A more damaging argument is supplied by Mr Airy’s assertion that he had never met with one church pointing to the place of sunrise between the 1st of May and the 9th of August[557]. If this testimony were unassailable, it would seem to indicate at least an approximation to the rule of Durandus--a pushing back of building operations towards early spring (March 22nd) or forwards to autumn (Sept. 22nd). Unfortunately, we do not know the church which, according to Mr Airy, has a deflection of 30°; unless it be situated in the Northern part of Britain it must fall between the 1st of May and the 9th of August. One cannot agree with Mr Airy’s contentions that festival orientation involved too great a knowledge of the variation of sunrise, that the “ancients” [apparently the Mediaeval builders] did not observe such “refinements,” that they had no idea of the sun’s movements “in his course Northward and Southward to the τροπαὶ ήελίοιο” [the solstices]. On the contrary, we believe that from very early times rough observations of this kind must have been taken with some frequency. The possession of this knowledge by our forefathers does not, of course, prove the case for the Saint’s Day theory. But between the idea of dedication alinements and Mr Airy’s conclusion that the variations are accidental, and represent attempts to make the fabrics bear to the East as nearly as possible, there is room for a middle opinion. Should Domville’s assertion prove true, then, whatever may have been the practice of the early Fathers, the Mediaeval builders were imbued with the symbolism of festival dedication and alinement, and would not fail to give effect to their ideal. Granted that this practice cuts at the root of the equinoctial precept, there is yet a possibility that, as with so many other customs, uniformity was never attained. Exceptions alone, numerous though they be, do not disprove the theory altogether. A smaller number of positive cases would have force against them. I incline to the belief that the Mediaeval builders followed partially the Saint’s Day principle--perhaps invented it--and that a limited list of churches may conform to the idea. If John Leake’s map, already mentioned, be really accurate, the various City churches dedicated to St Mary have approximately parallel axes, and this evidence, so far as it goes, would be confirmatory. That the explanation covers even the majority of instances is very doubtful. And, at most, it accounts only for the symbolism which produced variations of the axial line, and for the actual routine of the builders. The master reason for orientation, whether equinoctial or solstitial, is to be sought, as we have seen, in much more primitive times.