in 1792, but the rather imperfect plans and general draw-
ings which have been preserved are sufficient to show the
strong influence of Aix, and of S. Vitale, its prototype. The
plan was that of a basilica ending in a large rotunda with
an open eye, as originally at Aix and at the Pantheon at
Rome, and the first church of the Holy Sepulchre.

William had hardly begun St. Benigne when Richard II,
Duke of Normandy, invited him to to reform the
abbey of the Trinity there, and generally "to found mon-
asteries and erect buildings." The abbot of Cluny himself
had refused to go, and so did William at first, because "the
Norman Dukes were barbarous, more disposed to destroy
than to raise sacred buildings," but he consented later, and
before he died is said to have founded forty monasteries
and restored many others. Two or three points should be
noticed which seem peculiar to his work. The tower of
St. Benigne is probably the first example in Western Europe
north of the Alps of the type first seen in S. Satiro, Milan
( ninth century) (p. 175 ), and in the basilican part of the
church first appeared the eastern chapels in the transepts,
characteristic of that plan of Norman churches which first
appears in those built by William's pupils, or by Lanfranc.
At Dijon, also, is found for the first time the cubic capital
which became so characteristic of Norman work, while the
great triforium galleries were either invented by the Cluniacs,
or else adopted by them from Lombardy, partly perhaps in
order to light the space between the upper surface of the
aisle vault and the lean-to roof. In the original basilica, with
its wooden-roofed aisle, no such gallery was needed, nor in
Provence where the roofing slabs rested directly on the
surface of the vault.

But the church which William built at Bernay for Judith
of Brittany, the wife of Richard II, still remains (p. 188 ). It
had apparently been begun before she died in 1017, and her
husband Richard continued it after her death till at least
1030. Its construction is a puzzle. M. Poret, who is the
most recent investigator of the building, considers that
the original piers of the nave were rectangular as at St.
Remi of Reims, but that they had not been carried to any
height. To these he thinks that Richard added a shallow
pilaster, and a half column carrying a roll of the same size.

-190-