s^n [\im Be van An Archaeological Survey of Herfordshire bris 3DEN THE LIBRARY OF THE UNIVERSITY OF CALIFORNIA LOS ANGELES AN AECHAEOLOGICAL SURVEY OF HEREFORDSHIRE. THE EEV. J. 0. BEVAN, M.A., F.S.A.; JAMES DAVIES, ESQ.; AND F. HAVERFIELD, ESQ., M.A., F.S.A., MEMBIiRS OF THE WOOLHOPB NATURALISTS' FIKLD CLUB. ..S.i5^^?'S>i WESTMINSTKli : i'iUNTED BY NICHOLS AND SONS', 25, PARLIAMENT STREET. 1896. AN ARCHAEOLOGICAL SURVEY OF HEREFORDSHIRE THE REV. .J. 0. BEVAN, M.A., P.S.A.; JAMES DAVIES, ESQ.; AND F. HAVEEFIELD, ESQ., M.A., F.S.A., ME.VIBKRS OP THE WOOLIIOPE NATURALISTS' FIELD CLUB. i1^*©©^^ ■ ■' : oi a WESTMINSTER : rrilNTED BY NICHOLS AND SONS, 25, PARLIAMENT STREET. ISUG. ^DA 67 O Mh& AN AKCHAEOLOGICAL SURVEY OF HEREFORDSHIRE. By the Bev. J. 0. Sevan, M.A., F.S.A. ; James Davies, Esq. ; and F. Haverfield, Esq., M.A., F.8.A., members of the Woolhope Naturalists' Field CInh. Read November 23, 1893. PREFACE. The idea of carrying out an Archaeological Survey of the pre-Conquest antiquities existing in the county of Hereford seems to have occurred simultaneously to several persons in the course of the year 1891. In the month of July of that year, a meeting of the Woolhope Field Club vras held at Llanthony, and the pro- posal to draw up such a survey was formally put forward by the Rev. J. 0. Bevan, at that time vicar of Vowchurch. Shortly after this, a committee was appointed to carry out the scheme, and two secretaries, Mr. James Davies, soHcitor, of Hereford, and Mr. Bevan, were deputed to send a circular, with appropriate forms, to all persons in the county who were likely to help in preparing the survey. The circular, unfortunately, received little attention: the forms were not returned to any great extent, and the burden of the work fell on the two secretaries. They examined the printed literature on the archaeology of the county, worked through the Ordnance maps, collected first-hand information, and drew up a list of "finds." r This list was revised by another member of the Woolhope Club, Mr. F. Haverfield, Student and Tutor of Christ Church, Oxford, who also compiled the accompanying sketch of ancient Herefordshire, and the result is now presented to archaeologists, and in particular to the members of the Woolhope Field Club and of the Society of Antiquaries of London. The survey is drawn up in two forms, that of an index of places and enumera- tion of the chief objects found at them, and, in addition, that of a map, with suitable coloured symbols, summarising for the eye the general purport of the index. In general, the plan of the index is identical with that used in the other county surveys lately issued by the Society of Antiquaries. A brief bibliography, 2 A7i Archaeological Survey of Herefordshire. with abbreviations, precedes the actual index, while, for the purposes of identifi- cation of place, the numbers of the 6-inch Ordnance sheets have been placed after all the place-names. Two appendices, by Mr. Haverfield, contain (1) some details relating to the roads, which it was difficult to incorporate in the alphabetical index, and (2) some notes on mounds, possibly Saxon burhs, which form a con- spicuous feature among Hereford earthworks. In accordance with the rules of the Survey, neolithic relics have not been noted. It may be said here that very few have been recorded in the county. As in other Surveys of this series, the symbol d) denotes vmdateable earthworks, some of which may of course be later than the Conquest. In conclusion, the compilers desire to return sincere thanks to those many correspondents who have furnished them with information in relation to their work, as well as to those friends who contributed to the fund raised in Herefordshire to defray the necessary initial expenses, such as the purchase of Ordnance maps, and to the Society of Antiquaries, at whose expense the map and index are now printed. ANCIENT HEREFORDSHIRE. 1. The county of Hereford lies between the English midland plain and the highlands of South Wales ; its early fortunes were largely influenced alike by its geographical position and its internal configuration. By position it is a border land between hill and plain ; before the Roman conquest it was, perhaps, a border between two British tribes, and in historical times it long formed the limit of the English advance into the difficult uplands beyond the Severn. From the first settlement of the West Saxons in the seventh century, until the final conquest of Wales, Herefordshire was a scene of border warfare and its inhabitants were to some extent organised as borderers. As late as 1406 Owen Grlendower swept across it with his French allies, burnt the suburbs of Worcester, and, according to local theory or tradition, encamped for eight days on the Herefordshire Beacon. The internal configuration of the county has affected its history no less than has its position. Steep wooded hills cover a great part of it, and sometimes reach a considerable elevation ; the valleys are narrow and secluded, and the rivers, descending from the Welsh mountains, partake the nature of mountain-streams in the swiftness of their currents, the suddenness of their rise and fall, and their impatience of bridge and boat. In early times the land had few attractions for ordinary settlers ; the fertile plain round Hereford, the mining districts round Ross, alone offered encouragement to the arts of peace. Hence the charac- An Archaeological Survey of Herefordshire. 3 teristics whicli appear most markedly in tlie archaeological record of early Here- fordsliire. The remains discovered are few, and occur principally near Hereford or near Ross ; it would appear that, in the British, Roman, and Saxon periods, the county was thinly inhabited and imperfectly developed. 