In an address at the Inaugural General Meeting of the Society of Architectural Historians of Great Britain, Mr. Howard Colvin distinguished three traditions in the study of architectural history. First, there is ‘the study of buildings as monuments whose history is contained in their own structure’; secondly, there is the ‘use of documents … to throw light on the history of a structure, on the mind of an architect, or on the building practice of an age’; thirdly, ‘there is the art-historical approach which is concerned chiefly with tracing aesthetic concepts’. The study of railway architecture in Britain, in so far as it has been studied at all, has hitherto been undertaken by the first and third of these methods. With regard to railway architecture, this is true of Mr. Barman's book and of the important studies of Professors Hitchcock and Meeks although they make some use of original drawings and plans. It is the purpose of this article to show that the second method also—the use of records—is necessary to a full understanding of the subject.