Special attention has often been called to the semi-circular buttress, because this shape is uncommon, though something like it is found at St. Peter's, Northampton, at the Church of St. Remi at Rheims, and elsewhere. The outstanding portion of it measures almost exactly three feet in circumference. It cannot have been made (as some have supposed) to contain a staircase, because there seems no reason whatever for a staircase at this particular place, the rood-loft being several feet eastwards. Others have conjectured that the old church might have ended somewhere near this point, and that then the buttress would have had something to do with the support of the western front, or have been a staircase up to the old belfry. But there is no foundation for this surmise, which is disproved by the fact that the external plaster extends on each side of the buttress, and the character of the south wall is absolutely unbroken. This external plaster, indeed, is probably not Roman, though it is composed to some extent of pounded brick. The buttress bears little or no resemblance to the lofty semi-circular projection occasionally found in Saxon towers. Its object must be left in a state of obscurity, and it may perhaps have been a mere freak of the builder. [Illustration: Erudito viro et Amicissimo Johi Hardy de Nottingham. _Tabulam hanc vevet W. Stukley_ Date 1722.] At a distance of 10 ft. 6 in. from this circular buttress we come to a _nearly circular panel_, immediately behind the Norman piscina, which has always been a puzzle to antiquaries. The dimensions of it, as now seen, are roughly 4 ft. by 3 ft. 8 in. It is sunk 6 in. into the wall, is unevenly splayed, and in parts plastered. In Stukely's engraving of the church (1722 A.D.) it is represented as a round-headed doorway, but there are no voussoirs or arch stones. The result of excavations beneath the surface are doubtful. Generally speaking, there are courses of two Roman bricks running along this part of the nave wall, below which are Kentish rag stones, and a foundation of concrete. Singularly enough, the _top_ row of Roman bricks (just below the opening) has been interrupted for a space of 3 ft. 8 in., and it looks at first sight as if the _lower_ row were the sill of a doorway, from which a slight suspicion of a rough vertical joint goes upwards for a little distance. But against this theory we must state that the _one-brick_ course does not extend the whole width of the panel. The immediate back of the Norman piscina was discovered on investigation, not to be of stone, as we might have expected, but of coarse thin plaster, and it is not impossible that this back was taken out sometime in the Early English period, and that the opening thus made was used as a hagioscope. No plausible theory has been advanced as to the use of this _panel_. It was once suggested that it was a niche for a churchyard-light, which would shine on the south side of the church. This,