Stones of Venice, The, volume 1


Fig. XXXV.

I. In the preceding enquiry we have always supposed either that the load upon the arch was perfectly loose, as of gravel or sand, or that it was Mont-Cenisian, and formed one mass with the arch voussoirs, of more or less compactness.

In practice, the state is usually something between the two. Over bridges and tunnels it sometimes approaches to the condition of mere dust or yielding earth; but in architecture it is mostly firm masonry, not altogether acting with the voussoirs, yet by no means bearing on them with perfectly dead weight, but locking itself together above them, and capable of being thrown into forms which relieve them, in some degree, from its pressure.

II. It is evident that if we are to place a continuous roof above the line of arches, we must fill up the intervals between them on the tops of the columns. We have at present nothing granted us but the bare masonry, as here at a, Fig. XXXV., 145 and we must fill up the intervals between the semicircle so as to obtain a level line of support. We may first do this simply as at b, with plain mass of wall; so laying the roof on the top, which is the method of the pure Byzantine and Italian Romanesque. But if we find too much stress is thus laid on the arches, we may introduce small second shafts on the top of the great shaft, a, Fig. XXXVI., which may assist in carrying the roof, conveying great part of its weight at once to the heads of the main shafts, and relieving from its pressure the centres of the arches.


III. The new shaft thus introduced may either remain lifted on the head of the great shaft, or may be carried to the ground in front of it, or through it, b, Fig. XXXVI.; in which latter case the main shaft divides into two or more minor shafts, and forms a group with the shaft brought down from above.

IV. When this shaft, brought from roof to ground, is subordinate to the main pier, and either is carried down the face of it, or forms no large part of the group, the principle is Romanesque or Gothic, b, Fig. XXXVI. When 146 it becomes a bold central shaft, and the main pier splits into two minor shafts on its sides, the principle is Classical or Palladian, c, Fig. XXXVI. Which latter arrangement becomes absurd or unsatisfactory in proportion to the sufficiency of the main shaft to carry the roof without the help of the minor shafts or arch, which in many instances of Palladian work look as if they might be removed without danger to the building.

V. The form a is a more pure Northern Gothic type than even b, which is the connecting link between it and the classical type. It is found chiefly in English and other northern Gothic, and in early Lombardic, and is, I doubt not, derived as above explained, Chap. I. XXVII. b is a general French Gothic and French Romanesque form, as in great purity at Valence.

The small shafts of the form a and b, as being northern, are generally connected with steep vaulted roofs, and receive for that reason the name of vaulting shafts.

VI. Of these forms b, Fig. XXXV., is the purest and most sublime, expressing the power of the arch most distinctly. All the others have some appearance of dovetailing and morticing of timber rather than stonework; nor have I ever yet seen a single instance, quite satisfactory, of the management of the capital of the main shaft, when it had either to sustain the base of the vaulting shaft, as in a, or to suffer it to pass through it, as in b, Fig. XXXVI. Nor is the bracket which frequently carries the vaulting shaft in English work a fitting support for a portion of the fabric which is at all events presumed to carry a considerable part of the weight of the roof.

VII. The triangular spaces on the flanks of the arch are called Spandrils, and if the masonry of these should be found, in any of its forms, too heavy for the arch, their weight may be diminished, while their strength remains the same, by piercing them with circular holes or lights. This is rarely necessary in ordinary architecture, though sometimes of great use in bridges and iron roofs (a succession of such circles may be seen, for instance, in the spandrils at the Euston Square station); but, from its constructional value, it becomes the 147 best form in which to arrange spandril decorations, as we shall see hereafter.

VIII. The height of the load above the arch is determined by the needs of the building and possible length of the shaft; but with this we have at present nothing to do, for we have performed the task which was set us. We have ascertained, as it was required that we should in VI. of Chap. III. (A), the construction of walls; (B), that of piers; (C), that of piers with lintels or arches prepared for roofing. We have next, therefore, to examine (D) the structure of the roof.


1 of 2
2 of 2