain they built the embanked structures called causewayed camps, such as the famous Windmill Hill on Salisbury Plain. They are known to have inhabited the Peak District, the Yorkshire and Lincolnshire Wolds, and both the Bridestones, near Congleton and Five Wells, Taddington,near Millers Dale, are believed to be examples of their tombs. The attraction of Castle Hill is a little hard to see now except that it was an area of light soils with thin vegetational cover which could be easily cleared. Below were the heavy forests and swamps of the valleys, rich in game, then as later. A polished, volcanic stone axe, suitable for felling trees and digging soil, was found on Castle Hill, together with a few flints.

Phase One, Stage II: The First Interregnum

There is no evidence that these people built any defences on Castle Hill, or made pottery there, or buried their dead in the immed- iate locality. How long they remained there we have no idea, but eventually their settlement was abandoned, and a layer of soil shows that their hut-floors were overgrown for a very long time. I have called this period ” The First Interregnum”, i.e. the first natural break in the continuity of human occupation.

Page 13

--" ~~ “eee weer -- _ ws er


0 180 eed Ft

Fig. 2 A. Pian of the Univallate Enclosure, Castle Hill, Almondbury, in relation to the floors of he First Settl i i d

B. Plan of the Univallate Fortlet, Castle Hill, Almondbury, in relation to the hut-floors of he Open Settl di it ted

Page 14

Phase One, Stage III: The Univallate Enclosure

The lay-out of the first defences erected on Castle Hill is shown in Fig. 2A. An area of five acres at the southern end of the hill, inside the Small Transverse Ditch of today, was enclosed within a single bank which, with the notable exception of the northern portion, was placed at the edge of the summit plateau overlooking the steep natural slopes with which it is surrounded.

The northern boundary ran from north-west to south-east, bisecting the summit. It had a simple gap entrance achieved by leaving a natural causeway between the two halves of the bank. The banks underlay those which are still visible and the entrance (E1) is still used by all who come to the Castle Hill Hotel by car.

The single vallum (bank or rampart) was defined by two parallel rows of large flagstones, set on edge, upright, in two narrow slit-trenches, twelve feet apart. Between these limits there were two low stone banks, one outside, the other inside. The space between was filled with false clay, the soft top-surface of the weathered Elland Flags. We suspect there may have been a wooden palisade on top of the bank. There is no evidence of a contemporary ditch but there was a small guard-room with a central hearth on the south side of the entrance (Site 9). All that remains of it now are low cobbled walls, joining up with those, of the first bank. We believe that the occupants of these defences manned them from positions just behind the enclosing bank, and probably hidden by it.

Phase One, Stage IV: The Second Interregnum

There is now abundant evidence from several sites dug in 1972 that the first enclosure rampart, throughout its extent, fell into ruin and was covered with a second land surface. This second break in continuity was that I have called the Second Interregnum. We are still not certain of when it occurred or how long it lasted. We are certain that it was succeeded by a period of open settlement, i.e. a period when the site was occupied but not defended.

Phase One, Stage V: The Open Settlement

At certain places the land surface of the Second Interregnum was broken through to admit stone slabs, set obliquely on edge in slit trenches, identical with those of the Univallate Enclosure below. Elsewhere, the old foundations were carefully levelled out and covered up with a packed clay floor into which the hut-kerbs were inserted.

Page 15

Fig. 3 Reconstruction of Section 35, Castle Hill, Almondbury.

1: Bedrock. 2: Land surface of First interregnum. N-N: Neolithic Floor. -3: Rampart of Univallate Enclosure (U.E.). 4: Second 5: Hut-floor of Open Settlement. 6: Rampart of Univallate Fortlet (U.F.) ei

h: Hearth



Fig. 4 Reconstruction of the defences of the Univallate Fortlet, Castle Hill, Almondbury.

Page 16

They were probably circular, some twenty feet or more in diameter, They were covered inside with a thin scatter of ash and burnt bone, radiating from a central point, probably a hearth.

Their sites, so far as is known, were not confined to the original enclosure but existed outside it, though still on the top of the hill. Their relationship to what lay below and what lay above is shown in Fig. 3. Their distribution is shown in Fig. 2B.

