A Roman Land Settlement near Rochester

A ROMAN LAND SETTLEMENT NEAR ROCHESTER By MICHAEL D. NIGHTINGALE, B.Sc.(Agric), B.Litt. WITH AN INTRODUCTION BY COURTENAY EDWARD STEVENS, M.A., B.Litt., F.S.A. THE attempt of Mr. Nightingale to reconstruct from the evidence of modern field-divisions a Roman centurial grid in the grain-growing areas north of Rochester, the ancient Romano-British Durobrivse, may not carry conviction to ah (there are purists who demand either literary evidence or the discovery of a typical boundary stone); nevertheless, I think that he can fahly claim that his discoveries conform more clearly to Roman traditional methods of land surveying and division than other attempts elsewhere in the country. It is necessary, therefore, to justify this claim by explaining in brief what these traditional'methods are and what can be said of theh possible application to Britain in general and to the Rochester neighbourhood of Britain in particular. Centuriation is a method of laying out regular rectangular areas for cereal cultivation (it was not used for the delimitation of pasture or woodland)—a method employed by a guild of technicians (agrimensores) according to strict professional doctrines, large portions of which are preserved to us in the Corpus Agrimensorum.1 The instrument of survey was the groma,~a, kind of cross-staff, with which the surveyor began by establishing two lines (limites) intersecting at right angles ; these lines, the framework of the grid to be constructed, are the decumanus maximus, which should run from east to west, and the kardo maximus, from north to south. This was the strict doctrine, based on old religious concepts, but it could be modified, notably in the dhection that the decumanus maximus might occupy the largest extension of the land to be surveyed, or that one or other—or both—of these lines might be laid down with reference to the line of an important road. The point of intersection was normally, but not invariably, taken in the centre of the town to which the land belonged. With the aid of stations for the groma along the two lines, the surveyor constructs a rectangular grid of secondary limites to be, like the original lines, fixed on the ground by permanent metalling. In the early days, these rectangular areas might be of a variety of different sizes and shapes, but from the time of the Gracchi (133-121 B.C.) they 1 There is an extensive literature on the theory and practice of Roman surveyors, mainly in German and Italian which I have utilized for my general remarks. A valuable summary in English mentioning air-surveys in southern Europe is given by Bradford, Antiquity, XXI (1947), pp. 197-204. 150 A ROMAN LAND SETTLEMENT NEAR ROCHESTER tended to become standardized. Indeed, the doctrines of the professional writers, reinforced by observations on the ground and from the ah, teach us that during the imperial period of Roman history there is only one type of centurial lay-out that we can seriously expect to meet. This is the grid of units, each a square of 20 by 20 actus, the actus comprising 120 Roman feet. The unit thus formed wih consist of a square of 2,400 by 2,400 Roman feet (776 by 776 yards, 710 by 710 metres). Since the standard measure of Roman land, the iugerum, is a double square, 2 by 1 actus, we see that our unit wih comprise 200 iugera, in fact 100 of those plots of 2 iugera each which made up the traditional land-holding of the primitive Roman. Hence the unit of this size, which forms, as we have said, the rule, is appropriately styled the centuria, whence the term " centuriation ". Each of these centuries, has an individuality which should express itself by stones at each corner measuring the " latitude " and " longitude " (if one may so express it) from the original station, and there may well be land divisions inside the centuria, known to the professionals as limites intercisivi, which may have boundary stones, but which wih lack the schematism of the formal survey. We have next to enquire to what types of land in the Roman empire these methods of survey and lay-out were applied. According to the professional writers, they were intended exclusively for Roman colonial allotments, and thus we might expect to find traces of such centuria} surviving in the neighbourhood of our British colonies, but none of the attempts have been very successful.1 We are told that land in the provinces was normahy simply measured along the frontiers of the city territories with its internal division left as the responsibility of the local magistrates, which is why, no doubt, traces of " centuriation " are so difficult to find in this country and in Gaul; while imperial estates could be divided up into a system of rectangular plots reminiscent of but fundamentally differing from " centuriation ", a method which Mr. Margary has, perhaps, discovered in south Sussex.2 But the professionals make the important observation that in all these types of land where " centuriation ", according to the strict rules of the profession, should not occur, it occasionally does (a fact which somewhat annoyed the professional conscience of one of them !), and research on the ground in Africa has shown that imperial estates, private estates, as 1 Richmond is—rightly—critical of attempts at Colchester and Lincoln (Arch. Journ., CIII, pp. 61, 66). Sharpe's attempt to find centuriation in Middlesex is quite fantastic (Antiquities of Middlesex). 