St Mildred’s, Canterbury

BY R. HUSSEY, ESQ., F.S.A.

It is not uncommon to find fragments of Roman work in the walls of medieval buildings which occupy or are adjacent to Roman sites, and these remains are sometimes the only evidence of the earlier settlement,¹ as at Eynesford Castle, in this county. They usually consist of tiles or thin bricks, generally more or less broken, and sometimes of pieces of hard concrete or mortar built in at random with the ordinary materials of the walls;² but when the tiles are abundant they are used, especially in districts where stone is scarce, for quoins and for relieving arches over openings. Occasionally less rude and insignificant remnants of Roman work are met with in later erections, but they seldom amount to more than a few squared stones; these should always be carefully examined, to see whether they retain any original features of interest, and their geological character should be noticed, especially if they are not the produce of the neighborhood.

The church of St. Mildred, at Canterbury, has many fragments of Roman tiles built into the walls among the flints; and it is my firm belief that most of the stones of the two quoins of the south wall of the nave have also been taken from a Roman building; the majority of them are of larger size than are usually found in medieval work, especially of a date so early as this wall; and five of those in the western quoin, and six in the eastern, are of oolite, a material very rarely found in this district, in buildings contemporary with this church.

The lowest and largest stone in the western quoin is about 4 feet, by 2 feet 9 inches, by 1 foot 5 inches, and there are indentations upon it which show that it has been used for some other purpose. There is also, on one side, what appears to be a hole for a lewis (now stopped with cement), which, if it is so, implies that it once occupied a higher position in an earlier building. Each of these quoins contains one stone taken from a large arch; that in the western is 1 foot 11 inches long, 1 foot 3 inches wide at the upper or broadest end, and, to speak technically, 1 foot 8 inches deep in the bed; the other, in the eastern quoin, is broken, and I could not reach to measure it, but it seems to be of corresponding dimensions. I have not been into the church, and do not know whether the interior presents any characteristics to determine the date of the south wall. Externally the original features have been obliterated by subsequent alterations, but the construction shows that it is not later than the Early English period, and it probably may be older. It is very unlikely that at that time such stones as these could have been taken from any but a Roman building. I am too ignorant of geology to be able to say from whence these pieces of oolite have been brought, but it may be hoped that someone better informed will determine their native district. The Romans certainly carried oolite into this part of the country for building purposes, for fragments of it were found at the villa at Hartlip, when re-opened a few years ago; it is also met with at the villa at Bignor in Sussex; and a block of it lies within the walls of Richborough Castle, which was probably taken there by the Romans. The late Dr. Buckland, on a slight examination of a piece which he broke off this last-mentioned specimen, said he believed it came from the neighborhood of Weymouth.

Endnotes

1. Some small lumps of concrete which were found in the walls of the church at Frillenden, in this county, during the repairs in 1846, were, until recently, the only signs of Roman occupation in that neighborhood; but in the course of last year two sepulchral urns were dug up within about a mile of the church.
2. It is well known that Roman mortar and concrete may very frequently be distinguished by the redness of the pounded tiles and pottery with which it is made, instead of sand and gravel. The same materials may perhaps, occasionally, and to a slight extent, have been used in medieval buildings; but, with the exception of some very late works at Colchester, no specimen which could be mistaken for Roman has ever come under my observation. Mr. Hudson Turner, in the introduction to his 'Domestic Architecture in England' (p. xxvi.), discredits the idea that this peculiarity in mortar is a certain evidence of Roman date; and he gives a translated extract, from an account of the repairs of Newgate in 1282, in proof that pounded tile was used in mortar at that time. But his quotation is inconclusive; the document to which he refers is probably written in Latin, with contractions, and it may be that the broken tiles which he has supposed to be for making mortar, were provided pro aement., that is, "pro cementariis," for the masons, or wallers, and were intended to be used for some other purpose, perhaps to be laid under some of the thinner stones, where required to bring them up to the general level of the courses, as was frequently done in medieval erections. Tiles, mostly broken, were also much used in medieval times for the backs of fireplaces, and were considered to be peculiarly fit for works exposed to the action of fire. Antiquaries who wish to gain credence for opinions based on ancient documents, must set forth the documents fully, and in their original language.