THE MIDDLESEX DISTRICT IN ROMAN TIMES (Part II).

The Romanization of the District.—The Pagus.—The Saltus.—Allotments and Farm Holdings.—Places of Pagan Worship and Middlesex Churches.—Compitalia.—Paganalia.—Silvanus.—Policy of Christian Missionaries.—Village Settlements.—Vectigal Lands.—Tenure of Land.—Descriptive Perambulation through Middlesex.—Tables.

We will now consider the Roman system under which the country side was surveyed, divided into districts, and planted with immigrants settled in village communities. For this research there are, besides the spade of the antiquary, other sources from which evidence can be obtained as to the extent of the Romano-British settlements within the quadrilateral known as Middlesex.

Such evidence would principally include:—

(a) Vestiges of Roman ways laid out in parallel courses, having, in different divisions of the County, distinct and separate alignments.

(b) Indications that the sites of many Romano-British chapels and sacred groves are now marked by the mother Churches of ancient parishes.

After a brief description of the Romanisation of the district, the different links in the evidence of the Roman occupation within the County will be examined seriatim, and the conclusions finally summarised in a descriptive perambulation through Middlesex. {4}

The Romanisation of the District.—When Agricola commenced the peaceful settlement of the Imperial Province of Britain, primeval woods, wastes, and swamps covered the greater part of the Middlesex district. Its northern borders were enclosed by an extensive oak forest, and its three other sides were fringed by marshes which marked the courses of the Thames, Colne and Lea, so that probably not more than 1-6th of this quadrilateral was under tillage by the Catuvellauni, whose fields and clearings lay scattered chiefly over the Thames Valley, and on the plateau to the west of the lower Brent. (4.1)

Now it was a dictum of the Romans that “where they conquered there they inhabited”; their practice being to plant settlers where the ‘axe and plough’ of the natives had already opened up portions of the land. Hence by many of the tribal clearings Romano-British settlements arose; in the next age to become Saxon vills and tonships, the extent and names of which, as given in the Domesday Survey of Middlesex, have in most instances survived to the present day, either in ancient parochial areas or in manors. But their former or Romano-British designations are now unknown, with some few exceptions, such as:—‘Londinium’ which in the 3rd century was for a time called Augusta—‘Sulloniacæ’ (Stanmore)—‘Pontes’ (Staines) from the bridges across the Thames and Colne to carry ‘Tamesis Street’ the great main road to the S.W. of Britain—and perhaps also ‘Ruislip,’ probably so called from the entrenched enclosure there by the marshes of the river Pinn.

Twenty years prior to the introduction by Agricola (A.D. 78–85) “of the language and civilization of Rome” a military colony (Colonia Victricensis) at Colchester, and a Municipium at Verulamium had been established within the territory of the Catuvellauni. The remainder of their former tribal lands, here designated the Londinium Civitas, being probably administered as a Prefectura (4.2) with its official headquarters at Londinium, which by A.D. 61 had become known as a port and centre of trade.

The area under this administration was considerable. On the east it seems to have embraced South Essex with the small towns of {5} Cæsaromagus (Chelmsford), Durolitum (Romford), and the castellum at Othona, St. Peter’s Head, Bradwell (5.1). On the north it extended to the upper waters of the Lea and Colne, which rivers also marked the southern limits of the Verulamium Municipium, on the west to the Chiltern Hills, and on the south to the Thames. (5.2)

Under Roman law the ownership of the soil in a conquered province vested either in the Emperor (manu Cæsaris) or in the Senate at Rome; the occupiers only enjoying the possession or usufruct from it. (5.3) As the province of Britain was in the former category, the collection of taxes and dues (vectigalia) was not under the control of the Quæstor, as in a Senatorial province, but under that of a Procurator, an Imperial officer who also defrayed certain responsibilities undertaken by the Emperor, together with the costs of military defence, and the transport of troops; (5.4) while the supreme command was in the Proprætor who was principally engaged with military and judicial matters.