We do not know what their superstructure was like, nor is there any occupation material definitely of the same period. We do not know whether their occupants were the descendants of the people who occupied the first defences, or why their ancestors or predecessors abandoned the site, or why the people who returned, if they did, chose to live without defences. All we do know is that this sequence of events took place in the order given.

We will now describe the successive defences which were built on Castle Hill until the burning in 431 B.C.

In the broadest terms, we see the beginning and end of changing ideas about what a particular community, living in a particular place at a particular time, considered appropriate for their defence. I say “community” advisedly because those who went to such trouble and labour must have included more than a single family and, judging from the results of their labour, defence must have been a growing preoccupation. Likewise a considerable portion of their resources in labour and skills must have been employed. We shall also try to throw some light on the final fate of that community which brought their occupation of this particular site to an end. Using scientific methods developed since 1950 we can date these structures accurately. So far as the Iron Age* occupation is concerned this has been the main justification for our protracted activities.

Phase One, Stage VI: The Univallate Fortlet The Open Settlement gave way in the southern half of the hill-top to a small five acre fort, nearly but not quite on top of its predecessor. The lay-out is shown in Fig. 2B. Fig 4 is a reconstruction of the defences, based upon many sections.

The rampart, like its predecessor of Stage One, lay between two parallel

* The Iron Age in Britain is considered to span the period 600 B.C. to A.D. 43 or later.

Page 17

rows of stone slabs, set upright, ten feet apart. Inside these were dry-stone walls sloping inwards, between which there was a core of false clay divided into cells by layers of stone leading out horizontally from a central pillar of piled-up stones. The top was stone-covered but in the centre were stone-lined post-holes at approximately ten foot intervals. I have suggested that these were the principals of a post-and-rail close-fence or palisade which would have increased the effective height of the defence.

Outside the front dry-stone wall of the inner rampart lay a narrow sloping platform (berm), beyond which was a wide V-shaped ditch with rounded bottom, cut in the solid rock. Beyond this again was a low clay bank, thrown up from the ditch, revetted on the outside with stone footings.

Two special features deserve note. Behind the back wall of the inner rampart there were occasional laid-clay floors carrying an inch or so of ash and burnt bone. One of these occupation floors used by the people who manned the defences yielded a few scraps of pottery which will be discussed later.

The second special! feature is that, in various places, the rampart core, normally of clay, contained occasional pieces of timber. One such instance is shown in Fig. 4 near to the inner face of the back wall. These pieces of timber had been cut from living trees, sessile oak from somewhere close at hand, but they were too small to have had any structural function in the position in which they were found. It is more likely that they had been taken from an earlier structure and thrown into the core as the rampart was being built. A sample taken from such a piece of timber was sent to Teledyne Isotopes of New Jersey, U.S.A., for radiocarbon dating. The date they arrived at was 590 b.c.*

The northern boundary of this small fort was remodelled by the addition of the Small Transverse Ditch with its own outer bank, as shown in Fig. 5, and an entrance on the same site as its predecessor, the original entrance to the Univallate Enclosure. Fig. 6 is an attempted reconstruction to show what that entrance probably looked like. Naturally, all such entrances had to be capable of being closed on Occasions, usually by means of a gate, as shown. Frequently, such entrances had a bridge over the gate to allow the defenders to pass from either half of the defences. to the other. We cannot be certain that the little guard room which belonged to the original Univallate Enclosure was actually retained in the new entrance on the same site, but it is included in the reconstruction.

Page 18




N NS ON / / aN a / NN \ \ \ YQ \ % eA 2 A SS aN a Bi cy

Lawes \) SS DSS oo qe aS ? Oo If PG

\\ \ A \ iN NA \


Page 20

Phase One, Stage VII: The Bivallate Fort

At some point in time after the Univallate Fortlet had been built and occupied, it was decided to extend the defences so as to enclose the whole of the plateau top within the 875’ contour, thereby doubling the defended area.

We were fortunate in finding the junction between the added defences and their predecessors in the north-west corner (Site 6). Excavation re- vealed what actually happened here, as shown in our reconstruction (Fig. 5).

