ABSTRACT
The graphic demonstration of negative evidence, however, is problematic. Whilst such connections as exist among this sample can be shown in the same manner as for the medieval layouts, the lack of more alignments can only be demonstrated by an absence of more lines. Without proof that even the medieval examples investigated here were so designed, this must remain a theoretical possibility for all layouts. Yet, in consisting of structural grids of repetitive bays divided into major and minor spans, even sometimes incorporating transepts and a crossing, such buildings are otherwise analogous to medieval abbeys and cathedrals. To begin with, Truro Cathedral was included because, being a Gothic Revival design, its layout was presumably meant to resemble that of a medieval cathedral. Even if the system did fit medieval plans by chance because of an intrinsic versatility, random coincidences would produce random connections, as in the case of Smithfield, without any governing pattern necessarily being apparent.