From the immigrants brought to Britain through the port at Londinium, many would be passed into the fertile lands of the Londinium Civitas, and so the district, now known as Middlesex, would be one of the first to be settled and yield returns into the Imperial Exchequer ( fiscus) at Londinium, the centre of the financial administration. “Day after day … and to Britain, not only officials and troops go, but countless private people.” (5.5)

The Pagus.—The boundaries of provincial districts were mostly natural ones, such as the ridges of hills, and the courses of rivers, and were therefore irregular (5.6). So also were those of its cantons or pagi. But within each pagus the portion surveyed for settlement and cultivation was rectangular in shape, and was divided into a number of square areas (saltus) by cross parallel lines laid out by the State Surveyors. As regards the making of these divisions it will suffice to {6} say, that a straight line (decumanus maximus) was drawn eastwards and westwards across the pagus, and at an appropriate spot a similar line (cardo maximus) crossed the former at right angles. On either side of these two lines a succession of parallels were drawn at distances of 612 Roman poles equal to 9 furlongs, the last in each of the four directions constituting the boundary line ( finitima linea) of the land marked out for settlements. These parallels were known as fifth lines (limites quintarii) because where requisite, four intervening and equidistant parallels could be added. In this way a large rectangular tract in each pagus was divided into squares like a gigantic chequer board: the quintarial lines creating the saltus (810 acres including roads) and the other lines, the 25 smaller land areas known as laterculi or centuriæ (31·155 acres).

For the public traffic, district ways (viæ) were designed to be laid out upon the quintarial lines, (6.1) and the lanes or bye-ways (deviæ) upon the intervening ones, but in practice as the land became cleared and cultivated they followed the parallel divisions of the soil.

The finitima linea in a pagus, in default of appropriate natural objects, was marked by mounds of earth (botontini), usually containing charcoal and other matter foreign to the spot, and the internal divisions by roads, stones, trenches, blazed trees, etc. (6.2) The waste outside the pagi and on the confines of the Civitas, such as the Chiltern hills, was known as ager arcifinius or occupatorius; while that between the natural boundaries of the pagi and the finitima linea of the rectangular areas marked out for settlements, as ager extra clausus.

While the lay of the land, and the fertility of its soil would generally determine the natural size of a pagus, the rectangular or intra clausus portion of the pagi in the Middlesex district, seem to have been planned so that two of them, or an administrative area, contained forty saltus, equal to 32,400 acres. On the map (6.3) the finitima linea of the intra clausus within each pagus is shown by the outer border line (shaded) of every saltus, beyond which there was insufficient land within the natural boundaries to form any further complete saltus. It is submitted, that the positions assigned by the writer to these border lines as well as those of their fellow parallels are {7} correct, for not only do they follow the alignment of ancient ways still existing, but they also pass by the remaining botontini and the sites of village compita (cross road chapels) which, as will be explained, were erected by the ways first to be cleared. With regard to the irregular borderland in extra clausus a few portions of the original woods and waste (silvæ et pascua) appear to have been leased and assigned as the number of settlers increased, probably during the palmy and later days of the Roman occupation. Middlesex district, diagram 1

The ways and bounds shown upon this diagram were chiefly taken from the fine 2 inch map drawn by Rocque in 1775, after his survey of Middlesex, {8} perhaps the first to be scientifically undertaken since the Roman era. This rare map shows the villages unenclosed, just as they had remained for centuries, besides including many rural features which have since been obliterated by the hands of the builder.

On the diagram it will be noticed that the local roadways in the divisions marked 1, 2 and 5 run due N. and S.: in the large central district No. 3, from N.N.W. to S.S.E.: in No. 4 from N. by E. to S. by W.: and that in all three instances they are crossed at right angles by other ways.

The discovery, that the ancient roads of the Middlesex district had so been laid out was made in this way. The writer having noticed that sections of ancient highways and parish boundaries (8.1) ran in parallel lines, which in each of the several divisions of the County possessed a different alignment or orientation, he had the diagram prepared, from which, by the omission of other details, these remarkable features could be readily seen.

