In alluding to discoveries like that of Pytchley, we approached the subject of cemeteries, rather than that of barrows. A few instances of pagan burial-grounds lying beneath Christian churches may be cited. The oft-quoted case of St Paul’s Cathedral does not properly fall under this head. It is true that, at various times, a number of ox-skulls and boars’-tusks have been discovered beneath the foundations. Tradition says that the cathedral stands on the site of a temple dedicated to Diana[231]. The legend may be fallacious, for the finding of a heap of animal bones scarcely warrants the assumption of a pagan temple, much less of a pagan burial-place. Rather is the indication towards one of those foundation sacrifices, which might profitably engage attention in another volume. Moreover, the site of St Paul’s has always been a prolific field for Roman relics, hence it is within possibility that the bones are accidental items of a greater depository of rejected remains. Other records, however, may pass unquestioned. At Lewes, in Sussex[232], and at Mentmore[233], in Buckinghamshire, churches have been built, if not on the actual sites of Saxon cemeteries, at least, very near them. With respect to the Mentmore interments, it is to be noted that those bodies of which the positions were recorded lay East and West[234]. According to the view, now widely held, such a position indicates Christian burial, but, as will later be shown (p. 248 _infra_), the rule is by no means absolute. Unfortunately, the concomitant relics were so few as to yield little support to either theory. This difficulty is peculiarly noticeable in churchyard discoveries. Either the records date from the pre-scientific period of excavation, or, from the nature of the case, little modern exploration can be attempted. Thus, numerous relics have been dug up at various times near the West side of another Northamptonshire church,--that of Whittlebury. These relics comprise a bronze celt, Roman coins, an inscribed legionary tile, and several uninscribed tiles[235]. Such articles may suggest a burial-ground, or, at least, an inhabited site, but obviously no systematic excavations can be made.