2. One feature breaks the uniformity and contradicts the barrenness of the archaeological record. Herefordshire is a land of camps. The enthusiasm of antiquaries has noted more than seventy earthworks and entrenchments of various kinds, and though some of these are insignificant in themselves, and others are erroneously considered artificial, the real examples are numerous enough to form a striking feature in a county scarcely thirty miles wide and forty miles long. It is very difficult, unfortunately, to interpret these remains. They have usually been ascribed to the Britons and connected with the wars waged about A.D. 45 — 55 between the native Silures and the invading Romans. Local patriotism has even been rash enough to recognise in Oyster Hill and Capler Camp the names of the Roman general Ostorius Scapula, and to trace on Coxwall Knoll the last struggles of Oaratacus. This is, of course, an error. Six counties have been the scenes of such identifications ; in no county have they any justifica- tion. Briton and Roman doubtless fought in these districts, but the wise student will renounce the pleasure of pointing out their battle-fields. The narrative of Tacitus, on which we mainly depend for our knowledge of these struggles, is notoriously inexact in all details of warfare and of geograjahy, and the camps themselves do not aid the inquirer. Only one of them has l^eeu excavated ; very few have yielded characteristic remains. But the excavation of the Here- fordshire Beacon" shows that part at least of that camp belongs to the middle ages ; the coins found at Ivington testify to occupation in the fourteenth century ; Aconbury and Dinedor were occupied during the civil wars, and the history of the county shows that earthworks may have been used or erected in it at any period down to the seventeenth century''. One class of small earthworks can, however, be separated from the general crowd. The mounds, which occur frequently in Here- fordshire near the church and the court-farm, are doubtless the mounds of Saxon burhs or Norman castles, and indicate by their mere number the position of the county as the frontier against Wales, both before and after the Conquest." 3. In the British period, Herefordshire probably belonged partly to the Silures, " Pitt-Rivers in Archaeolugia, xlvii. 435. F. G. Hilton Price in Authwjiological Journal, xi. (1881) 330. *• W. T. Watkin (Archaeological Journal, xxxiv. 366 — 369) draws out a long list of Roman earth- works in Herefordshire. There is no proof that any item in this list is reallj' Rom;in. " Sec ApjJendix II. a 2 4 An Archaeological Survey of Herefordshire. T^hose special territory was Monmoutlisliire and Glamorgan, partly to the Dobuni, who inhabited Gloucestershire, and who may have reached westwards as far as Ken Chester." The land was thinly populated. Discoveries of coins and other objects suggest that British towns or villages existed at Kenchester in the Here- fordshire plain, and at Weston near Ross, while a few camps like Credenhill above Kenchester, Wall Hills near Ledbury, Caplar, Oldbury, Risbury, may be ascribed to the same people. But until excavations or chance discoveries have provided us with more evidence, it will be necessary, as I have said, to observe the greatest caution in discussing the Herefordshire camps. 4. The recorded Roman remains are more numerous, though not actually abundant. They occur principally (i.) along the one certain Roman road of the county, and (ii.) in the vicinity of Ross. (i.) The Roman road which ran from Viroconium (Wroxeter) to Isca (Caerleon on Usk) crosses the county from north to south.'' It can still be traced with con- siderable precision, under the name of Watling Street, north of the Wye, and of Stoney Street, south of that river. It passes two " stations," Leintwardine and Ken- chester. At Leintwardine, near the Shropshire border, the researches of two local antiquaries have established the existence of a rectangular entrenchment, fourteen acres in extent, within the area of which Roman tiles, coins, potterj^, etc. have been dug up. The mileage of the Antonine Itinerary shows that a " station," called Bravonium," was situate in the neighbourhood of this place, and we may plausibly argue, as Reynolds long ago conjectured, that Bravonium is Leintwar- dine ; it was, in any case, a small and unimportant place. Kenchester, near the Wye, about four miles west of Hereford, has yielded more considerable evidence of Roman occupation. It appears to have been a small town, in shape an irregular hexagon, with an area of some seventeen acres, surrounded by a stone wall with four gates. The principal street, fifteen feet wide, ran from east to west; the houses contained tessellated jjavements, hypocausts, leaden and tile drains ; coins of various periods, fibula3 (one of silver), glass, pottery, and the like abound, wliile two inscriptions (one dated a.d. 283) lend a distinctive Roman colouring. Suburbs lay outside ; a mile to the west was a "villa" at Bishop- ' The letters E P C D on tlie milestone of Numerian, found at Kenchester (which I have re-examined), may possibly stand for respuhlica civitatis Dohunorum. See my Boman Inscriptions in Britain, part iv. '' Ititi. p. 484: Isca Burrin, mpm. viii, ; Gobannio (Abergavenny), mpm. xii. ; Magnis, mpm. xxii. ; Bravonio, mpm. xxiii. ; Viroconio, mpm. xxvii. See Appendix I. § 1. ■= Bravonium seems to be the spelling of the better MSS. of the Itinerary. The Brannogenium of Ptolemy tuul the Ravenna Geogr. 