The excavation of the subsoil and the analysis of the standing structures resulted, in the first place, in a better understanding of the extent of the successive alterations and renovations that had been made in the building since 1800 with the aim of turning it into a "modern" bathhouse. It had always been used as a bath, even after the hamman-style heating system had become inoperative in the 17 th century. The research also revealed the scope of the restoration work which took place in 1953 and 1961-63. These interventions had caused the elimination of various structures and most of the strata dating from before the 19 th century, making a global understanding of the building's precise historical phases difficult.
The structural layout of the building reproduces faithfully one of the variations of the hamman design. As we have seen, it consists of three adjacent, rectangular, rooms, parallel to one another from east to west, and perpendicular to the main axis of the vestibule, which is located on the northern side. The latrines are next to the cold room, on the far eastern side, and the boiler room is next to the hot room, on the western side.
Because of the need to maintain high temperatures, the wet rooms were solidly constructed and they constitute the only part of the original building that is still standing. Here the walls are 0.8 m thick. They were built using the cob method, with a very solid lime mortar reinforced by a high content of gravel and by an irregular filling of pebbles and brick shards. The vaults, the arches and the jambs between the rooms were made by alternating courses of brick and mortar slurry.
The foundations were built at different depths, using different approaches. The foundation for the wall separating the cold room from the vestibule was made 1.6 m deep. In order to build it, a considerable strip of earth was dug out, which was then filled as the building got higher, so as to make the workers' job easier. This procedure generated successive work surfaces that were easily recognised during the excavation process. The other foundation ditches were more conventional. They were trenches, built parallel to the walls, around 0.4 m wide on each side, reaching depths of only around 0.7 m.
The cold and hot rooms, as well as the two alcoves in the warm room, were covered with half-barrel vaults, whereas the central third of the warm room was topped with a octagonal dome supported by "elephant trunk" arches. Skylights shaped like eight-pointed stars were bored through the thick ceiling in a funnel-shape and arranged symmetrically in both the dome and the vaults. The latrines, however, have a cross-vaulted roof.
The examination of the southern walls of the steam-rooms led to the discovery of the plumbing system that distributed the water from the boiler room. The pipe was made of assembled ceramic pieces, sloping from east to west, which allowed the water to circulate from that room to the cold water basin, located in the south-west corner of the cold room. There was also a brick archway forming an opening in the eastern wall of the hot room, which allowed the users direct access to the water from the boiler. A small segment of this arch still remains. In the same wall and to the north of the boiler arch, another brick arch was discovered, which may have been part of the basin containing hot water.
In the subsoil on the edges of the hot room, brick pillars, 90 cm high and approximately 62 cm wide, were found. These belonged to the hypocaust chamber, where the air that came from the boiler circulated and heated the floor, before rising up the two pairs of chimneys built into the walls. This chamber was covered by brick archways that were supported in the central area by a series of pillars, built by twos, and by ledges in the walls on the ends.
The spilt water followed the slope of the floor, flowing through the rooms up to the latrines, where an underground drainage canal began. During the excavation of the warm and cold room floors a thin layer of mortar was found, used as a primer for the original pavement, whose characteristics and finish are still unknown. This layer covered the levelling underlay, made of stamped earth.
The openings that connected the central room—the warm one—with the lateral ones—the cold and hot—were enlarged, both in height and width, at later periods, although we know that the original width must have been between 0.8 and 0.9 m.
The excavation also confirmed the progressive deterioration suffered by the most delicate parts of the building, caused mainly by the extreme conditions of heat and humidity. This situation must have resulted in the need for a series of minor renovations during the 15 th century, which did not affect the building's general structure.
The work done during the fourteen-hundreds mainly consisted in the reinforcement of the arches, and the placement of new flooring slightly above the level of the first one. This pavement was made of bricks and had some medallions with a suboctagonal outline, formed by a chequered-pattern of diamond-shaped ceramic pieces, some glazed in green and others in white, reminiscent of some of the ornamental fixtures on the Mudejar towers of Teruel. Some of these pieces were found during the restoration work in the 1950s, which led to the belief that all of the original pavement was made in this way. For this reason, costly reproductions were made to repave the floors in 1963.
