We find ourselves in front of one of the best singular buildings of all Spanish architecture.

From this monument, that had not been thoroughly studied until 1947, we know it was recorded on a document of the 8th century, where the upper church of Santa Eulalia is mentioned. This indicates it was a double plan building and that its lower vault was damaged when the upper plan was tried to be fitted out for burials. It is now manifested as a series of unknown factors, extremely difficult to decipher due to the existing incompatibilities between some typical characteristics of the Roman architecture, as the painting in the interior decoration and the design of its structure, clearly classical; and others of basically barbarian and pagan sign, like the sculptured decoration of the external walls or the existence of a nave's entrance door with the most ancient horseshoe arch in Spanish architecture as a structural element, as before then, it had only appeared in the decoration of some Roman steles.

This makes dating very difficult, since if it were Roman or Paleo-Christian, it should be located in the 4th century, whilst we consider the carved decoration on the outside and, above all, the existence of the horseshoe arch, it could be thought as a Swabian construction of the 6th century.

The original building consists of a sort of double vaulted crypt, with a structure similar to that of San Antolín in the Palencia Cathedral; squared shape, around 6 m by side, floored in alabaster with a pool in the middle that has at one end a quadrangular niche of around 3 m wide by 1 m deep, and the entrance at the opposite with a rectangular nartex and a portico before two columns that supported three arches, now disappeared, being the central one the widest. The building's interior was later modified, reorganizing it longitudinally by adding two three arch rows, so each row was supported upon the walls and in two columns added for that purpose.

Some traces of the nave that existed originally upon the lower one that we have described are still preserved. This lets us assign Santa Eulalia de Bóveda to the group of double vault monuments that follow the syle started in Spain with the Mausoleo de la Alberca. It seems obvious that the upper nave had the same characteristics as the lower one, though probably with the entrance door at the opposite side. In the attached reconstruction we have had in mind that in the rear side the floor level was at the same height than the lower vault, since the building is buried on that side. We can also suppose that the upper nave was also covered by a barrel vault with a saddle roof. It is also possible that the referred door was a horseshoe one, as the three portico's arches, like the one that is still preserved.

In the main façade formed by stone blocks in horizontal courses are worth mentioning the rectangular windows with a triangular discharging arch that flanks the door. This one has a horseshoe arch in bricks radially placed, prolonging a fourth of the radius, what would have been less surprising for a Visigothic church of the 7th century. In some of the stones of the façade a sculptured decoration may be noticed, clearly of the same period of the original construction, that consists of dancing figures that have no relation with the Roman sculpture.

In the interior, the barrel vault, recoverd with stucco, keeps a decoration based on geometrical, vegetal and animal motifs, maintaining the Roman pictorial tradition recalling also in the drawings of birds, some of the sculptured decorations of later Visigothic churches, such as Quintanilla de las Viñas that had a great influence in the Asturian painting of the 9th century.

After all that has been described we can reach the conclusion that the building was originally a nymphaeum devoted to a pagan divinity between the 4th and 5th centuries, since the only opposition to this theory is the existence of the horseshoe arch, but we know that these had appeared earlier as a decoration of some Roman steles in the north of Spain. Accordingly, the sculptured decoration could have Celtic influences on account of the style and clothing of the dancers. This theory is based on the existence of the pool and of a complicated system of water conduction, that entered through the bottom niche allowing to keep the level of water in the pool, what makes us think on a religious kind of rite related with water. The ensemble is a singular example of Hispanic-Roman syncretism.

The later modification of the building is also worth mentioning. Its only nave was divided by adding two three arch rows, possibly horseshoe arches, that supported upon four columns with capitals, two of them attached to the portico's walls and the chevet. Each one of these ensembles supported a separation wall that, besides reinforcing the vault, artificially confer a basilical structure of three naves, possibly with the intention to make it look like more as a church. For that reason the pool was filled and remained undiscovered until 1947. The restauration date is unknown though, according to its characteristics it seems to belong to the Visigothic period, probably from the 6th or 7th centuries, influenced by the impulse of conversion to Catholicism generated by San Martín de Braga (+580).

>OTHER INFORMATION OF INTEREST