aOTHIC
668
GOTHIC
we may trace the progress year by year and at the
hands of diverse peoples. By the beginning of the
tenth century, the available supply of ancient columns
having become exhausted, square piers built up of
small stones had everywhere taken the place of circu-
lar monolithic shafts, but the old basilican system
remained intact (except in the polygonal, Carolingian
churches), arcades supporting roof-bearing walls
pierced by narrow windows, and an enclosing wall in-
dependent in its construction and forming aisles cov-
ered by lean-to roofs of wood. In Sant' Eustorgio at
Milan (c.'JOO) we find evidences that transverse arches
were thrown from each pier of the arcade to the aisle
wall, so necessitating the addition of a flat pilaster to
each pier to take the spring of the arch. These arches
may have been evolved for the purpose of strengthen-
ing the fabric, or for ornamental reasons, or in imita-
tion of similar arches in the Carolingian domical
churches; but whatever their source the fact remains
that they form the first structural step towards the
evolution of the Gothic system of construction. Next,
transverse arches were thrown across the nave, the
first recorded example being the church of SS. Felice
e Fortunato at Vicenza, dated 985. Neither for struct-
ural nor a?sthetic reasons was it necessary that these
nave arches should spring from every pier, so every
alternate pier was chosen, the intermediate transverse
aisle arch being suppressed and the pier, that no
longer had a lateral arch to support, reduced in size.
To support the great nave arches, pilasters were of
course attached to the nave face of the pier, and these,
as well as the aisle pilasters, were made semicircular in
plan. If we assume, as we may, that in other exam-
ples all the transverse arches of the aisle were retained,
while only each alternate pier bore a nave arch, we
shall have a plan made up of compound piers support-
ing longitudinal and transverse wall-bearing arches
that divide the entire area into squares, large and
small, the great square of the nave being four times the
area of each aisle square.
The next step for a people on the highway of prog- ress would be the vaulting, in masonry, of these Sfiuares, for the wooden roofs were inflammable; more- over theCaroliiinian liviildcrs had constantly so vaulted their smaller siiuaic roof areas. The process began at once, and of course with the aisle squares, where the structural problem was simplest. The date is not re- corded; no early examples remain in Lombardy, but in Normandy we find, about 1050, churches which pos- sess aisles covered by square, groined vaults, with the transverse arches showing. The next step was of course the vaulting of the great squares of the nave, but before this was attempted the rib Vault was de- vi.sed, and the task rendered structurally more simple. The old transverse aisle arches had given the hint;
where an aisle so spanned was to be vaulted, the arches
already in place formed a very convenient shelf on
which some of the vault stones might rest, and, by so
much, a portion of the temporary centering might be
dispensed with. Intelligence could not fail to suggest
that an expedient useful in the case of the transverse
arch might be equally useful in that of the diagonals,
which were far more difficult of construction, as well
as the most liable to give way in the case of ribless,
groined vaults. When did this era-making invention
take place, and at the hands of what people? Where,
we shall probably never know, nor yet the exact date;
but it could not have been earlier than 1025, nor later
than 1075. San Flaviano at Montefiascone, authenti-
cally dated 1032, has aisles with rib vaults which are
possibly original and, if so, are the earliest on record,
while the nave vault ofSant'Ambrogio at Milan (c. 1060)
is of fully developed rib construction. " The most re-
cent authorities (such as Venturi, Storia dell' Arte
Italiana, 1903, who cites Stiehl, 1S9S) accept the view
that the vaults are of foreign fashion derived from
Burgundy, and were about contemporaneous with the
campanile [1129]. . . . It seems that on the evidence
we are compelled to suppose that vSant' Ambrogio de-
rived its scheme of construction from Normandy. It
may be that the origin of the vault is to be sought for
m Normandy, or even in England; but there are many
reasons for thinking that the seed idea, like so many
others, came from the East." (W. R. Lethaby, "Me-
diaeval Art", IV, 109-111.)
In all probability the Lombards are the originators of this device so pregnant of future possibilities. The new vault, groined, ribbed, and domed, was in a class by itself, apart from anything that had gone before. Particularly did it differ from the Roman vault in that, while the latter had a level crown, obtained liy using semicircular lateral and transverse arches and elliptical groin arches (naturally formed by the inter- section of two semicircular barrel vaults of equal ra- dius), the "Lombard" vault was constructed wifli semicircular diagonals, the result being that domical form which was always retained by the Gothic build- ers of France because of its intrinsic beauty. Finally, the new diagonals suggested new vertical supports in the angles of the pier, and so we obtain the fully devel- oped compound pier, which later, at the hands of the English, was to be carried to such extremes of beauty, and to form a potent factor in the development of the pure logic of the Gothic structural system.
The last step in the working-out of the Gothic vault- ing plan remained to be taken — the substitution of oblong for square vaulting areas. This was finally accomplished in the Ile-de-France after various Nor- man experiments, the evidences of which remain in the vaults of St-Georges de Bocherville and the two great abbeys of Caen. The sexpartite vaultings; of the latter, together with that of the five other similarly vaulted Norman churches and of the choir of St-Denis at Paris, has always been an architectural puzzle, since it is manifestly a stage in the development of the oblong quadripartite vault, and yet is found in these cases some years after the latter system is known to have been fully understooil in France, and nearly three- quarters of a century later than the vault of Sant' Am- brogio. There is reason to suppose that it is a revival of some of the curlier experiments in the development of the large, oblong, high vault from the small, square, aisle vault. It is conceivable that sexpartite vaults may once have existed in Lombardy and before the quadripartite vault was evolved; this would explain the persistence in Sant' Ambrogio of the vaulting sliafts on the interninliate piers, for which no apparent rea- son exist s. The vault of the Abbaye aux Dames may be consiilered either as a ribbed quatlripartite vault of square plan, bisected and strengthened by a trans- verse arch with solid spandrels, or as a series of trans- verse arches, one on each pair of nave piers, with the