427, 3, was a town of the Ordovices and probably much further noith than Leintwardine; it cannot, therefore, be another spelling of Bravonium. A71 Archaeological Survey of Herefurdshire. 5 stone, celebrated by Wordsworth in an indifferent sonnet. Tlie town, tbough small, had pretensions to comfort and civilisation, and is the only important Romano-British site in the county ; it lies under the shadow of Credenhill, and may have succeeded a British oppidimi. In itself it represents more probably the Romanised Briton than the Roman. The mileage of the Itinerary permits us to identify it with Magni (or MagnaY, a name which the wholly unmilitary character of the place forbids us to expand into Magna Gastra. Camden, Stukeley and others supposed it to be Ariconmm, but the identification with Magn{a), first pro- posed by Horsley, seems tolerably certain. (ii.) In tlie south-east of the county, at Weston near Ross, we have con- siderable remains of a town or village, which may have existed in British times, and was certainly connected with the iron mines of the neighbouring Forest of Dean. Its Roman name may perhaps have been Ariconmm. The Itinerary mentions a road from Glevum to Isca, which must have passed through south Herefordshire and, though no traces of this road exist, one of its "stations," Ariconium, may well have been at Weston. It is even possible that the name actually survives in Archenfield, the name since Saxon times of part of southern Herefordshire." Three large hoards of coins, " scorite " and traces of mining," and a villa, also mark this district as inhabited under Roman rule. (iii.) A few detached Roman sites remain to be mentioned. At Blackwardine, near Stoke Prior, skeletons, pottery, coins, and remains resembling kilns were found in making the railway. At Stretton Grrandison,'' some dwelling or dwellings " In the Itinerary, as in the Ravenna Geogr., we have only the form Magnis, presumably from a nominative Magni or Magna. The name reappears on Hadrian's wall at Carvoran, and is perhaps Keltic, connected (Prof. Rhys tells me) with " maen " a stone. As " maen " is masculine, the nom. of Magnis may have been Magni. The name may survive in " Magonsetun " the oldest recorded form (a.d. 811), as IVlr. W. H. Stevenson tells me, of the name of the English Magesaetas who settled in Herefordshire and who plainly took their name from some place. ^ Bin. p. 48.5 : ab Isca Burrio, mpni. viii ; Blestio (h Monmouth), mpm. xi. ; Ariconio, mpm. xi. : Grlevo, mpm. xv. Ai-chenfield appears in Domesday as Arcenefelde, in the time of Edward II. (a.d. 1316) as Irchinfield, in the A. S. Chronicle, a.d. 915, as Ircingafeldes (various readings : Yrcingafeld, lercingafeld). Professor Napier tells me that this latter can be phonetically connected with Ariconium, though he considers the element inga as possibly indicative rather of a Saxon dei-iva- tion ; in Welsh it is Erging (Liher Landavensii, Red Book of Hergest) or Urcenevefeld (Liher Land.) ; Leland spells it Herchinfield. It is ciu-ious that the deanery of Archenfield (the only modern survival of the name in use) does not include Ross or Weston. " T. Wright, Wanderings of an Antiquary, mentions " scoria; " in the following parishes of Here- foi'dshire : Ganarew, Walford-on-Wye, Whitchurch, Goodrich, Bridstow, Peterstow, Weston, Llan- garran, Hentland, St. Weoiiards, Tretire. The mining district is indicated in the map. "^ Identified by some writers with Circutium (or Circutio) mentioned by the Ravenna Geogr. The identification is absolutely witliout proof or probability. G A71 Archaeological Survey of Herefordshire. existed, to which we owe the names Blacklands and Cromwell's Walls, and a few small objects have been discovered; a road, still traceable, connected this spot with Kenchester. Further south a villa has been noted, but not explored, at Putley, while on the south-west, near Walterstone, another villa stands close to Stoney Street on the border of Monmouthshire. The sum of Roman remains in Herefordshire amounts, therefore, to two small towns, one insignificant "station," five villas or inhabited spots, some mining industry, and one certain road, the Watling and Stoney Street, in total a scanty sum, which is no doubt due in part to the unfavourable physical features of the county. Certainly the district played no prominent part in Roman Britain. Its datable remains, the Kenchester milestone and the coins, belong mostly to what is styled the Lower Empire. There are no traces of troops, or of permanent garrisons, or of the Roman civil administration, or of definitely Roman municipal life ; we deal in Herefordshire with the Romanised Briton, not with the Roman. The period between the departure of the Romans and the coming of the English is an entire blank. Names of villages and dedications of churches suggest that the British held parts of the county as late as the tenth century, and many Saxon place-names {e.g. Kington) testify similarly to a late-Saxon occupa- tion ; there is reason also to connect the British saint Diibricius (about 600 a.d.) with the district of Archenfield. Beyond this, all is conjectural or worse. The compiler of the Liber Ln.nclavensis, either Urban or his nephew Geoffrey of Mon- mouth, includes Archenfield in the diocese of Llandaff and mentions a " college " founded by Dubricius at Hennlan on the Wye, but he is more concerned with a twelfth-century controversy than with history, and his statements do not merit full credence. Some writers, including the Rev. W. J. Rees, identify Hennlan with Hentland and find the " college " in some ruins at Llanfrothen in that parish, but the ruins are unexplored, and later scholars place Hennlan at Dixton in Monmouthshire." There is still less reason to accept the statement of the lolo MSS. that there was a British bishop at Hereford in a.d. 601. 