(most likely of the pre-Norman period), 3 ft. 3 in. wide, and 3 ft. from either end; the jambs not splayed but running straight through at right angles to the walls. Among the debris in this porticus were found several pieces of glass and of fused bronze, and portions of a door and late Tudor window; and, deeper down, two or three fragments of Roman pottery. Under a close layer of brick earth, seven or eight inches below the tile pavement, is a floor of concrete, shewing in parts marks of fire. This concrete floor seems to extend beneath the present altar, and is also traceable in adjacent portions of the nave, and again at the approach to the chancel, where we discover something like steps. I need only add that the floor of the porticus was originally on the same level as that of the rest of the nave, but was raised one step above it when the later church was built.
   Everywhere throughout the excavations are evident traces of burnt earth and other calcined substances.
   It is to be borne in mind that the foundation-walls throughout (being twenty inches wide) are composed of Roman tiles bound together in some places by salmon-

coloured mortar, in others by mortar made from sea-shells and pebbles, and even later material.
   These are the facts and the data. What conclusion then are we to draw? That there was on this spot some early Roman building, whether of a secular or religious character, is indisputable. There is a vague tradition that there was once here a Romano-British church, and this having fallen into decay may have been partially restored and used by Ethelbert for a heathen temple. We have Thorn's story, written 500 years ago, and it is of course possible that he had consulted earlier records. It seems to me incredible that he should have written as he has done if the first church on the spot had been of Norman work, built only some 200 or 300 years before his own time; for he was a monk of St. Augustine's, and had free access to their chronicles. His testimony therefore (though not to be implicitly received) must, I think, be entitled to some weight.
   But we must chiefly rely on the excavations themselves. The Roman tiles are pronounced to be of a good time, and