The Small Transverse Ditch (which had been cut through part of a hut belonging to the Open Settlement resting on top of the foundations of the rampart of the Univallate Enclosure) was blocked up at the end. The inner rampart of the extension, i.e., the inner rampart of the Bivallate Fort, was carried on top of this blocking right round the western edge of the plateau, as far as the north-eastern end of the long axis. The same thing was done at the other end of the Small Transverse Ditch and the new inner rampart was carried round the eastern edge of the plateau top. The additions to the inner rampart were on the same scale as their predecessors, and like them, they had a berm, a ditch and a small second rampart, running paralle! with the inner rampart, lower down the slope. A feature of the additions, particularly on the east side, was that they rested on a base of clay with layers of turf at intervals of a foot, into which considerable amounts of timber had been inserted more or less as a raft, inside the rampart core. Otherwise, the rampart presented the same sort of appearance as its predecessor, namely, a stone-lined box, 10 feet wide, defined by large flagstones standing on edge, and internally braced by layers of stone. There were also stone-based, stone-walled shelters behind the inner wall of the inner rampart. In one such shelter, we found a hearth on which there were two lids and part of the base of rectangular ‘salt-pots’ (Site 4).

The old entrance was presumably abandoned and a new one, a simple gap, placed at the north-east end of the long axis (E 2).

A piece of timber from the core of the Inner Rampart was radiocarbon dated to 555 b.c.*

Stage VII probably followed Stage VI by a reasonable interval, say half a century, because if it had been intended originally to build a bivallate fort defending the whole hill-top by defences placed all the way round and overlooking all approaches, surely it would have been done initially and not as an afterthought. The third hill-fort makes. full use of the defensive advantages of the hill-side.

Page 21

@RSeD “38/404 Aema}ed ay} jo uo! 9 “64

Wy i tH tl} rt

i i} LJ I Ry HI cH HH HY]


@ 1 1 H Uy {| 8 a LH tJ

ry ti ® @ a

err ee Ca


) 9 ry y Y ae LY

I a v N ny


ry Hy



on Tete erry aa

/ if POE AS AIM ee WM 7" CT OTT mT uf Ht uy Puy lI An



(imi y

‘worsuarxg °Z — “| -Aunqpuowyy *suossaoepesd Sj! 19 '

jo uonoun! ug “BIg”

Page 22

Phase One, Stage VIII: The Multivallate Fort

The occupants produced an even more sophisticated set of defences in Stage VIII. As Fig. 7 shows, they remodelled the defences on the east side of the N.E. entrance by adding another pair of ramparts and ditches, linked by a bank connecting the revised second rampart with the new fourth. They left the original gap intact but added a pedestrian approach in the form of a defended Hollow Way leading to the bottom of the hill.

As Fig. 8 shows they also widened and heightened the inner rampart. The widening was achieved by taking in the berm and part of the earlier ditch to make a new rampart twenty four feet wide. They cut a new ditch further down the slope with a deeper V profile. The inner rampart was raised by adding more clay, stones and timber. The amount of fallen stones outside the rampart, but not derived from it, suggests that there may well have been a rampart walk behind a stone wall on the crest. Behind, there certainly were lean-to shelters, post- supported at each end, with burnt bone and ash on the floor. In many places round the hill-top there was no room for anything quite so intricate in the way of alterations but everywhere it was practicable, the inner and second ramparts were widened and heightened, and the ditches re-aligned; the object was presumably to strengthen the defences at the north-eastern approaches so as to enclose a flattish area just below. the main summit in this part of the site. Elsewhere, there was no room for any such addition.

A conventional radiocarbon date of 464 b.cwas provided by the Nuclear Physics Division of Harwell for the raised portion of the Inner Rampart of the Multivallate Fort. Ash from the floor immediately behind the back-wall of the same section yielded a conventional radiocarbon date of 447 b.c* for occupation behind these defences as distinct from the date of their construction.