Now it is obvious that from various causes during a period of 17 centuries, many deviations from the original alignments have arisen, and numerous lengths of the original ways have dropped out of use, but fortunately a sufficient number remain to indicate the peculiar features, and the comprehensive character of the survey. It is however clear, that neither the rude Saxons nor their Norman successors were capable of designing or carrying out such a big undertaking, and that only the Roman agrimensores were able to thus lay out the country side. The latter were a corps of land surveyors trained in practical geometry using scientific instruments for their work, for “in practical mensuration a daily necessity for men who were perpetually alloting lands, or marking out camps, the Romans were experts.” (8.2) That the work was Roman, is further shown when the distance between several of the old ways was found to be 9 furlongs or 612 Roman poles, which in the ancient survey was the interval between the quintarial lines separating the saltus, and in the majority of instances, {9} it will be found that they run in parallels 120 R. p. apart, or multiples of this distance, the net widths into which the saltus could be divided (9.1).

The Saltus.—The side of a saltus used by the Surveyors in Britain measured along any one of the four quintarial lines which bounded the model saltus, was 612 Roman poles or 1980 yards, and the square of this gives 1300·5 jugera, equal to 810·353 acres as the superficial area of the saltus. (9.2) Now a side of a centuria measured 120 R.p., and since 25 centuriæ when grouped in rows of 5 each way went to a saltus, the length of a row was 600 R p., and the superficial area of the group was 1250 iugera equal to 778·886 acres. The difference therefore between a saltus and its 25 constituent centuriæ was practically 50 j, which, though equivalent to another centuria, viz. a 26th, could not form it, for the centuriæ used in the Middlesex district was in shape a square. These 50 j, as will he shown, were to provide the necessary ways giving access to the 25 centuriæ, and to their several minor divisions where needed.

The width of the district ways was not great (9.3) for the breadth of Roman vehicles was about half that of those in modern use. This is shown by the wheel ruts in the pavement being only three feet apart, between the stepping stones across the streets of Pompeii. (9.4).

Now the width of Watling Street, one of the most important Roman thoroughfares across Britain, when disclosed near the Marble Arch in London, was 2½ Roman poles equal to 24 feet between the curbs, and in another part of Middlesex an undoubted Roman vicinal way averaged about 18 feet between its banks, which appeared to have remained practically undisturbed. The width of a via through the woods in the Londinium civitas may therefore fairly he taken to have been fixed by the officials at 2 R.p. about 19·4 feet, and that of a devia at 1 R.p. or 9·7 feet. In colonies founded by Augustus they were respectively 12 feet and 8 feet.

The superficial area of these district ways when open at both ends to the public was not included in the assignments of the adjoining {10} land (10.1) though its occupiers had, under the supervision of the Magister Pagi to keep in repair their respective portions of the way up to a dividing line, running length ways down the centre of the roadway (ad medium filum viæ) (10.2). This was essential during Roman times for the agricultural traffic in parts of Middlesex must have been heavy, since there are fourteen ancient rural ways, which still bear their old name of ‘street’ (strata) from formerly having a paved surface of stone.—See map.

In order to understand the allocation of 50·5 j or 14,544 R. sq. poles between the roads, lanes and agricultural paths of a saltus, some further particulars are necessary. We will first take the actus quadratus a small measure of land containing 144 R. sq. p., traditionally derived from the length of a furrow made by a plough team of oxen. The yoking together ( juga a yoke) of two square actus made the well known jugerum (10.3) a little oblong field equal to about 5/8 acre.

Next we have the laterculus or early form of the Roman centuria with its 50 jugera or 100 actus, hence the latter name, and to lay one of them out into 50 plots as an ager jugarius, or village allotment field, it would have to be divided into 10 equal strips, and each strip into 5 plots of one jugerum each; four paths or plough balks either way, being provided to give access to the plots. Assuming for example .3 R.p. (2ft. 10ins.) to have been the width of the path, then the side of an allotment field plus its four paths would have extended to 121·2 R.p., and the area of the eight paths less overlaps would be 289·44 R.s.p. (10.4), and at this rate for 25 centuriæ 7236 R.s.p. Again since every centuria in a saltus might form a separate possession, provision for eight lanes 1 by 610 R.p. could be laid four each way, had to be made, and which, less 16 overlaps, would absorb Examples of Roman centuriation in Middlesex Diagrams used in the writings of the Gromatici Veteres {11} 4864 R.s.p (11.1). Lastly the remaining 2444 R.s.p. would provide for one half of the width (1 R.p.) of the four roads, down the centre of which ran the quintarial limites marking the extreme bounds of the saltus.—See diagram II, p. 13.

