It is outside our purpose to discuss the date of St. Pancras, though many authorities maintain the possibility of its Roman origin. But, granting (for the moment) that St. Pancras' Church was built or restored by Augustine (and this is the latest date assigned to it), the identity in plan and character of the two churches is disputable. Of course, taking St. Martin's as it now exists, there is no similarity whatever, either in regard to the masonry of the nave, or the general outline. There is more similarity (with the exception of the points above mentioned) between St. Pancras and the assumed shape of St. Martin's chancel. But here, too, are points of difference. The walls of _St. Pancras_ are only _1 ft. 10 in._ in thickness; they are constructed almost entirely of broken bricks, roughly cut to a triangular shape and fitted together in the core, the interstices being filled up with small bits of brick. The walls of St. Martin's chancel are _2 ft. 2 in._ thick, and contain a much larger proportion of whole bricks, about 12 inches wide, laid side by side in each course, the interval between them being filled up with mortar and small stones. We may mention also the difference in the treatment of the division between nave and chancel. In the churches of St. Pancras, Reculver, Brixworth, Peterborough, Lympne, and Rochester there was a triple chancel-arch. In St. Martin's the space is too narrow to admit of any such arrangement. If we carry back the original building of St. Pancras to Roman times (and we must remember that King Ethelbert is said by Bede to have allowed the Italian Missionaries to build and repair _churches_ in all places) we do away with the difficulty as "to the temple of the heathen god being built after the fashion of a Christian church." We may pass over, as unworthy of serious discussion, the argument that St. Martin's cannot be a Roman church, because no existing Roman churches have yet been discovered in this country! and that it is not Roman because its ground-plan does not tally with the ground-plan of the Roman Church at Silchester. In the first place, we do not know what the original ground-plan of St. Martin's was, and it has not yet been definitely settled whether it may not have possessed side-aisles. And secondly, to contend that it cannot be Roman because it is unlike the church at Silchester would be to limit the capabilities of Roman builders to one monotonous design, perpetually and exactly reproduced for a century or more, which would be contrary both to reason and experience.