place, the centre of the outer or the inner sanctuary? Again, the actual shape of the buildings is important. Ethnologists tell us that it is ultimately derived from a type of dwelling commonly occupied by primitive man, a circular wattle-hut with conical top; or even more directly, as some would have it, from a similarly shaped abode which the ancient Phrygians used to excavate in the ground, constructing the top of withies laced over beams converging to the apex and covered over with earth, while they tunnelled out an approach from one side where the ground sloped conveniently away[169]. From this it is argued that the domed chambers of Mycenae must be tombs, on the ground that ‘men in all ages have fashioned the dwellings of the dead in accordance with those of the living; but the dead are conservative, and long after a new generation has sought a new home and a new pattern for its houses, the habitations of the dead are still constructed in ancestral fashion[170].’ I readily admit conservatism in all religious matters; but how does the argument touch Mycenae? Archaeologists, and among them Schuchhardt himself[171], are agreed that the shaft-graves in the citadel are earlier in date than the bee-hive structures of the lower town. There was therefore a breach in the continuity of the ancestral fashion. Reversion to a disused fashion is a very different thing from conservatism in upholding an unbroken usage. But even supposing that there were good evidence of the uninterrupted continuity of this type of sepulchre, may not the temples of Chthonian deities have been built on the same plan? The use of the old word μέγαρον suggests that the sanctuary of Demeter and Persephone, though subterranean, was modelled on the dwellings of men, and, to borrow an argument, religious conservatism may well have preserved for the gods’ abodes the hut-like shape of primitive man’s dwellings long after a new type of house had become general among mortals. Concrete instances of this actually existed in much later times[172]. In Rome the temple of Vesta was of this primitive shape, and so also most probably was the Prytaneum of Athens, which, though not a temple, contained the sacred hearth of the whole community and a statue of Hestia[173]. Demeter then, as one of the deities of primitive Greece, might well have been provided with a temple constructed on the same primitive pattern as that of Vesta, but subterranean, as would befit a Chthonian deity, and thus analogous to the cave wherein she had been wont to dwell. The large domed chamber would be her _megaron_, wherein her worshippers assembled just as guests assembled in the _megaron_ of a prince. The small square apartment, where such exists, opening on one side of the main room, might be the παστάς or ‘bedchamber,’ an inner sanctuary which temples of later ages also possessed. The approach or ‘dromos’ would represent the natural cave which had given access to her fabled palace in the bowels of the earth.