5. With the English invasion we pass into more fully recorded history, but our archaeological material remains as scanty as ever. Early in the seventh century,'' » See W. J. Rees, Liher Landavensis (Llandovery, 1840), p. 324; J. G. Evans, The Book of LI an T)'tv (Oxford, 1893), index s.v. Hennlan. The Herefordshire Hentland is mentioned in one of the cliarters of the Book of Llandaff. •> Green, with perhaps unintentional precision, dates the English occupation of Herefordshire to A.D. 626-634 {Makimj of England, maps on pp. 261, 272). Duncumb (i. 222) places the synod of Hertford at Hereford, and on the whole our historians do not help us much. — Credenhill appears An Archaeological Survey of Herefordshire. 7 tte West Saxon Hecanas and Magesaetas" crossed the Severn and tlirust tLemselves between Wales and Mercia, with which hitter they soon became incorporated. Here- ford was shortly founded ; about 676 we meet with its first bishop. In the next century Offa, king of Mercia, constructed his frontier dyke from the Dee to the Wye. Portions of this dyke are still visible in Herefordshire, at Kington on the Radnor border, and at Bridge Sollers on the Wye near Kenchester," but its best preserved portions lie outside our county, and its problems do not specially concern the Herefordshire antiquary. It may be stated, however, that there is no evidence in the county for any continuation of the dyke south of Bridge Sollers ; from that point onwards through Herefordshire the Wye appears to have formed the frontier itself. There is equally little evidence to decide whether the dyke was erected by Offa, or whether, as Ormerod and Hughes have conjectured, that prince adapted it out of earlier earthworks ; that problem can only be solved by excavation in other counties. Besides Ofl'a's dyke we have, according to tradition, at Sutton Walls the remains of a Mercian royal residence, occupied about a.d. 800,° but with that our list of Saxon remains immediately closes, and our survey is brought to an end. F. Havebfield. to be called after some Creda ; one Creda died in a.d. 593 (A. S. Chi'on.) and has been assumed to be the first king of Mercia (Henry uf Huntingdon, Roll Series, p. 53 note). The name itself seems to be West Saxon and is otherwise known ; it was not, therefore, invented out of Credenhill. " On the derivation of this name, see note " to p. 5. * The Kington fragment \\inds along the hills in a generally S.E. direction near Discoyd, over Evenjobb Hill, Middler wood, Rushock Hill, to the neighbourhood of Knill (0. S. x. N.W., S.W., S.E.). It crosses Lyonshall park and vanishes in Rusholme marsh (xvii. N.E.. S.E., S.W.). It reappears near Upperton, crosses Bishopstone and Byford Hill, ending on the Wye, half a mile west of Bridge Sollers (xxv. S.E., xxxii. N.E), where it marks a parish boundary. South of the Wye no trace can be found, the course indicated by Sir F. Meyrick (OentlemMi's Magazine, cii. 2, 580, A.D. 1832) being wholly imaginary. Recently the late Sir J. Maclean (Bristol and Gloucester Arch. Soc. xviii. (1894) 19-31) has suggested that the dyke ran fi-om Bridge Sollers along the left bank of the Wye, but the proofs given are veiy ir adequate, and it seems to the present writer far more probable that the Wye itself was the frontier. Whether it was the frontier only in Herefordshire or throughout its course to the Bristol Channel, is not easy to decide. " This is a conjecture, based on the accounts which Giraldiis Cambrensis, Richard of Cirencester, and other late writers give of St. Ethelbert's murder in 794. According to these accounts, Ethelbert was killed at " Villa australis," and buried first at Harden on the Lugg, then at Hereford. Giraldus was a canon of Hereford, and may well have meant " Villa australis " for Sutton — certainly it was so identified as long ago as Leland (Itinerary, viii. 87). But the earlier accounts of Ethelbert's murder (A. S. Chronicle, Florence of Worcester) mention no place. The Ethelbert legend was connected with Hereford as early as about 1030, when a church was dedicated in honour of the saint : it is not impossible that Giraldus, in medieval fashion, localised the details of the legend without historical warrant. The remains at Sutton Walls have never been excavated. BIBLIOGRAPHY. KEY TO ABBREVIATIONS. A. A. B. C. A. B. I. A. C. A. J. A. s. I. B. A. A. B. B. Clark D. H. Ephem. Fosbroke. Archaeologia. Society of Anti- quaries of London. Ancient British Coins. By Sir John Evans. Ancient Bronze Implements of Great Britain. By Sir Jolin Evans. Archaeologia Cambrensis (series and volume quoted). Archaeological Journal. Royal Archaeological Institute. Ancient Stone Implements. By Sir John Evans. Journal of the British Archaeolo- gical Association. Beauties of England and Wales. By E. W. Brayley and J. Britton. Vol. vi. 1805. Medieval Military Architecture. By G. T. Clark. 2 vols. History of Herefordshire. By the Rev. J. Duncumb (I. II. 1804), continued by the late W. H. Cooke (III. IV. 1882-1892). History from Marble. By Dingley (or Dineley). Camden Society, 1867. Ephemeris Epigraphica. Vol. vii. (Berlin, 1888). Supplement by Haverfield to I. B. L. Ariconensia. By T. D. Fosbroke. Ross, 1821. H. B. R. Britannia Romana. By John Horsley. 1732. G. C. Camden's Britannia, with Supple- ments by R. Gough. Vol. iii. 1806. G. M. Gentleman's Magazine. Heath. Excursion dovrn the Wye. By Charles Heath. Monmouth. 1799. I. B. L. Corpus Inscriptionum Latinarum. Vol. vii. Bi-itannia (Hiibner). Berlin. 