Pottery from the shelters behind the inner rampart is a variety of flat-rimmed, native coarseware cooking pots, grey in colour and con- taining chalk grits which are not found in the local clay. It closely resembles pottery from various sites in East Yorkshire where chalk grits, used to stiffen coarse pottery from the Neolithic period onwards do occur. The salt-pot lids and base found a similar shelter behind the inner rampart at Site 4 likewise hint at a connection with the east coast since salt-workings of pre-Roman date have been found on the Lincolnshire coast at Ingoldsmells.

These facts need not necessarily imply that Castle Hill was

occupied by intruders from the east coast in the fifth and sixth centuries B.C. The pottery could have come by way of trade or barter.


Page 23

What the occupants of Castle Hill had to offer could have been skins since we found post-holes which could have been used to support skin-drying frames behind the inner rampart at Site A.

Finally, in our last season, 1972, we have at long last come to a conclusion about the remaining features of the Iron Age defences. Earlier versions of this short guide put forward the view that the Annexe, the Hollow Way and the Outer Ramparts, those which run round the base of the hill, were added to the main fort after the inner portions had been built. We now believe that all of them formed part of the Multivallate Fort.

I have previously referred to the Annexe as a quadrilateral enclosure set within a bank almost in the same style as that of the very first defences, namely a clay core lying between two parallel rubble walls. The Annexe possessed its own asymmetric ditch with an entrance mid- way along its eastern side. We placed this construction in the pre- Roman period, not merely because of the method of construction, but principally because the lower portion of bank and ditch were separated from an upper portion of quite different construction by a land surface which we were confident was the counterpart of that which we knew belonged to the Great Interregnum. In 1947, when the Annexe was first excavated, we were not aware that the Great Interregnum had two predecessors inside the main fort. In 1972, however, we discovered that the Annexe, which we had previously suspected of being only three-sided, had in fact a fourth side, a low bank, running parallel to, and made up of materials coming from, the ditch of the fourth rampart. The fourth rampart, far more extensive than this small portion, was of one construction upcast from the ditch outside it and so was the single bank of the Annexe. The fourth rampart, ditch and Annexe were therefore built at the same time, fell into decay at the same time and were covered by a land surface which came into being after the fourth rampart, Annexe, and main fort, fell into disuse. As the fourth rampart was not built until the multivallate fort was constructed, the land surface growing over both it and the Annexe must belong to the third or Great Interregnum, and not the second, or first.

Likewise, the Hollow Way is clearly a feature prolonging the north- eastern entrance far beyond the confines of the first form of that entrance which belongs, as we have seen, to the Bivallate Fort. On the eastern side, the bank which lies on the left side as one ascends the Hollow Way, joins up with the bank which protects the entrance where it runs through the fourth rampart.

Finally, but no less satisfactorily, we have to come to a conclusion about the Outer Series, the pair of ramparts running round the base of the hill with the suggestion of an entrance (E4) in the south-west

Page 24

Fig. 7.

Castle Hill, Almondbury

Plan of the Multivallate Fort,

A: Annexe.

£4: South-western entrance.

a: Site ot drying frames.

S: Shelters behind rampart.

Fig. 8 Reconstruction of the Inner Rampart of the Multivallate Fort, Castle Hill, Almondbury.

Page 25

corner. At present, this pair of ramparts is made up of revetted shale. There is no sign of any land surface separating an upper from a lower portion and the space between the two banks does not contain a ditch, silted or otherwise. They do not resemble, therefore, any of the other ramparts and ditches provenly pre-Roman. They could be, and in my view, probably are, the outer banks of the medieval earthwork, built in the same style.

Separating the two banks is a flat shelf, cut into the bed-rock, and now covered up to a depth of several feet with silt, run-off from the adjacent banks. These banks never rested on such a shelf, they merely reposed on the sloping ground. Flat shelves of these dimensions were used to form the base of the Iron Agethird and fourth ramparts, likewise built as shale banks, but which were invariably sealed from later medieval additions by the land surface of the Great Interregnum.

In conclusion, therefore, although we have now discarded the Outer Series from the inventory of Iron Age earthworks, we think there may have been at least a single bank running in the flat shelf between the invisible banks, and running round the base of the hill at about 800 ft. above sea level, to complete the pre-Roman multivallate defences.

