VI. That limitation, however, being by such circumstances determined, by what means, it is to be next asked, may the actual magnitude be best displayed; since it is seldom, perhaps never, that a building of any pretension to size looks so large as it is. The appearance of a figure in any distant, more especially in any upper, parts of it will almost always prove that we have underestimated the magnitude of those parts.
It has often been observed that a building, in order to show its magnitude, must be seen all at once; it would, perhaps, be better to say, must be bounded as much as possible by continuous lines, and that its extreme points should be seen all at once; or we may state, in simpler terms still, that it must have one visible bounding line from top to bottom, and from end to end. This bounding line from top to bottom may either be inclined inwards, and the mass therefore, pyramidical; or vertical, and the mass form one grand cliff; or inclined outwards, as in the advancing fronts of old houses, and, in a sort, in the Greek temple, and in all buildings with heavy cornices or heads. Now, in all these cases, if the bounding line be violently broken; if the cornice project, or the upper portion of the pyramid recede, too violently, majesty will be, lost; not because the building cannot be seen all at once—for in the case of a heavy cornice no part of it is necessarily concealed but because the continuity of its terminal line is broken, and the length of that line, therefore, cannot be estimated. But the error is, of course, more fatal when much of the building is also concealed; as in the well-known case of the recession of the dome of St Peter's, and from the greater number of points of view, in churches whose highest portions, whether dome or tower, are over their cross. Thus there is only one point from which the size of the Cathedral of Florence is felt; and that is from the comer of the Via de' Balestrieri, opposite the south-east angle, where it