1873. I.e. Itinerarium Curiosum. By William Stukeley (ed. 1, 1724; ed. 2, 1776). J. Anthrop. Journal of the Anthi'opological Institute. L. I. Leland's Itinerary. Ed. Hearne. Vol. V. 1711. N. C. Numismatic Chronicle. 0. S. Ordnance Survey (.six inch). Price. History of Leominster. By Price. Ludlow. 1795. Roy. Military Antiquities. By General Roy. 1793. W. C. T. Transactions of the Woolhope Naturalists' Field Club. Williams. History of Leominster. By the Rev. J. Williams. Leominster. 1808. W. W. Wanderings of an Antiquary. By T. Wright. 1854. Note : — The reader will find two references constantly recurring in the following index, A. J. xxxiv. 349-372 and W. C. T., 1882, 236-258. These refer to two comprehensive articles on Roman remains in Herefordshire, the first by the late W. Thompson Watkin, the second by the late Dr. Bull. The conclusions of these two writers have not been by any means always adopted, but it will be convenient to readers to refer to their articles. The areas of the camps given in the Inde.'t have been computed by a professional surveyor from the 25-inch Ordnance Maps. They represent the interior areas, excluding ramparts and ditches. References to roads and to Offa's Dyke will be found not in the Index, but on pp. 6, 14, 15. TOPOGRAPHICAL INDEX :— HEREFORDSHIRE. Locality. Abbey Dore (xliv. N.E.) Aconbury (xxxix. S.E.) Almeley (xxiv. N.E.) Asliton" near Eye (vii. S.E. lii. N.E.) A.ston Ingliam (lii. N.E.) Ditto, in Coombe Wood Aymestrey (vi. S.E.) Bache Bacton Bishopstone (xxxii. N.E.) Bishopswood (liv. N.E.) Black wardine Bolitree Brampton Bryan (ii. S.E.) Brandon, (ii. S.E.) Breinton (xxxiii. S.W.) Bridge Sollers (xxxii. N.E.) Brilley (xxiii. S.E.) Brinsop (xxxiii. N.W.) Brockliampton(O.S.xIvi.N.E.) Bucknell (ii. S.W.) Burrington (iii. S.W.) Buiy Hill Caplar Camp Canon Pyon (xxvi N.W.) Clifford (xxxi. N.E.) Colwall (xxxvi. S.W.) Combe (xi. N.W.) Coxall Coxall or Coxwall Knoll (ii. S.W.) Credenliill (xxxiii. N.W.) Croft Ambrey (vii. S.W.) ROM. PRE-R. ROINI. (D PRE-R. ROM. ROM. PRE-R. ? (D POST-R. © (D ROM. (D PRE-R. Nature of Discovery. (D PRE-R. Section of Stoney Street opened 1893 '; Camp, IS acres Two circular mounds (one near Cliurch) Camp Castle tump Bronze sword Hoard of 4tli century coins, said to number many thousands Camp in Pyon wood, 6 acres See Kimbolton Tump at New court Farm Bronze celt Villa, tessellated pavements, coins of Con- stantine : road to Kenchestcr Hoard of about 18,000 "third brass" (a.d. 300—350), in three urns See Stoke Prior See Weston See Walford Camp, 9 acres (near Leintwardine) Barrow and urn, probably the same as that recorded under Walford Small entrenchment (near Church) Offa's dyke Pentwyn camp Small earthwork (near Church) Well, pottery Pawley camp Winged celt (? in Shropshire) Camp See Weston See Fownhope Camp Newton Tump Herefordshire Beacon, 20 acres, part post-Conquest Earthwork Tumuli ? (near Buckton) Camp, half in Shropshire Camp, 50 acres. For the name see p. 6, note Camp, 24 acres h Where recorded. See p. 14 B. B. 636 ; G. C. 74 W. C. T. 1885, 294 (plan) H. C. Moore W. C. T. 1884, 174 A. B. I. 250 A. J. xxxi-v. 365 B. B. 560 0. S. xxxviii. S. W. In possession of the Rev. G H. Davenport, Foxley A. xxiii. 417 A. J. xxxiv. 361 W. C. T. 1882, 257 W. C. T. 1896, 4 N. C. 1896 (not yet published) B. B. 551; G. C. 78 ; A. Xll 92; G. M. 1853, 1. 39 ; W W. 190 G. C. 78 Roy, plate xl. D. H iv. 13 0. S. O. s. D. H iv. 39 W .C T. 1887 ,144 (plan ) W .C T. 1887 .128 G. C. 88; D. H. iii. 240 A. B. 1. 74 B. B. 574 B. B. 579 O. S. B. B. 597 ; G. C. 81 W. C. T. 1880, 220 J. Anthrop. 1881, 330 A. xlvii. 435 0. S. O. S. ii. S. W. and S. E. B. B. 551 ; G. C. 78 W. W. 198 Roy, pi. xl. and pp. 171-5 D. H. iv. 80 A. C. iii. 13 p. 407, 14 p. 93 W. C. T. 1882, 236 (plan) B. B. 560; G. C. 84 10 TopograpJiical Index : — Herefordshire. Locality. Dewchurch, Mucli (xlv. N.E.) Dewclmrcli, Little (xliv. N.W.) Dinedor (xxxix. N.E.) Docklow (xiii. S.W.) Dormingtou (xxxiv. S.E.) Dorstone (xxxi. N.E.) Ditto (xxxi. S.W.) Doward Downton-on-the-Rock (ii. S.E.) Eardi.sley (xxiv. N.E.) Eastnor (xlii. N.W.) Eaton Bisliop (xxxiii. S.W.) Ewias Harold (xliv. S.E.) Eye Fawley Camp Ford (xix. N.E.) Fownhope (xl. S.E.) Foy (xlvi. S.E.) Ganarew (liv. S.W.) Garmsley (xiii. N.E.) Goodricli (li. S.W.) (liv. N.E.) Grendon Bisliop (xx. N.E.) Haffield (xiii. S.W.) Hentland (li. N.W.) Hereford (xxxiii. S.E.) Herefordsliire Beacon Humber (xx. N.W.) © 1) d) d) d) PRE-R. d) PRE-R. ROM. PRE-R. ROM. d) d) d) d) PRE-R. ROM. POST-R. Xatiire of Discovery. ROM. POST-R. d) PRE-R. ? Wormelow tump (now destroyed), per- haps a prehistoric burial-mound ('low') Small oval earthwork (180 feet diam.) Tump in Trilloes wood Oyster Hill camp, 10 acres Ul)hampton (Upperton) camp St. Ethelbert's camj) on Backbury, 5 acres Arthur's stone (cromlech) Two tumps near Fowmine or Vowmynd Farm, one originally moated See Ganarew, Whitchurch Entrenchment Two mounds, one near Church (probably Saxon burh mentioned in Domesday) Red Earl's dyke Midsummer hill camp, 21 acres Wain Street (old road) Camp, Ruckhall or Eaton, 30 acres, per- haps pre-Roman Bronze celts, near Stonej' Street Two lamps, urn, near Stoney Street Burh (perhaps built 916 a.d.) See Ashton See Brockhampton Doubtful Capler, or Woldbury, camp, 10 acres Coin of Lucilla (in Capler camp) Cherry Hill camp, 5 acres Eaton Hill camp Little Doward camp, 19 acres Camp, 8 acres (near Stoke Bliss) The Queen stone Hoard of 4th century coins, scorias Castle Westington (or Netherton) camp Camjj, the Vineyard, 5 acres (near Led- bury) Gaer Cop camp Tump at Treaddow Altars, bronze ligurine (waifs from Ken- chestei') Burh, now removed. Fortifications by Harold Pavement (? Pre-conquest) See Colwall Risbury camps, 17 and 8 acres (not walled with stone) Where recorded D. H. ii. 313 0. S. G. C. 74; B. B. 502 B. B. 587 B. B. 590; D. H. iii. 74 A. C. ii. 5, p. 94; B. B. 545 W. C. T. 1882, 175 0. S. ; H. C. Moore O. S. 0. S. W. C. T. 1880, 220, 1889, 376 (plan) B. B. 585 ; G. C. 74 > In Hereford Museum Clark, i. 21, ii. 39 D. H. iii. 241 ; B. B. 607 W. C. T. 1883, 45 W. C. T. 1883, 45 W. C. T. 1883, 38 D. H. ii. 347 B. B. 608 W. C. T. 1884,212; G. C. 72 W. C. T. 1893, 143 Heath's Wye, 67 (ed. 1799) W. W. 14 A. xxxiv. 