For those who have merely examined the plans, the Annexe may appear to be no more than an eccentricity to satisfy a need no longer apparent. We know, of course, that originally it housed a two-roomed hut or bothy (Site 17).

Naturally, at this distance in time, we have no notion why the Annexe was built or what social need it satisfied. Enclosure for the animals and their custodian, private demense for a chieftain, or a compoundfor those without the law have all been suggested by my ingenious colleagues, but the siting and the relationship to the main multivallate fort make most use of the natural defensive advantages of the hill-fort.

Castle Hill is surrounded by slopes which are really steep, especially on the west, south, and south-east. From these points, Castle Hill stands up like the bows of a giant ship. But in the north- east quadrant of the oval, the ground falls in a series of gently sloping shelves. The fourth rampart itself is placed on the outer edge of one such shelf and commands a complete view of a second, the higher part of which contains the Annexe. The north-western rampart of this little enclosure itself commands a view of the long approach from even lower shelves. In other words, the Annexe defences prolong and extend the defended area in the least well naturally-defended area of the whole site. Likewise, the Hollow Way makes a right angle turn

Page 26

approaching the south-west corner of the Annexe, then links up with the single bank below.

The latter, could either have been the boundary of the community, rather like the linear earthworks of the Wolds, or a single line of defence placed round the bottom of the hill so as to throw the natural slopes into the total defensive lay-out. In terms of the contours of Castle Hill, the sequence of defensive lay-outs has its own logic. Equally, of course, it expresses genuine expansion. From the five-acre Univallate Fortlet to the far more extensive Multivallate Fort was a fairly far cry, and whether it resulted from genuine need, or desire for grandeur, it was a remarkable transform- ation, particularly in an area where such forms of grandeur were rare.

That it was a continuous progression to meet the needs of one community and its descendants seems clear. The decisive pr oof is not so much in the continued use of purely local materials in particular ways, e.g. the use of stone-revetted box ramparts for the last line of defence and the very distinctive use of flagstones in ramparts and shelters, which need reflect no more than the sensible use of purely local materials, but rather in the fact that there are no true breaks in construction represented by intermediate land surfaces. \t would appear that the builders of these successive fortifications were planning to stay. The reason for their not being able to do so needs now to be examined.

Phase One, Stage IX: The Burning

From the evidence of our very first section in 1939, it has always seemed likely that the pre-Roman Phase ended with a catastrophe. The difficulty has always lain in trying to decide whether it was a natural catastrophe or a deliberate slighting of the fort by an enemy.

Earlier, I took the view that it was the latter, and suggested that the burning was probably carried out by the Romans when they finally conquered this part of Britain and built their fort at Slack. We now have additional evidence which proves that this was not the case.

Dr. Martin Aitken, of the University of Oxford Institute of Applied Archaeology, applied the technique he has so successfully developed for the dating of pottery to the study of the burnt clay in our Inner Rampart at Site 31 and the answer he has arrived at is that the Inner Rampart of Castle Hill was burnt in 431 B.C. Even allowing for the customary margin of error (plus or minus 180 years) this date means that the prehistoric defences were burnt, and abandoned, long before

Page 27

the Romans came to this country. They found the site destroyed and overgrown; they did not destroy it.

Since we have a conventional radiocarbon date of 447 b.c* for an occupation floor immediately behind the same section of the Inner Rampart, and a second conventional radiocarbon date of 464 b.c.* for a piece of timber taken from the upper part of the same rampart, it would appear to be the case that the Inner Rampart of the Multivallate Fort was burnt not very long after it was built.

The question is, therefore, was the burning an accident or was it the handiwork of another hostile Celtic tribe?