2 Roman Ways in the Weald, pp. 204-7. I do not feel convinced that Margary really found a grid, but if he did, it has nothing to do with a " colony " (for where is it ?), but an " assignatio per scamna et strigas " which the professionals associated with publio land in the provinces. It would be possible to make a case for public lands in this district. 151 A ROMAN LAND SETTLEMENT NEAR ROCHESTER well as land in the territories of cities which were not Roman colonies were treated to this operation of gridding in centurice of 200 iugera, just as if they were colonial land. No doubt the snobbish desire of pretending that one's land was really as Roman as a Roman colony played its part.1 In short, though " centuriation " is not likely to have been common in Britain outside the colonies and its appearance outside them would be professionally irregular, there is no reason why we should not find it here and there. In the light of this, let us examine Mr. Nightingale's discoveries. We cannot be sure where the original station for the groma was, but it does not look as though we are to look in the town of Rochester ; in fact Mr. Nightingale makes a good cause for supposing it to be at the intersection of the straight lines of the Watling Street and the Maidstone Roman roads, which, if produced, would meet at Frindsbury. The grid itself is orientated to the cardinal points in true professional style ; indeed while.it is orientated on the roads as well, the course of them seems to have been readjusted to fit the grid, rather than the grid laid down to fit them, as occurred, for example, with the grid outside Anxur-Terracina, laid down to fit the Appian Way.2 We would seem then to have the interesting result that the grid is probably contemporary with the present course of Watling Street, which would presumably assign it to an early phase of the Roman occupation. The unit is the centuria of 200 iugera which is as it should be, and Mr. Nightingale seems to have traces of the limites intercisivi. Finally his discovery invites us to consider the status of Durobrivse itself. The evidence suggests that for a long time it was merely a village in the territory of the Cantii with their capital at Canterbury (Durovernum).3 On the other hand there are its massive (and seemingly early) Roman walls and the fact that aheady in A.D. 604 Augustine chose it as the seat of a bishopric* Bede indeed cahs it by its Roman name, which persisted for some centuries in documents.5 The rule in the west that episcopal sees must correspond with politically independent units of local government was very precise. One would expect, therefore, that Augustine's choice was dictated by the fact that Durobrivse had been in the past an independent unit, that the area of the Cantii had 1 Hadrian once expressed surprise at the eagerness of ordinary towns .to pretend that they were real Roman colonies (Qellius, XVI, p. 13). 2 See the illustration in the Corpus Agrimensorum reproduced in Grenier's Manuel d'Archeologie, II, p. 16. 3 Durovernum alone has the tribal name of the Cantiaoi attached in the Ravenna list, where Rochester appears as an ordinary place name, and Rochester is not mentioned in Ptolemy's list of towns of the Cantii at all. (See Haverfleld and MacDonald, Roman Occupation of Britain, p. 190.) 1 See V.O.H. Kent, III, pp. 81-3, on the Roman name and town walls. 5 Bede, Hist. Heel., II, p. 3. Of. English Hist. Review, LII, p. 198. 152 A ROMAN LAND SETTLEMENT NEAR ROCHESTER been divided at some time and given a new capital; and there are parallels for this elsewhere. One would like to think that it was at this moment that the men of Rochester in a kind of eashy imaginable parvenu sphit determined to equip themselves, in spite of professional head-shakings, with a truly Roman land system. But the relation of the grid to Watling Street, pointing to an early date, gives pause, and leads us to hope that Mr. Nightingale may be able to follow up his investigations from map and field study with excavations of the ditches, which, according to the professionals, accompanied the metalled limites. A British expedition is looking for datable material in the ditches of a centurial system in Apulia. There is work to do here at home, and it would be worth doing. * * * * CLIEPE Church is situated about five miles north of Watling Street and Rochester Bridge. An arable plateau stands out above the marshes, and this land consists of a thick fertile loam (medium to light) overlying a raised table of chalk in the north and the Thanet Beds in the south.1 To-day, these fields carry a large acreage of market garden crops, for which the soil is most suitable. Edward Hasted, writing in 1778, refers to the state of the village in his time :2 "The land in this parish is in general rich; great part of it lies in a common unenclosed field of more than 2000 acres. Close under the cliff below Courtsole at the bottom of Church Street are the marshes, containing many 100 acres which reach from there northwards to the Thames, which encircles them in the form of a crescent." If reference is made to a small scale map of the area it will be seen that the road plan of Cliffe appears peculiar in comparison with that of the surrounding country. There are several roads running in a northsouth dhection parallel to one another, and others intersecting them at right angles. It would be more natural perhaps to find roads radiating from Rochester, and in a more haphazard fashion, (v. figs 1. and 2.) After measuring distances between these roads, it became evident that they were separated by multiples of Roman actus* and that the 1 For Geology see Report of the Land Utilization Survey of Britain, Part 85 (Kent), p. 562. 