433; B. B. 517 G. C. 70 B. B. A. J. xxxiv. 367, G. C. 88 0. S. A. J. xxxiv. 357 W.C.T. 1879, 165; 1882,247 Clark, i. 29, 108, ii. Ill G. M. 1828 (i.), 358 W. C. T. 1885, 334 Topographical Index : — Herefordshire. 11 Locality. Ivington (xxi. S.E.) Kenchester (xssiii. N.W.) Ditto (xxxiii. S.W.) Kenderclnirch Kilpeck (xliv. N.W.) Kimbolton (xiii. S.W.) King's Caplc (xlvi. S.W.) Kingsland (xii. S.W.) King's Pyon Kington (xvii. N.W.) Ditto (xvii. S.W.) Ditto (?) Kinnersley (xxv. N.W.) Ledbury Ditto (xli. N.B.) Leintwardine (ii. N.E.) Ditto (ii. S.E.) Leominster (xii. S.E.) PRE-K. ROM. ROM. d) (D (D (D POST-R. d) PRE-R. (D PRE-R. ROM. PRE-R. ROM. PRE-R. POST-R. Nature of DiseoTeiy. Camps, 17 and 1.5 acre.s, with coins of A.D. 1340-90, two hatchets of uncertain age British gold coins Walled town of 17 acres, streets, pavements, hypocausts, etc. leaden and other drainpipes, pottery, fibulae, coins (many later than A.D. 250) and miscellaneous objects, especially : Illegible altar Milestone of Numerian (see above, p. 4) Oculist's stamp Coins in Hereford Mus. Masonry and tiles, at New Weir, not a bridge " Tumulus " (near Howton) Supposed British village, an error British (?) and Saxon (?) mound Bacli camp, 10 acres Tump (a curious ceremony here on 1st May) Mound (near Church), possibly a burh Tumulus (?), near Butthouse Bradnor Hill camp Moated mound (?) at Hergest, Castle Twts Stone hammer (perhaps neolithic) Earthwork Kilbury camp Wall Hills camp, 3.3 acres, and in it : Spear and arrowheads, pottery, and worked flints Roman bronze coins Hoard of bronze weapons (found at Broadward, i.e. in Shrop.shire, but often called a Hei'efordshire find) Brandon camp, see s.r. Rooftiles, ashes, coins of fourth century, pottery, quern ; earthwork, 14 acres Arrowheads and spears, pottery, burnt bones, tiles, etc. (Dates uncertain.) Gold coin inscribed Ersv Supposed A. S. inscription in church; a forgery (A. S. Napier) I b2 Where recorded. B.B. 575 Price's Leominster, 25 W. C. T. 1882, 213 In possession of C. Hard- wick, Esq.. Old Weir farm L. I. V. 66 H. B. R. 465 Stukeley's Diaries, ii. 289 ( = Iieliq. Galeanoe, p. 120) I. C. 66 (69 ed. ii ) G. C. 66, 74 ; B. B. 583 A. J. xxxiv. 352 ; A. J. xiv. 83 W. AV. 35; D. H. iv. 113 Birch's History of the Royal Society, ii. 274, 347 Ephem. 868 ; Hereford Mus. A. XV. 391 1. B. L. 1165 ; Hereford Mus. B. A. A. iv. 280 1. B. L. 1320 W. C. T. 1896, 6 W. C. T. 189.3, 56 Antiquary, Oct. 1893 0. S. Ivi. S.E. Clark, ii. 165 W. C. T. 1884. 170 (plan) O.S. B. B. 355 ; Clark, i. 108 W. C. T. 1890, 48 O. S. xxvi. N.W. B. B. 548 J. Davies In possession of C. Fortey, Esq., Abbey Villa, Ludlow 0. S. xxxvi. S.W. A. J. xxxiv. 361 ; B. B. 593 G. C. 81 W. C. T. 1883, 24 W. C. T. 1883, 26 A. B. I. passim B. A. A. xxiv. 64 A. C. iv. 3, p. 338 A. J. xxxiv. 350 A. Civ. 5, p. 163 W. C. T. 1882, 252 (plan) AVilliams's Leominster, 44 A. B. C. 494 G. M. 1827, i. 414, 503 12 Topographical Index : — Herefordshire. Locality. LeWon (near Leintwardine) (vi. N.W.) Lingen Llaroillo (xlix. N.W.) Llangarren (li. N.W.) Longtown (xliii. S.E.) Lyonshall Marcle, Much (xli. S.W.) Newton Clodock (xliii. S.E.) Oyster Hill Camp Oldbury Camp Peterstow (li. N.W.) Putley (xli. N.W.) Risbnry Camp Rose Hill Ross, near Rowe Ditch (Pembridge) St. Ethelbert's Camp St. Margaret's (xxxviii. S.W.) St. Weonards (1. N.E.) Sellack (xlvi. S.W.) Staunton-on-Arrow (xi. N.W.) Stoke Blis.s Stoke Prior (xix. N.E.) Stretton Grandison (xxxv. N.W.) Sutton St. Michael (xxvi. S.E. and xxvii. S.W.) Sutton St. Nicholas (xxiv. N.W.) Thornbury (xiv. S.W.) ROM. ROM. PRE-R. ? POST-R. PRE-R. PRE-R. ROM. ROM. PRE-R. ROM. (D PRE-R. PRE-R. ROM. (D ROM. Nature of Discovery. ROM. POST-R.? (D Gold coin of Tiberius Tumulus, near the Farland Coin (undecipherable) Stone cist and skulls Rectangular earthwork, 3 acres (part certainly post Conquest; see p. 16) Earthwork and mound at Ponthendre Offa's dyke Oldbitry camp, 14 acres Winged celt, in camp Brouze celt See Dinedor See Marcle Smelting works, coins of Philip, etc. Villa, h^'pocaust tiles, etc. (near Church) See Humber See Weston Winged celt, pi'obably from Weston Coin of Trajan do. Dyke about 1 mile long See Dormington Cruciform earthwork (perhaps late) Ferrule Earthwork (South of Whitehouse mansion) Mound, cremation burials Coins, scorias Mound or camp Wapley camp, 21 acres (near Combe and Eyewood Warren) See Garmsley At Blackwardine : coins of early and late empire, skeletons, pottery, armlet of Kimmeridge clay, kilns (?). A Cold Harbour is near Earthwoi'k, on hill On lower ground : scales, coins, bracelet, fibular, lamp (with figure of Actaeon) Sutton Walls, 30 acres, said to beMercian Palace (see p. 7) The Devil's stone, Wergin 's stone, ( ? Crom- lech or medieval) Wall Hills, 26 acres Where recorded. G. M. 1789, 742 0. S. vi. S.W. Hereford Mus. A. C. iv. 9, 76 D. H. ii. 275 0. S. O. S. xvii. N. E. D. H. iii. 43; W. C. T. 1883,38 A. B. 1. 90 In possession of Mrs. Jenkins, of Holmer D. H. i. 29 W. W. 15 A. J. xxxiv. 364 D. H. iii. 97 A. J. xxxiv. 363 A. B. I. 91 B. B. 508 O. S. xi. S. W., xviii. N. W. A. J. X. 358, xi. 55; A. C. ii. 5, p. 94 ; G. M. 1853, 387 A. B. I. 340; A. J. xi. 55 J. Davies A. C. iii. 1, p. 168 A. C. iii. 1, p. 168 ; W. W. 16 G. C. 88 W. C. J. 1873, 59 ; G. C. 84 A. C. iv. 4, p. 338 A. J. xxxiv. 