Close study of the burning on the ground, by ourselves and by officials of the Yorkshire Division of the National Coal Board who have much experience of the effects of burning in coal waste tips, followed by studies in the laboratories of the Physics Department of the Huddersfield Polytechnic, directed by Mr. Dougherty, and supplemented by our own experiments, have shown that in the best authenticated sections (Site 31) the source of heat lay in or near to the timber raft inside the core of the rampart. Here the timbers were turned into charcoal and the clay adjacent to them was turned into purple slag in which the escaping gas had bored tiny holes, almost like those found in pumice. In the laboratory such effects can be achieved in identical materials at temperatures between 700° and 800°C. It is important to note that although these areas of burning were considerable they did not extend to the whole of the rampart core. Some, quite near, were only faintly discoloured, an effect which could be matched under laboratory tests at temperatures as low as 250°C. There were large areas of rampart core in which neither the timbers nor the clay were affected by heat so that they retain the texture and colour of the original materials. Furthermore, the stones on the outside and inside of this rampart in the same section were not burnt at all as our photographs show (Plates 1a and b).

Our own experiments in the field have shown that it is not possible by any means at our disposal to produce the effects observed inside the rampart by applying heat outside without causing the outer stones to be fire-reddened, fractured or otherwise changed. In other words, the sum total of our efforts has been to suggest that the ram- parts were not fired from outside, but that the source of heat lay buried inside the rampart core. If so, then the likeliest source of that heat was spontaneous combustion, generated by the timbers in the rampart and fed by oxygen contained in the clay core and the hollows between the stones.

In short, if this view is correct, and it certainly fits the facts

Page 28

available better than any alternative anyone has suggested, the burning at Castle Hill was accidental and not contrived by human agency.

Although this view rests mainly on the evidence provided by Section 31, it is supported from elsewhere on the site. Some areas are not burnt at all, even though they lay cheek to jowl with sections burnt to a cinder. Any visitor coming by road to the Castle Hill Hotel can confirm that fact for himself. The road passes right through a section of the inner rampart which is totally unburnt, visible on the right hand side as one ascends. Twenty yards on either side, the ram- part is burned to varying degrees.

Moreover, even in those areas which are most intensely burnt on the south and west of the hill, so much so that the timber inside the core has been reduced to charcoal, no sign of that burning is visible in the shelters or hearths which lie behind the inner face of the inner rampart. There is no discoloration of clay or stones except on or under the hearths themselves and the pottery fragments left lying on them when the site was finally abandoned were not affected by heat in any way.

But whatever the cause of the burning, there is no doubt of its effect. In our forty sections excavated, none were repaired after the burning. The burnt areas were allowed to tumble down into partial ruin and decay. This is proven by the fact that a new soil-profile, stained by the reddened debris beneath, developed over rampart remains and fallen debris alike. Likewise, there is no evidence that the interior of the burnt fort continued in use. After the burning, there is no suggestion that the occupants continued to live there with- out defences.

Phase Two: The Great Interregnum

This period of 1700 years lies between the burning of the Iron Age fort and the re-occupation of Castle Hill in the twelfth century A.D. It is represented on the ground by clearly defined and recog- nisable land surfaces which were formed by natural agency over the ancient ruins and which were only disturbed by the medieval builders where necessary. Everything below those land surfaces is pre-Roman, everything above is post-Roman.

Phase Three, Stage I: The Medieval Earthwork

The Great Interregnum was ended by the building of a medieval earthwork which in its ruined state is mainly what can be seen today.


Page 29

The lay-out is shown in Fig. 9.

To achieve it, a broad steep sided trench, thirty yards wide, ten yards deep, and ninety yards long, was dug from north-west to south- east at the south-western end of the hill. This dry moat (I.t.d.) isolated the south-western end of the hill and created an Inner Ward (1). Around this was placed a shale rampart carrying a stone curtain wall on the crest. The old inner ditch was recut and the second pre-Roman bank covered with a shale bank. On the highest portion of the knoll, a eep was erected where the Victoria Jubilee Tower now stands, and the occupants of this inner area were served by two wells, one under the tower itself, and the second in the south-east angle at the point marked W. The first well was discovered when the foundations of the Victoria Tower were being dug in 1897 and was explored only to a depth of 32 ft.

A second or Middle Bailey (M) was created by building the inner rampart all round the area between the Large Transverse Ditch (I.t.d.) and the Small Transverse Ditch (s.t.d.), recutting the pre-Roman ditch, and rebuilding the second rampart.