2 Edward Hasted, History of Kent, 1778, Vol. I, p. 536. 3 Actus = 116-496 English feet (120 Roman feet of 11-6496 English inches) according to Sir Wm. Smith's tables, whose figures I have used. Both I. D. Margary in his " Roman Centuriation at Ripe " (Sussex Arch. Collections, LXXXI) and John Bradford in " A Technique for the Study of Centuriation " (Antiquity, XXI) have taken the actus = 116-05 English feet. For other centuriation terminology see the original texts and diagrams of the classical writers Frontinus and Hyginus (ed. Carl Thulin and published by Teubner in Corpus Agrimensorum Romanorum, Leipzig, 1913). A description of them is given in French by M. A. Schulten (Bull. Arch., 1902, p. 129). See also Haverfleld (Eng. Hist. Review, XXXIII, p. 289), Henry C. Coote (Archceologia, XLII, p. 127) and C. E. Stevens in the Cambridge Economic History, Vol. I. 153 A ROMAN LAND SETTLEMENT NEAR ROCHESTER T H A M JJECyllANUi MAWMU^ Cross © © © © Roman road found while excov/ah'nq Strood docK C.I8W Roman hoost, PolTeru and ptouqhshon:, found tm-1 Komon poHiru kiln found or Broomy Form fcj>rnano-Btihjh crcmo+ion. txrshrva roads Roman alianmenrs M E P W A V COCr S,TEfe LAND SETTLEMENT NEAR ROCHESTER.. Acolt in milts Fia. 1. 154 A ROMAN LAND SETTLEMENT NEAR ROCHESTER area had been planned rather than developed from natural causes. This plan takes the form of a chess board grid of which the orientation of the cardo is almost exactly north-south (2 degrees east of north) and correspondingly the decumanus east-west (2 degrees south of east). Taking the road which runs north from Cooling Street past Gattons and Berry Court as the Cardo Maximus or principal axis, it wih be found that at a distance of 40 actus west another cardo coincides with West Street. Within these two limites the road past Perryhill, James' House and Woodview House is at a distance of 10 actus west from the Cardo Maximus, whilst the Station Road is at a distance of 30 actus further west. Now take the road which hes east and west a little to the north of Mortimers Wood as the Decumanus Primus} then Reed Street near Cliffe Church coincides with a decumanus at a distance of 80 actus. Within these latter limites Salt Lane is at a distance of 50 actus, the road lying a httle to the north of it 53 actus and the road leading to Cooling Castle 55 actus north from the Decumanus Primus. If the Cardo Maximus is extended in a southerly dhection it passes through Frindsbury to the north bank of the Medway at a place which is opposite Gashouse Point on the southern bank. Extend this hne across the river over Gashouse Point, and it wih pass through the junction of Eastgate, Star Hhl and High Street, Rochester and then coincide with the Star Hhl—Delce Road ahgnment of the Roman Rochester- Maidstone-Hastings Way.2 At the present stage of investigation it is impossible to say whether there was ever a through road and bridge across the Medway on this ahgnment or whether it was only the result of planning on the part of the agrimensores. It may be significant that a warped extension of the road which coincides with the Cardo Maximus in Cliffe is called " Port Way " where it passes through Berry Court Wood (v. plan : fig. 2). The present accepted course of Watling Street from Gravesend to Rochester turns to the south at Salter's Cross in order to make for Rochester Bridge. If, however, this turn is ignored and the original ahgnment extended it whl pass over Strood Dock (close to Roman Foundations marked on the O.S. 6 in. map as found in 1828)3 and finally wih meet the Cardo Maximus Cliffe-Delce Road alignment at a position near the north bank of the Medway opposite Gashouse Point. Within 400 yds. of this junction foundations of a Roman house were 1 If the Decumanus Maximus was Watling Street, as I shall suggest later, then this limes which I take as the Decumanus Primus was 120 actus north from it. 2 For alignments and large scale plan of this road refer to " Roman Roads in West Kent ", I. D. Margary (Arch. Cant., LIX, p. 30). 3 This was a piece of paved way running east and west across the site of Strood Dock, actually found during excavations c. 1819. It has been thought if Roman to have connected the house at Frindsbury with the London Road. Such a road would have followed the course that I have outlined. Refer to V.C.H., Kent, I I I , p. 116. 155 T^^&zgssrci Romano- £>n|\sh ^ . . , ^ 0^ cremation %^"' ~ * ° / \ ^ ^ f e ^ = — . *° / Quarry B u l l * ' A enw Hoiua L [fiwi Sircei 1 ,*M - ^^?\& v • l V - ^ - - > y 1 »Berrv ,:Court... Wood'"-.: J ^ > ^ 4 o M >*1' FIELDS AT CLIFFE WITH GRID MAR.KEP IN ACTUS SCALE OF FURLON6S I ', J, L i I ' I i i I ,'a • 3 SCALE OF ACTUS strip tomWaries added from 1840 Hthe. map ; # f ^4 Reproduced from the Ordnance Survey map with the Sanction ofllie Controller ofB.M. Stationery Office -Mortimers -::Wood