372 W. C. T. 1885, 341 B. B. 589 ; G. C. 88 A. J. xxxiv. 368 W. C. T. 1882, 255 A. C. V. 5, p. 187 L. I. V. 55 D. H. ii. 173 B. B. 586 G. C. 86 W. C. T. 1887, 120 Topographical Index : — Herefordshire. 13 Locality. Tretire (1. N.B.) Trewyn (xlviii. S.E.) Turnaston (xxxviii. N.W.) Uphampton Camp Vowchurcli (xxxviii. N.W.) Walford, near Brampton Bryan (ii. S.E. S.W.) Walford, near Ross on Wye (li. N.E.) Wall Hills Wapley Camp Walteistone (xlix. N.W.) Wellington (xxvi. N.E.) Weoblcy (xxv. N.E.) Weston under Penyard (Hi. S.W.) Also called Bolitree, Bury Hill, Rose Hill [See also Ross] Whitcliurcli (liv. N.W.) Wigmore (vi. N.E.) Woldbury Camp Woolhope (xl. N.E.) ROM. d) PRE-R. ROM. ROM. ROM. ROM. PRE-R. ROM. ROM. PRE-R. ROM. F Nature of Discovery. POST-R. POST-R. Where recordeil. Scoriae ; a coin Inscribed altar (now used as font in church) : original provenance unknown Camp Near Cothill See Docklow Camp. Also mound near Chanstone mill Tumulus, urn with bones {see Brandon) Coins Camp, 20 acres (for the ford, see p. 14) See Ledbui-y, Thornbury See Staunton Camp At Coed-y-Gravel, villa with tessellated pavements Oven (?), pottery Fourth-century coin Coins of Cunobeline, etc. Celt Foundations, pottery, bronze, and other objects ; scoriae, etc. over a large area ; coins Coins (earliest gens Gordia, latest Magnentius) Bronze statuette of Diana Tessellated pavements, coins, scorice G reat Doward Camp King Arthur's cave (flint chips and Romano-British ('r') pottery in it) Darvold (Deerfold) hill, doubtful Castle or burh (a.d. 921) See Fownhope ? Chapel of Dubritius (now destroyed) W. W. 16 I. B. L. 163 A. J. xx.xiv. 365 W. C. T. 1882, 248 W. W. 17. W. C. T. 1884, 153 A. J. xxxiv. 3 D. H. ii. 274 G. C. 70; B. B. 544. W. C. T. 1882, 254 Roy, pi. xl. A. C. iv. 5, p. 163 A. J. xxxiv. 366 0. S. W. C. T. 1882, 258 A. vi. 13 W. C. T. 1882, 258 Part in possession of J. Arkwright, Esq., Hamp- ton Court Dingley ccxlv. A. B. C. 456, 568 B. B. 514 B. A. A. xxvii. 203 A. B. I. 78 A. J. xxxiv. 358 W. W. 25 U. H. iii. 214 B. B. 514 A. C. ii. 5, p. 98 Fosbroke, 36 B. A. A. xxvii. 203 [hence N. C. xi. (1871) 155] A. ix. 368 W. W. 14 W. C. T. 1882, 258 A. J. xxxiv. 363 A. C. iv. 3, p. 74 G. C. 71 W. W. 13 W. C. T. 1884, 217 A. S. I. 475 Geological Magazine, viii. 433 British Assoc. Reports, 1871 B. B. 659 A. S. Chron. ; Clark, ii. 531 D. H. iii. 239 14 An Arcliacological Survey of Herefordshire. APPENDIX I. Roman Roads in Herefokdshiee. 1. Watling and Stoney Street. — This road forms part of the itinerary road between Viroconium (Wroxeter) and Isca (Caerleon). Its course; and that of two small branches from it, is well known. It comes down the Clun valley from Shropshire to Leintwardine {(). S. ii. N.E. S.E.), passes Pay toe, Green Lane, Leinthall Moor (where it is lost) Pyon, and Aymestrey (vi. N.E. S.E.). South of Mortimer's Cross it appears to keep a strai<;rhi line, forming, as else- where, the parish boundary, while the modern road diverges to the east fxi. N.E. S.E.. xii. S.W., xix. N.W.), but the two roads meet at Stretford " and pass Oldfields. At Canon Pyon they diverge ; the modern road turns east, while the Roman road follows a line soutiiwards nast Tillington Court, and slopes round Crcdenhill to Kenchester (xxxiii. N.W.) and the Romano- British town there. Here two branches diverge. One road leads west to the villa at Bishopstone (xxxiii. N.W. , xxxii. N.E.), the other east to Stretton Grandison (xxxiii., xxxiv., xxxv. N W.), where remains have also been found. (The question whether these roads continue further than Bishopstone and Stretton will be considered in o [h) (c) (rf) ). From Kenchester the Itinerary road proceeds S.W., crossing the Wye probably at Old Weir by a ford or causeway rather than a bridge. From the Wye it can be traced under the name of Stoney Street over Worm Hill, Great Brampton, and Kerry's Gate (xxxiii. S.W., xxxix. X.W. N.E.) to Abbey Dore (xxxviii. S.E.), where a section was made in 1893. It is described as being 13 feet wide, pitched with largo pieces of local limestone larger than a man's head, and showing two distinct wheel tracks four feet apart (letter from ('apt. Freke Lewis). Its furtiier course is obscure ; probably it ran past Ewyas Harold, Llancillo, and the villa at Waltei'stone into Mon- mouthshire (xliv.) to Abergavenny {Gobannium). I have had no means of ascertaining whether the name Watling Street in Herefordshire is a genuine old name or one bestowed by antiquaries. Besides the two chief Watling Streets, the one which connects London with Wroxeter and the one in Northumberland, there are in Enoland several lesser ones, in Herefordshire, Lancashire (Watkin, Roman Lancashire, p. 70), Cheshire (Watkin, Roman Cheshire, p. 42), etc. and the antiquity of the names, as applied to all these lesser examples, requires investigation. The name has perhaps a fair claim to be considered " Stukeley {Itiu. Our. p. 7.3, cd. ii.) mentions this road as passing through Stretford : he adds, that it then went " through Biriton, two miles north of Leemster, where they dig up the pavement of it, made of squarish i-ag-stone." I do not know where Biriton is ; if it is Buiy, two miles north of Leominster, it lies far out of the track of the real road. t' An Archaeological Survey of Herefordshire. 15 genuine in Herefordsiiire, for Horslcy {Brit. p. 388) seems to say that it was known to the " country peojtle " in his chiy, i.e., abont 1720 — 80, Imt his statement is rather vague. 