To the north-east of the Small Transverse Ditch, the first, second and fourth ramparts were rebuilt on the east side and the ditches recut. On the north-west side of the Outer Bailey (O) the inner and second ramparts were rebuilt, and the inner ditch re-cut. The entrance in the north-east was where it had been in Iron Age times. The Annexe ram- part was re-established and re-occupied, and the Outer Series of ram- parts were replaced by shale banks on a different alignment, with a new entrance at E4.

The medieval earthworks took the form of great wedges of shale over twenty feet wide, revetted front and rear with unmortared blocks of stone. No attempt was made to dig through the land surface formed in the Great Interregnum, presumably because the builders wanted to take maximum advantage of the extra height created by what lay underneath. The ditches were cut to a truncated V-profile, flagged at the base, usually a few feet above the bottom of the Iron Age ditch.

The date of the medieval earthworks as established by our excav- ations rests firstly on the radiocarbon date provided by Harwell for a wooden stake, one of a pair, found in its original position in a pit behind the medieval rampart on the east side of the Outer Bailey, and secondly by the date of a silver short-cross penny found in a second sealed pit. The stake is dated 1147 a.d* and the coin was minted about 1160 A.D. in the reign of Henry II. The date of the stake coincides exactly with that of the return to the de Laci family

Page 30

of the Honour of Pontefract and Manor of Almondbury. The archaeological evidence thus appears to confirm the documentary since the castle at Almondbury is referred to in a grant of Stephen to Henry de Laci some- time between 1142 - 1154 A.D.

A complete resistivity survey by Mr. Arnold Aspinall of the University of Bradford, tested by our excavation, has shown that considerable areas of the Outer Bailey were occupied in the twelfth century initially but were then sealed off by filling in the many pits and covering them with a paving of stones, preparatory to putting this part of the interior into arable cultivation. A quarter of a short-cross silver penny, the original farthing, found in the self-same pit as the earlier coin, was dated to the reign of Henry Ill, being minted by William the moneyer in London about 1251 A.D.

The abandonment of the outer bailey for its original purpose and especially the sealing of the occupied areas suggests that there was a major change in the function of the medieval earthworks towards the end of the thirteenth century.

That change can best be demonstrated by the story of the great well- shaft in the south-eastern angle of the inner bailey(W).

Phase Three, Stage II: The Hunting Lodge

The mouth of the well was discovered some eight feet or more below the present ground level. Inside a rectangular well-head built of dressed stones, there lay a rectangular shaft, 6 ft. 10 ins. x 7 ft. 3 ins., cut as a tapering rectangle, to a depth of 74 ft. 6 ins., from which point it became a cylinder 6 ft. 3 ins. in diameter at 77 ft. 9 ins. Thence it tapered smoothly to end in a cylindrical sump, 4 ft. 10 ins. in diameter and the same height. From the bottom came two wooden buckets, now restored, and the up-ended staves of a circular wooden casing which could only have fitted the mouth of the well. These features, and probably a beam now broken across a central groove around which a rope had probably been used to haul the buckets up and down, clearly belonged to the time when the well was in use. A blackened water mark showed that the well had never held more than 10 ft. 6 ins. depth of water. Presumably it was abandoned at the same time as the contemporary earthworks. Thereafter, the well was filled with clean silt interspersed with thick black deposits at 50 ft. and 32 ft. respectively, full of broken cooking pots and the remains of animals involved in the chase and of their quarry. There were skulls of fifteen dogs of assorted breeds and sizes, but five at least were large breeds with powerful jaws capable of tackling wild boar


Page 31

Plate 1a. Unburnt inner wall, inner rampart of Multivallate Fort (Site 30)

Plate 1b, Burnt core of same rampart.

Page 32

Mir ---

Fig. 9 Plan of Castle Hill, ee en (Phase Three, Stage 1) Key A: Annexe. 1, 2, 3, 4 ale Banks. Q Outer Ward. ae ‘Middle Ward. I Inner were td. Seal transverse ditch. I.t.d.: w: Well K: Knoll on ch castle s tood. £:, South-west entrance, 0.S.: Outer series of rai

Page 33

and stags. The animals hunted comprised red, roe and fallow deer, wild boar, badger, partridge, woodcock, woodpigeon, dove and a greylag goose. In addition to the wild animals, there were bones of domestic animals, pigs, oxen, fowls, ducks and geese. Falconry was indicated by the bone of a male goshawk and a falconer’s bell.

