{489}

Notes and Documents

Centuriation in Middlesex

Professor Haverfield’s assertion that there is, ‘so far, no trustworthy evidence for centuriation in Britain’, for lack of ‘traces of roads laid out accurately straight, running in direct lines or at right angles’¹ is too sweeping. Evidence of centuriation more or less distinct is to be found in most Romanized districts of outlying Britain, but I propose to confine my remarks to briefly indicating how traces of the survey were first discovered in the Middlesex district, together with the historical information obtained therefrom—a research which extended over a period of ten years.

¹ Ante, p. 292.

The Romans are known to have been great agriculturists, and it can hardly be supposed that they continued the primitive methods of native cultivation, and did not extend the area brought under the plough. Had they not done so, Britain could not have become one of the fertile portions of their empire from whence grain was exported to the Continent. In the middle ages Middlesex was known for the excellent quality of the corn it produced, and prima facie in an earlier age the Romans were equally aware of the fertility of the soil in the valley of the Thames, which also contained their commercial town and port of Londinium.

Evidence that this area had been settled by a Romano-British agricultural population was obtained in this way. For some time past it had been noticed that many fragments of its ancient rural ways ran in parallel lines, and were crossed at right angles by similar ones, which in the several districts of the county were distinguished by a different orientation. Thus in the northeastern division the direction of the cardinal ways was from north to south: in the southern portion between the Brent and the Lea rivers, and into Essex, they pointed south-south-east. Over the south-western area and beyond the Colne into Buckinghamshire the course was south by west, and in the north-western district they were again south-south-east. Passing into that part of the Middlesaxon province lying south of the upper Colne and Lea, but now in Hertfordshire, the two orientations were {490} respectively south-east by south, and south by east. A further feature was that many crossways occurred at equal intervals, and along one road five in succession were found at distances of 120 Roman poles or 388 yards, two being roads, two foot paths, and the other an ancient field boundary, presumed to have been formerly a plough balk or a footway.

It was manifest that this laying out of land amounting to 181,000 acres could not have been the result of chance, but must have been carried out at a time when the soil was mostly in its primitive condition, by a conquering race who had seized it, and who were accompanied by skilled land measurers. All this pointed unmistakably to the Romans and their corps of agrimensores, trained in applied geometry and using scientific instruments. The writings of the Gromatici Veteres were next consulted for information as to the manner in which Roman lands were surveyed and laid out, and it is worthy of note that one of the most eminent of these writers was Sextus Frontinus, Propraetor over Britain from a. d. 74. Among the more enduring bench or land marks used by Roman surveyors were mounds of earth (up to the size of a small haystack), stones, and trenches, and in these three respects important discoveries have been made in the county. A mound (botontinus) is to be seen both in Cranford and in Syon parks, also at Hampstead, Stanmore, Hadley—where there are two—and just out of the county at Salthill, Slough. Two others have not long ago been levelled, one by Bushy Park and the other at Hillingdon, while local names apparently preserve the sites of half a dozen more. Four stones are still in situ; two marked on old maps no longer exist, and the former positions of several others can be located. Two trenches are still to be seen.

A map showed that these boundary marks and the remnants of the oriented ways were naturally co-related, that each district had been of nearly equal area, rectangular in form, and contained by a boundary line, the course of which was disclosed by the botontini and stones. It was also seen that these districts or pagi were in general identical in area with those of the later hundreds of the Saxon period, as set forth in Domesday. From the orientation of the pagi, the territorium of the Londinium canton appeared to extend from the foot of the Chiltern hills across Middlesex and into Essex; the pagi had been laid out by lines (quintarii) crossing one another at right angles, and so forming possessae, each of which according to the text-book, and in fact, contained 1,300 jugera equal to 810 statute acres. These in turn could be divided into 25 laterculi or small centuriae of 50 jugera lying in rows of five, plus an area equal to a centuria distributable over a possessa for lanes and paths. This provision, equal to one-{491}twenty-fifth of a surveyed area, was later on found to have an important bearing when comparing the total acreage of the Roman and Domesday surveys of the county, for the latter did not include road surface. A side of this square centuria measured 120 Roman poles or 388 yards, and five of them lining the face of a possessa accounted for those five successive equal intervals formed by crossways which were noticed upon a Middlesex road between Greenford and Ealing as above mentioned.

It is difficult to see how the large centuria of 200 jugera referred to by Mr. Haverfield could have been utilized, if the surveyors used the possessa as a measure of land, or the saltus with 1,250 jugera, its net or productive area less the road surface. I hope he will follow up his suggestion as to a possible trace of centuriation south of the Braintree–Dunmow section of the Roman road from Colchester to Bishops Stortford, though military roads or streets appear to have been laid down independent of the agrarian and centuriated ways through which they passed. In Middlesex, Ermine, Watling, and Tamesis streets bore no relation to these rural ways.

Two curious discoveries came to light after the quintarial cross-lines had been drawn, making each pagus appear like a gigantic chequer-board. The first was, that 47 out of 56 mother churches of parishes in Middlesex were situated upon one or other of these lines, the apparent explanation being that Romano-British chapels (compita) adjoined the principal rural ways, which were designed to follow the quintarial lines. In the next age these little edifices were adopted by missionaries for Christian worship, following the astute and well-known direction of Pope Gregory to utilize the pagan sacra where the people had been accustomed to assemble. If so, then such sites have been associated with public worship, first pagan, then Christian, for nearly 2,000 years.

The other discovery had an important bearing on the correct reading of the Domesday Survey of Middlesex, for it became evident that the centuria of 50 jugera, with its known area of 31·158 acres, was identical with the virgate of the Saxon period, the size of which has caused much controversy. The proof of this lies in the fact that if the Middlesex Domesday measures are worked out on this basis the total acreage for the county, which has not been changed in area since the ninth century, agrees, when the road surface is included, with that of the modern Ordnance Survey. All this bears testimony to the accuracy of the Imperial Survey, and to the diligence of the Domesday Commissioners.

Such evidence shows a more intimate connexion between Roman Britain and Saxon England, especially in matters relating {492} to rural economy and in the common law bearing upon it, than has hitherto been supposed. Further points can be adduced, of which the headings of only three can here be given: (1) The Roman settlers’ heredium of two jugera (a Saxon aker) in non-contiguous plots, and upwards to a centuria, all having compascua: followed in Saxon and later times by scattered holdings in the village farm in acre and half-acre strips, and amounting to virgates and halfvirgates, while all possessed appendant common pasturage. The average amount of land held by a bordar in Domesday Middlesex was five akers, and similarly that by a cottager two akers; of the larger holdings 438 villanes held each a virgate, and 426 each half a virgate lying in half-acre strips in the common farms of the villages. (2) The tributarius and colonus in Britain under decurions with the natives appear to survive in the geneat, gebur, and cosetla in their tithings during the Saxon period. (3) The Domesday geldage for Middlesex, with its decimal foundation upon the constant geld unit of five on the vills, curiously amounts to the same total as from the number of possessae when multiplied by that unit.

Montagu Sharpe.