2. If tlie identitication of Ariconium witli AVeston near Ross be accepted, tlie Itinerary road connecting Glevitm (Gloucester) and Inca (Caerleon) must have passed through South Hereford- shire, as indeed general considerations suggest. It has been thought tliat it ran Ijy Dursley Cross and Huntley (li. S.E.) near the modern road from Gloucester to Ross, but practically its course is unknown. West of Arlcunium it may have crossed the Wye at the ancient ford of Walford, in Domesday Walecford, i.i-. the Welsh ford. 3. It may be convenient to add here some details about the roads which have been traced in Herefordshire on what seems insufficient evidence: [a) A road has been thought to branch from Watling Street at Leintwardine and run parallel to it at a little distance east through Blackwardine to Weston. The i)art of this road which most resembles a Roman road is that between Ashton, Stockton Ridge, Patty's Cross, Stretford, and Blackwardiue (xix. N E. S.E., xx. S.W.). The rest is pure imagination. The course supposed is by Satiron's Cross, Preston (xxvii. N.W. S.W.), Withington, Lugwardine (x.\xiv. X.W. S.W.). to Fownhope (xl.), Brockhampton, How Caple (xlvi.), and Weston (lii.), but no trace really exists The southern jiart, Fownhope to Weston, has also been called a road from Weston to Kenchcster, but this is as imaginary as the rest. {!)) A road has been thought to continue the Kenchester-Bishopstone road into Radnorshire, according to some, to the Roman ren)ains at Cwm : as to which remains see Archaeologia, i. SOi, xvii. 170; Hoare's GiroMns Camhirii.^i'i, i. p. clvi. ; Williams' Radnorshire, 48, 241. There is no evidence for such a road. (c) A road has been imagined to lead from Stretton Grandison eastwards over the Malvern hills to Worcester and Kempsey (where a milestone has been found). This road also is wholly unproven. The Ivem})sey milestone (I. B. L. 1157) perhaps belongs to a Roman road along the Severn valley. {d) There is somewhat more evidence for a road south-east from Stretton Grandison towards Gloucester. A singularly straight piece of road, nearly 7 miles long, runs from Stretton. through Ashperton (xxxv. N.W. S.W.), Little Marcle (xli. N.E. N.W.), to Preston (xli. S.E.), and thence more deviously to Dymock and Kewent. In the large maj) of Isaac Taylor (Ross. 1754). the earliest on which roads are marked, this straight piece does not appear, but, if really an old road, it may perhaps be Roman. 16 An Archaeological Survey of Herefordshire. APPENDIX II. BoEDEE Mounds. The position of Herefordshii'e as a border county in both Saxon and Norman times led to the erection of many fortresses and strongiiolds, one class of wliich may be conveniently dealt with in an appendix. This is a class of small earthen mounds, usually moated, circular, and measurino; perhaps 100 feet to 200 feet round the base by 30 feet to 35 feet in height, which may often be found in proximity to churches and court or manor farms, and which occur in especial abundance in Herefordshire and on the Welsh border generally. These mounds are a local modification of the ordinary Saxon burh. Defended by a wooden stockade, they provided local strongholds against Keltic inroads, and correspond in function exactly to the pele-towers of Korthumljerland. The custom of erecting them appears to have lasted into the thirteenth century, and as this Survey includes only pre-Conquest antiquities, I have thought it best to include in the general index and map only such mounds as there are special reasons to consider earlier than 1066, and to add here a list of all such mounds, whether datable or not, indicating by an asterisk those which also find a place in the general Index. I have included in this list only such mounds as by their proximity to churches, etc. declare their use as places of refuge and local strongholds. Isolated mounds or tumps may often belong to this class, but are best excluded in default of definite evidence. For some discussion of the whole question I may refer to Mr. G. T Clark's Medieval Military Architecture, vol. i, ch. i. Mr. Clark is probably not wrong in asserting that most of the Norman castles in Herefordshire were built on earthworks of earlier date, and therefore our list of border mounds represents only a small part of those which existed in Saxon times. Some of the mounds may, indeed, be older than the Saxons, like that at St. Weonards, which is a pre-historie burial mound, but here again our evidence is too scanty for definite conclusions. I see no kind of reason, however, for assuming such pre-English earthworks at Longtown and Kilpeck, as is sometimes done. It will be undei'stood that my list is merely a tentative beginning for others to complete * Almeley (xxiv. N.E.) * Kilpeck (probably Castle) Aston (iii. S.W.) See W. C. T. 1879, 174 * Kingsland * Bacton Lancillo (xlix. N.W.) Birley Leysters (xiii. N.W.) *Bmntoii? Mansell Lacy (xxv. S.E.) * Brinsop ? Orcop (xlv. S.W.) Burglnill (xxxiii. N.E.). See B. B. 579. Rowlstone (xliv. S.W.) Probably = " Burh-hill" St. Weonards (1. N.E.). ,Seep. 12. Dilwyn ? (xviii. N.E.) Shobdon Court (xi. N.E.) Dorstone (probably Castle) Staunton-on-Arrow (xi. S.W.) Eardisland (xi. S.E.) Tliruxton (xxxix. S.W.). Opened in 1867 * Eardisley without results Edwin Ralph (xxi. N.W.) Wactou (xx. N.E.) * Ewias Harold Walterstone (xlix. N.W.). See D. H. ii. .S12 * Hereford * W la'more 52'—^ Prepared bv permission, fi-oin the Ordnance Map. ARCHAEOLOGICAL MAP OF HEREFORDSHIRE PRE-ROMAN ftOMAN Prepartd by permtiiion./roiH tht Ordnaait M.jp. .9o»/^b >>ui'i-'*. UNIVERSITY OF CALIFORNIA LIBRARY Los Angeles This book is DUE on the last date stamped below. THE LIBRART UNIVERSITY OF CALIFORNIA KX)6 AXCXUSS oevan - ■ I -"■ ^"''0 An archaeological H^B46 s urvey of Herford. shire *DA 670 H4B46 if