“Autor hujus operis in Deus crederi, esse dele
peccata omnibus te Christe hic orantibus instat
praesens memorea indigno famulo tuo Froila qui
optat et in Domino te conjurat O bone et delecte
qui legis ulme peccatorem memorea habeas sacrata
ex oratione tua.”
Of course, if Froila put up the inscription himself, we have a clue to the date of the building, but the question is—How are we to be sure that it was not put up by some one else after Froila’s death?
I had heard that there was another little church with horseshoe arches about two hours’ drive beyond Celanova, the church of Santa Comba (Columba), near the village of Bande; and, wishing to compare the two, we stayed the night at the little inn at Celanova, and drove to see Santa Comba the following morning. We started about six a.m.
Between the little town of Banda (once a halting-place on the Roman road between Astorga and Braga) and the church the scenery was very like that of the moorlands of Scotland, with few trees and a good deal of bracken and furze. Although it was April, the oak trees scattered here and there still wore their brown leaves from the previous year; white stemmed birches lined many of the fields, and reminded us that we had reached a higher zone, for these trees are never seen in the low valleys or near the sea-level in Galicia; they need a sharper and more bracing atmosphere.
The church of Santa Comba stands on the side of a hill about an eighth of a mile distant from the coach road, and a winding sandy path leads up to it, skirting a picturesque village as it approaches the church. This part of the journey had to be done on foot, and beneath a blazing sun. What a quaint, archaic little church it was! Outwardly it was divided into three sections, rising one above another like three steps, each with a red-tiled gable roof. It was constructed of irregular blocks of granite,[300] roughly cemented together; there were two entrances, the principal one being the western wall and a lateral door on the southern side. In front of the principal entrance was a small Gothic portico, evidently of much later date than the body of the church. The most striking characteristic of this little edifice are its extreme simplicity and its horseshoe arches. It is built in the form of a Greek cross, all four ends of which are of equal length, though the eastern end is lengthened by a small chapel which serves as an apse. The central part of the church is not unlike that of San Miguel de Celanova, square, and supported by four slightly horseshoe arches. The arch leading to the little chapel is also horseshoe in shape, and much more pronounced. Above the four arches is a cupola formed by four walls, in two of which, the eastern and the southern walls, there are small windows. The church is covered with intersected cylindrical vaulting. Running round the walls above the arches is a banded zigzag impost of a very rugged character. In the southern arm of the cross, the northern arm of the transept, stands a Roman altar, ara, of white marble, from which the inscription has been effaced, and at the end of the transept stands a large marble sarcophagus carved from a single block, perfectly plain, without a trace of carving or inscription of any kind; the lid is also a monolith. Tradition says that this is the sepulchre of San Torquato, whose skull we had seen at Celanova. From the white stone above this sarcophagus Portuguese pilgrims chip fragments to take away with them, believing that its dust will cure inflamed eyes and other troubles.
The eastern arm of the cross, prolonged by an apse chapel, contains the chief altar, and is reached by a horseshoe arch, like those of the Cordova mosque. On either side of this arch, but quite separate from it, are placed two pairs of grey marble columns, each of a single piece, whose lower ends disappear beneath the ground, showing that the original floor of the church must have been much lower down than the present one. The capitals of these two columns are Corinthian, the columns themselves are Roman, and it is believed that they must have been brought from the old Roman baths of which ruins are yet to be seen at Bande, where people still come in the summer time to drink the mineral waters. The chapel is almost square, and covered like the rest of the building with intersections of cylindrical brick vaulting. There is a little hole in the wall beside the altar, for the Host; the little window behind is filled with honeycombed marble fretwork, which has rather a Moorish look. The flooring is composed of granite slabs, each with a hole by which to lift it. The walls are a yard thick. Although all the arches, including those of the portico, are more or less horseshoe in shape, the windows (five in all) are Romanesque. One of the entrances to the porch has been closed up. Over the chief entrance is an inscription, and the date 1670; there is also an inscription on the wall of the portico, declaring the edifice to be a church of refuge. There were churches of refuge all over Galicia until the eighteenth century.
For many years this diminutive church of Santa Comba de Banda, on the borders of Portugal, has been, like San Miguel, a hard nut for Spanish archæologists to crack. It has been written about and discussed over and over again, but mostly by authorities who have not taken the trouble to go and see it. I doubt if, among all the archæologists and architects who have touched upon the subject, there are as many as three individuals who have examined it personally. Even Lopez Ferreiro, the greatest archæologist in Galicia, has not yet been to see it! This writer has suggested that San Comba and San Miguel may have been built to serve the purpose of mortuary chapels. Of Santa Combe he says: “It is one of those very rare examples which represent, in the history of art, the continuation of the Byzantine style in its last period, that of transition to the Romanesque style.”[301]
The plan on which the church of Santa Comba is built is, we have seen, Christian—that of a Greek cross. It must therefore have been built by Christians, for Christian worship, but at what date? And how comes it to have these horseshoe arches? Is it an example of Mudejar architecture? Did some Moorish slave design it at the bidding of a Christian master? Let us compare it with that other little Christian church with horseshoe arches, St. Juan de Baño, in Palencia. The plan of the latter is rectangular; it is divided by two rows of horseshoe arches into three naves, and had before its renovation three square apse chapels, one at the end of the central nave and the others placed at right angles with the heads of the right and left naves respectively. This plan is unique; there is nothing like it in any part of Spain. French architects who have been to see the church shake their heads over the suggestion that it is a monument of Visigothic architecture. “True,” they say, “that it bears the name of King Recesvinto, and the date 661, but how can you tell when that inscription was put up, and whether it is correct?” Spanish authorities, however, are now unanimous in pronouncing St. Juan de Baño to be an example of Visigothic architecture constructed in the seventh century, before the earliest date at which the Moors invaded Spain. For years they quoted it in their text-books as the only example of Visigothic architecture in the Peninsula, but now they are positive that Santa Comba de Bande is another remnant of the same architecture and of the same period, for they have found it mentioned in a charter given by Adozno to San Rosendo as a church that had already been established more than two hundred years in the year 910;[302] and Santa Comba too is to be ranked as a national monument, as a precious relic of pre-Moorish Spain. But should not a careful comparison be made between Santa Comba and the strange little oratory of San Miguel in the garden of the Celanova Monastery? A drive of three hours is all the distance that separates them, yet no comparative study of the two has ever been made. The roofing and the general sculpture of the two buildings, as well as their horseshoe arches, are strangely similar, and what differences there are may quite well be due to reconstruction. In fact, I fail to understand why Santa Comba should be thought to be so much as three centuries older than San Miguel.
A thick low wall of granite surrounds the little grass plot upon which the church of Santa Comba stands, and is quite in keeping with the rest of the picture. The village close by, whose houses are built of granite and thatched with straw, is also rather old and quaint. The granite lintel of one of the cottage doors bears the date 1713. We ascended its wooden steps at the invitation of the woman who lived there, and found the furniture of the rooms very curious; it consisted chiefly of some very large wooden chests that seemed to be hundreds of years old. The woman’s old mother was occupied in spinning a counterpane. As we drove along we had