[Footnote 77: We have the highest authority for believing that one type can symbolise two things quite independent of each other, in that the Jewish Sabbath, commanded from Sinai to be observed in commemoration of the Rest after the Creation, is enforced in Deuteronomy as the representation of the rest of the children of Israel from Egyptian bondage. 'Remember,' says Moses, 'that thou wast a servant in the land of Egypt, and that the Lord thy God brought thee out thence through a mighty hand and by a stretched out arm: _therefore_ the Lord thy God commanded thee to keep the Sabbath day.'--Deut. v, 15.] Another objection is as follows: If this theory be true, how will you account for churches with nothing but a nave, or with only one aisle; how for churches with neither cleristory nor triforio; or, on the other hand, for those with double triforia, or with four or five aisles? Now we never asserted that it was necessary that all, or indeed any, given things should be intentionally symbolised. We have pointed out that some things are essentially symbolical; others accidentally and occasionally. We might attempt to classify what _must_ be symbolised in church building, and what _may_ be. But we decline to do so because we do not think that the principles of symbolism are yet sufficiently investigated or apprehended. However, in a general way, _every_ building must, from the nature of things, have some accidents, as of material, of parts, of plan; every particular building must have particular accidents, as of use and purpose. These accidents _must_ be symbolical, from their nature, in a general way: they may derive, from purpose added to their nature, a further or modified symbolism in a particular way. With the first sort it is that Durandus chiefly concerns himself A building must have walls, must have roof, piers, windows, corners, and floor. For each then he finds a meaning. {cxvii} He does not quite neglect the second sort. Early English windows must have a splay: the spire may have a weathercock: for these then there is an appropriate signification. So we do not mean to insist that certain things _shall_ be symbolised, we say they _may_ be symbolised. Perhaps when more is known, we shall be able to criticise ancient buildings, to show their faults or their shortcomings in this particular. As it is, we have framed a sort of _beau ideal_ of a church, fully formed and developed, which we should propose as a perfect model. We are not qualified as yet to blame the ancient churches which do not come up to this ideal, but we cannot be wrong in praising such as do.