These facts suggest that the site was then used occasionally for hunting parties and the use of the well for its proper purpose was abandoned. The pottery found in the well up to this point in its history was mainly twelfth and thirteenth century in form and Mr. Aspinall’s analysis of the elements in the clay of that potter, compared with analyses of medieval pottery from several other sources, makes it virtually certain that the Castle Hill pottery was made from the same clay as that fired in the kilns at Upper Heaton, excavated by Mr. Manby.

It needs to be noted that this stage was not accompanied by any sign of the wholesale demolition of buildings.

Phase Three, Stage III: The Slighting

From 32 ft. upwards, the well had been deliberately filled in with building debris including ashlar, stone mullions and part of a stone spiral staircase implying the presence nearby of substantial stone buildings of more than a single storey.

The demolition did not stop with the filling of the shaft. Debris continued to be piled up in the form of a mound some 20 ft. or so high in the centre near where the Victoria Tower now stands. On the edges of this mound further buildings were then erected (see stage IV below).

Close dating of the changes is not easy. The abandonment of the Outer Bailey for its original purpose, and its transformation into arable land about 1260 A.D. would appear to fit best with the con- tinued use of the site for occasional hunting forays when it pleased the Lords of Pontefract to engage in that pastime within the Manor of Almondbury. Presumably the keep and its inner well were retained to meet their needs, the buildings meanwhile being occupied by a bailiff and his staff.

The demolition was obviously meant to be final. The purpose must have been to remove any suggestion of what people normally meant by a castle standing on this site, thus putting an end to what must have been a symbol of authority. There is evidence that the Castle still functioned in 1307 A.D. but it is referred to in the past tense in

Page 34

1340 A.D. It seems probable that the demolition took place within those limits. A likely date could have been when the Honour of Pontefract passed from the Earl of Lancaster, who had married a de Laci, to the Crown after the Earl's execution at Pontefract in 1322 A.D

Phase Three, Stage IV: The Final Buildings

The slighting was. followed by the erection of one or more buildings on top of the debris. One such has already been uncovered and has been left on view. Conservation now concluded has revealed further information about the date of these buildings and their function. At the moment we are disposed to think they had some connection with the continued administration of the Castle estate, possibly when it was returned to the Duchy of Lancaster, with whom it remained the accession of Henry IV in 1399 A.D.

So far as I am concerned, their erection brings down the curtain. As mentioned previously, regrettably, no doubt, to the romantically inclined, we have had to discard many of the hypotheses concerning the date and function of Castle Hill which have circulated, even in modern times. If the pre-Roman hill fort fell into disuse as late even as the third century B.C., then the people who lived in it were not the Brigantes. That name belongs to the last century or so before Slack was built. The coin which was once thought to have been found at Castle Hill and which was presumed to have been struck for Cartimandua was actually found at Honley and bears the abbreviation Cart for Cartivel and as Mr. Allen has shown it belongs to a family of Belgic coinage on the outer fringe of the area in which such coin- age was circulating in the first century A.D. Cartimandua’s story belongs to Tacitus but not so far as I know to Castle Hill. And if Camulodunum really was Castle Hill, as my old friend the late Sir lan Richmond, P.S.A. always maintained, followed by the editors of the Map of Roman Britain published by the Ordnance Survey, it was not as a Roman station, fort, town or villa, but merely as a salute to an ancient legend. Penda, the pagan Mercian, may well have burnt down many places but Castle Hill was not among them. There is no hint of Saxon occupation and the cathedral of Paulinus has failed to appear, as yet. Obviously, the medievalists were not wrong to cling to their beliefs, but neither were the prehistorians. At the moment of writing this brief account, honours are even, so to say, but if there are still genies on that site, I at least am prepared to give them best in the unequal struggle we have waged to wring their secrets from them. And if there should be a hereafter, I shall look forward to hearing the full story from those who really know it.

Return to the Huddersfield Exposed home page