of the supporting masses, the arrangement of structural lines; to the exterior, by the flying buttresses, the pinnacles, and the window tracery; to the interior, by the banded shafts, the capitals, the groined ribs of the vaults, and the openings of the triforium. Outside the church became a framework of glorified stone scaffolding; inside, an avenue of columns rising from the ground to the vaults, with intermediate spaces of tracery and coloured glass. But before this stage was reached there were many compromises and passing phases, and every considerable church in England, until the end of the fourteenth century, may be classified and criticised, not only for its beauty, but as a link in the development of Gothic architecture. The builders were grappling with both tendencies, the vertical and the horizontal; they were not consciously working on a theory of complete vertical development; they made progress by structural experiment, and a sensitive eye for possibilities of beauty; and in the meantime their problem, both structural and artistic, was to make a happy compromise between vertical and horizontal lines. It was a problem which probably presented itself to them in the question how they were to treat the different storeys of the building. Structural difficulties would be continually at war with their aesthetic ambitions, and the heavy stone vault made structural difficulties a serious matter. There was a growing desire for space, for height and width, for light and colour. With every increase of height and width the burden of the vault became more oppressive; with every enlargement of windows its supports were weakened. As a rule, the English builders were far less ambitious in their treatment of these problems than the French. Amiens Cathedral, begun at the beginning of the thirteenth century, is structurally as daring as can be. Salisbury, but for its spire, a later addition, is comparatively modest and timid. The French builders quickly reached the limits of structural possibilities, and their type became fixed. The English, with less economy of support, and a lower organisation of structure, were better able to play with their forms. So their churches present a series of continual and often inconsequent experiments in the treatment and proportion of every storey, particularly of the triforium, and in compromise between vertical and horizontal tendencies. Thus at Beverley, Salisbury, and particularly in the nave of Wells, the horizontal tendency is predominant, and the triforium is both important and continuous, without regard for the vertical division of the bays. In the Early English transept of the minster itself the triforium is the most prominent feature of the design. These are all examples of Early English work, but in the nave of Lichfield, which is Decorated, the triforium is still far more prominent than the clerestory. In the same way a various and experimental use may be noticed of the shafts dropping from the