Therefore when computing the gross area of the surveyed portion of a Roman pagus in the Middlesex district, an amount equal to 1-25th of 778·886 acres, or one centuria to every twenty-five, must be added in each saltus for roads and paths to make up its full extent of 810·353 acres. No allowance for roads and means of access to the fields is included in the figures of the Domesday Survey.

Allotments and Holdings.—The original allotment or heredium was distributed by ballot to each settler apparently to enable him to make a start in his new surroundings, and to bind him to his settlement (vicus). It consisted of two jugera, about 1¼ acres, (though the number was increased in later times to 3, 5, 6, 7, 8 and 10 jugera) and accordingly amongst the terms used by the Roman surveyors were, “ager jugarius” or a field divided into 50 jugera; and ‘ager meridianus’ or the south field in 25 jugera. (11.2) As this heredium was insufficient to support a man and his family, he had the further right to graze his cattle upon the neighbouring common pastures a right which it seems was also appurtenant to the holdings in perpetual lease (emphyteusis) within the vicus. (11.3)

To create these allotments (agri dati asignati), two hundred jugera lying in four centuriæ, but not necessarily contiguous, might be set apart for distribution among every 100 men, and a centuria when thus divided would appear to lie in strips. Hence subsequently arose the term ‘virgate’ (virgatus striped), which in course of time seems to have become generally applied in England to this Roman measure of land, and equal to 31·155 statute acres. Under the Roman system, farm holdings were generally sold and leased in quantities of one or more centuriæ, and so, as we might expect, it will be found in the Domesday Survey of Middlesex, that virgate holdings were the most numerous. {12}

This preponderance of virgates shows that when the Romans parcelled out the Middlesex area, the agricultural holdings were based on the centuria, into which measure the land intra clausus had been sub-divided: and field divisions when once established are not wantonly disturbed even by newcomers. (12.1)

In the same way the term ‘ager ’ as applied to a piece of ground afterwards in the Saxon period became ‘acer’ a little field, roughly a modern acre. This is shown from a passage in a charter of A.D. 860. “Dimidium agrum quod nostra lingua dicimus healve aker.” (12.2)

In some parts of England the virgate was not equivalent to the centuria of 31·155 acres, and this was probably the case in districts which had not been previously surveyed by the Romans. Antiquaries however, are agreed that the virgate generally contained about 30 statute acres, which goes towards establishing its semblance with the pre-existing centuria, which the Saxons found in evidence in every Romano-British settlement throughout the Middlesex district. But the identity of the two measures is proved by the Domesday returns of the County, for if the centuria with its 31·155 acres be substituted for the virgate therein recorded, and the ancient land measures of Middlesex are worked out on this basis, then it will be found that the total Domesday acreage with the allowance for ways practically agrees with that of the modern Ordnance Survey. (12.3)

As the plots were distributed by ballot amongst the settlers drawn by lot into groups of ten, a man’s two plots forming his heredium would not necessarily adjoin one another. In the course of time the number of ‘half acre’ or jugerum plots scattered about the common arable fields, and in the hands of one owner often became considerable. The manor rolls of Winslow, temp. Ed III. give an example of 68 ‘half acre’ strips of arable land held by one individual amidst those of his neighbours in the common fields of that manor. (12.4)

On the other hand 200 iugera might be divided into four holdings (quadrifinium) or into three (trifinium) and in both cases {13} without any intervening public lanes or balks, the Surveyor’s provision for them having been assigned with the land. Again several centuriæ might be leased to form one large possession (latifundus): and four would make a farm of 125 acres, called by the Saxons a “hide” and, in the Middlesex Domesday, of the Villani or farmers, twenty-three are described as holding 1 hide each, three with 2 hides, and eighty-two with ½ hide (13.1). Middlesex district, diagram 2