opinions were on that occasion (as often previously) expressed with that positiveness which is said to mark the true antiquarian !—a positiveness, in some instances, that had little foundation in real knowledge or personal enquiry, but rested chiefly on à priori arguments or purely negative criticism. In addition to the names mentioned above it is but necessary for me to allude to those of Mr. J. T. Micklethwaite, Mr. G. Fox, and others, to shew that no pains have been spared, and no professional attainments or special knowledge wanting, to determine the issue on a scientific basis. It may be true to the experience of human nature, but yet it seems a feeble conclusion, if we confess that, after all this apparently exhaustive discussion, the controversy on the main point is as much alive as ever.

Premising that by "the Chancel" is meant the original Chancel extending 20 feet eastward from the Nave, I may state the following four as the only theories that now hold the field :

(1) A Roman date for the Chancel, and a later Roman date for the Nave; or (2) vice versá—though this theory, formerly much in vogue, is at present out of fashion.

(3) A Roman date for the Chancel, and a Saxon date for the Nave.

(4) An early Saxon date for the Chancel, and a later Saxon date for the Nave.

It is indeed some consolation to friends and lovers of St. Martin's to know that even the anti-Roman disputants ascribe portions of it to such respectable antiquity as the time of St. Augustine (1300 years ago), and therefore, whatever ultimate conclusion may be arrived at, it is still the earliest existing Church in the island of Great Britain.

Many of the architectural details bearing on the subject are so minute, and so highly technical, that they are not suitable to the character of this Paper, so that I purpose to confine myself rather to broad general features, and to narrow the controversy (at any rate in the first place) to the question whether there still exists in the Church any Roman workmanship, or whether even the most ancient part of it must be assigned to the Saxon period. It is difficult to avoid tedious recapitulation of many points that are thoroughly familiar to those who have studied the subject, and some of which have appeared over and over again in print-but it seems advisable to record them in the pages of Archæologia Cantiana, as there must be many of its readers to whom the details of the dispute are still only partially known.

The principal arguments in favour of the Roman date of portions of the Church are these :

(1) History.—It is distinctly mentioned by Bede that there was (at the coming of St. Augustine in 597 A.D.) "on the East side of the city a Church dedicated in honour of St. Martin, built of old while the Romans still occupied Britain." Now this is direct testimony to which the greatest weight must be allowed, when we consider the character and authority of the writer. He was born in the year 673 A.D., i.e. only seventy-six years after the mission of St. Augustine and sixty-nine years after his death, and wrote his Ecclesiastical History in the first part of the eighth century (sometime before 735 A.D., when he died), taking the greatest possible pains to make it worthy of its subject. His information with regard to the history of Christianity in Kent was derived from Albinus, Abbot of St. Augustine's, who was

VOL. XXII.

с

himself a pupil of Theodore (Archbishop of Canterbury in 668 A.D.) the great consolidator of the English Church. We are told that Albinus referred to the records in his keeping, and sent Nothelm, a priest of London, to search the archives at Rome, where were preserved some valuable letters of Gregory the Great and other subsequent Popes. Considering then the extreme carefulness of Bede, and the sources from whence he derived his materials, we cannot imagine any evidence (short of first-hand) more trustworthy and valuable-and it seems to me that the Roman origin of St. Martin's Church might almost be accepted as proved if it depended upon the testimony of Bede alone.

That he should have written as he did, making a positive statement that the Church was built during the Roman occupancy of Britain, while all the time it owed its foundation to Queen Bertha or St. Augustine, is perfectly incredible. Were the latter theory true (as is maintained by some antiquarians now), would it not, in Bede's time, have been an easily ascertained fact, capable perhaps of documentary proof, especially among those who were inmates of St. Augustine's own monastery, and would have claimed St. Martin's Church as a specially precious inheritance-the legacy of their founder?

The only way that can be found out of this dilemma is to throw doubt on the genuineness or truthfulness of Bede's narrative, but no one has yet ventured in sober earnest to impugn his accuracy as a historian. The weight of historical evidence of this kind with regard to architectural facts cannot be too strongly insisted upon, for it is infinitely more valuable than any conventional ideas as to the supposed

character of a building, which confessedly varies to some extent with the materials ready to hand, the skill and capacity of the workmen, and whether it was erected in the zenith or decadence of the style adopted.

A priori then we may assume that there was a Roman Church in existence on St. Martin's Hill when St. Augustine came to Canterbury. Can we find any evidence in the present building which would strengthen the conclusion that portions of this Church are still standing?

(2) I have already alluded to the pink plaster, patches of which are found here and there throughout the Nave, and though well aware (as previously stated) that plaster of a somewhat similar kind has been met with in many Churches of a subsequent date, yet I must again lay stress on the point that this particular plaster has been pronounced by the greatest experts (including Mr. J. T. Irvine), after careful analysis, to be Roman, and to be distinguishable from later imitations by its hardness and texture, and the smaller admixture of sand. No perceptible difference can be detected between a piece of pink plaster stripped off the South Wall of the Nave and one taken directly from the undoubted Roman Villa at Wingham.

(3) The windows lately discovered in the West Wall of the Nave are by every one allowed to be built more Romano. The variation of the mortar used in their construction, from white mortar in the jambs to pink mortar in the voussoirs of the arch, is a very noticeable feature, and can be exactly paralleled in the Roman Pharos at Dover. It is certainly primá facie a strong evidence of Roman workmanship.

The objection that "Roman windows were never splayed may be met (a) by the general statement that the introduction of light by means of a splay is so natural that the idea could not have escaped a Roman builder, especially in countries where there was less light than in Italy. Isidore of Seville, a contemporary of Gregory the Great, living in the midst of Roman work, must be describing what was the distinctive features of windows around him when he says (1. xv. cvii.): "Fenestræ sunt quibus pars exterior angusta et interior diffusa est, quales in horreis videmus ;" and (b) Mr. Roach Smith, in his Collectanea Antiqua, gives several illustrations of Roman splayed windows at Arles, Vienne, etc. (see vol. v., p. 42; vol. vi., p. 241, etc.); and I am informed (though I have not verified the fact) that there is one at South Shields mentioned by Mr. Robert Blair, F.S.A.

(4) An ecclesiastical architect describes walls of Roman masonry in this country as "chiefly constructed of stone or flint, according to the part of the country in which one or the other material prevailed, embedded in mortar, and bonded at certain intervals throughout with regular courses or layers of large flat bricks or tiles which from the inequality of thickness and size do not appear to have been shaped in any regular mould." This account almost exactly describes the character of the walls in the Nave of St. Martin's Church up to a certain height, and especially where these walls have been practically undisturbed behind the present woodwork. Here, in many cases, the bonding courses are 9 inches apart. Roman tiles vary in length from 2 ft. to 15 in., and in thickness from 3 ins. to in.