300 NOTICES OF ARCHAEOLOGICAL PUBLICATIONS. admirable in its proportions and of an arrangement combining both beauty and convenience." lie goes on to say, " I fancied in" looking over a series of en'^ravings illustrating the mediaival architecture of this part of the south of France, that I could recognise this type or model in some of the principal churches of the eleventh and twelfth centuries." We refi^ret that we cannot follow this e.amj)le of genuine Roman archi- tecture by the semi-Koman of St. Jean at Poitiers, of which Mr. Petit gives an anastatic view ; this singular buihling has straight-lined arches, and hrick as well as stone is used in its construction. An excellent photo- ijraph of the front will be found in that fine series of photographic views of French churches now exhibited in the gallery over the nu'diaival courts in the Crystal Palace at Sydenham. Of that extremely rare class of buildings, which seems to belong to a period somewhat before the year 1000, we have a very remarkable instance in the church of Courcome, near Ruffec (See Woodcut, No. II.) ; there is much in it as in some of our probably Saxon churches, which resembles debased Roman rather than Romanesque or Norman. At about this period, several distinct schools of architecture make their appearance in France, the precise discrimination of which does not seem to be an easy task ; three of these Mr. Petit thinks deserve peculiar attention from the architect who wishes to revive mediaeval architecture in the present day ; these are the styles of Auvergne, of Perigord and Angoumois, and of Anjou. The first of these is principally characterised by the barrel roof, the second by being roofed with a series of domes, and the third by square compartments of cross vaulting much raised at the apex. The two last arc usually without lateral aisles. The first of these would appear to be the common type of the French Romanesque ; it as clearly originated from the Latin or Biisilican as the second did from the Byzantine form of church; the third, or Angevine style, appears to be a modification of the second. Of the churches of the first style, Mr. Petit has given us many very interesting examples ; two of the most important and remarkable of these, St. Scrnin, at Toulouse (See Woodcut, No. III.), and St. Etienne, at Nevers, are usually attributed to about the same date : the first, it is said, was finished (with the exception of the spire) between lODO and 1097,' while the latter, according to a writer in the Bulletin Monumental (vol. v, p. 17)*, was consecrated in 1097. They agree very nearly in plan, each liaving a long nave, with aisles, long transepts, a short choir ending in an apse, radiating apsidal chapels, and eastern apses to the transepts ; both liiive barrel vaults. Churches of this type, Mr. Petit says, prevail throughout Languedoc, Provence, and Dauphiny ; aiul he cites several examples in GuHcony. In the absence of a clerestory he observes, they resend)le the Lombard Homanescpie churches, with which, as well as with our Xorman, tlii-y nearly correspond in plan. St. I'jtienne, at Nevers, has some peculiar featur(!H in the straight lined arches in the transepts, and the slnnt shafts in th(! triforiunj of the apse, which swell out in tin! middle, and closely resemble the baliiHtrcH of our Saxon churches. The singular brackets which carry the cavcH of the ap.so arc almost exactly like those which occur in the church of Ainay, at Lyons. The churchcH of the second style, namely, of llu^ l>y/.anlinc type, are of • Mr. I'l'tit ilcK'H not t;ivc any 'lair l<p in;,' I" Iihm' liarril «liillnr ln' lidiovub lliin liuililin^ ; it wdiilil liavi- In. n imnihi tin .liitc nsiiilly ^ivi ii tn lie correct.