The pieces of ceramic retrieved from the filling in the foundation trenches and in the priming layers covered by the original pavement clearly date from the beginning of the 14 th century, which coincides with the documented evidence on the building's construction. This is also indicated by the collection of fragments from various vessels: large containers, some with etched decorations; jugs with manganese decorations; traditional grey pots from Catalonia; monochromatic glazed forms in green and amber (bowls, pots, pitchers, plates and platters); and finally, some green and purple ceramic fragments in the classic Paterna style.
Concerning the dating, it is particularly interesting to take into account the collection of shards found inside the well that was filled in order to build the northern wall of the vestibule over it, that is, the wall separating the bath from the oven from the time when Pere de Vila-rasa had them both built. The sediment inside the well contained nearly complete pieces of a very homogeneous stylistic repertoire, suggesting a synchronic filling. Furthermore, the materials were very similar to those described above.
The building also fits the typology of the period in question. We have already pointed out Azuar's conclusion that the Baño del Almirante corresponded to the "Late Granada" design, or to what Gómez-Moreno called "Advanced", in which there is little difference between the width of the warm room and that of the cold and hot rooms. This design was developed in the 14 th century, and more simplified examples can be found in rural baths.
The architectonic features of the vestibule are also interesting. The excavation revealed the original floor plan, which concurs with Laborde's drawings, allowing us to complete our information about the façade, destroyed in 1874. Basically, it was a construction with a rectangular floor plan, defined by a peristyle, or central portico, supported by eight columns which sustained three arches on the long sides, and one arch on the short sides. The engravings show that the porticoed room was surmounted by a gallery of windows on the western half, which corresponds to a renovation made in the 17 th century. However, it seems probable that the gallery originally occupied the entire upper perimeter of the peristyle. In a study on rooms lit by a central lantern or a skylight in the architecture of Granada and Morocco, Torres Balbás presented a series of examples where this solution was used for bath vestibules: the Room of Beds in the Comares Palace of the Alhambra; the Dâr al-cArûsa Palace bath in the Generalife; the Zagora bath; and the al-Mujfîya hamman in Fez, studied by Terrasse. All of them date from the 14 th century, allowing us to conclude that this kind of structure in vestibules "is not found, judging from known examples, in any bath prior to the 14 th century."
It is also interesting to point out the style of the columns in the warm room, which are characterised by inverted truncated-cone capitals, without any decoration, separated from the narrow shaft by a simple gorgerin and finished off by a plain abacus. To put it simply, they were capitals from the Romanesque tradition with ample parallels that generically refer to the last third of the 13 th century and the beginning of the 14 th . Among this style of columns, it is worth noting those in the annexed chapel of the church Sant Joan de l'Hospital, because of its proximity to the Baño del Almirante.
Finally, we should note the main changes produced in the building after the bath's original hamman-style heating system became inoperative. As we have said, at the end of the 17 th century, or perhaps shortly after, the hypocaust was no longer used and was filled with rubble. However, the establishment continued operating as a public bath, probably using tubs for immersion baths. The building was repaved again and, since the inconveniences of humidity and steam had been eliminated, the walls were decorated, very much in the Baroque style, with red geometric motifs on a yellow or ochre plaster background.
At the beginning of the 19 th century, coinciding with Laborde's visit, the building underwent considerable renovations, which were partially depicted in the famous engravings. Individual bath cubicles were built by dividing the bath's old steam rooms with partition walls. A marble bathtub was installed inside each compartment and a drainage system was built, which partly took advantage of the already existing pipes.
Nevertheless, it was between 1830 and 1874 when the establishment endured its most drastic alterations. On the one hand, the street line was regularised, forcing a reduction of the boiler room and a forward displacement of the entrance façade, which modified the doorways to both rooms. On the other hand, a three-storey building was constructed over the boiler room and the eastern half of the hot room and vestibule.
The work also included the demolition of the four peristyle columns located on the eastern side, under the new residential building. In addition, in the hot room, an underground space was made to house the structural elements to support the weight of the new building. In this same room, a well was dug to supply water for the new tenants. In 1874, the owner of the bath acquired the adjoining oven, located in what is now the assembly hall for the Ministry of Economy, Treasury, and Labour. Once the oven was demolished, the proprietor took advantage of the plot in order to enlarge the vestibule. This renovation completed the elimination of the medieval peristyle. In its place, a cloister patio was built, surrounded by a portico made of columns with six more bath cubicles. The floor was paved again with square tiles of white marble, raising the floor 0.3 m over its former height
At the beginning of the 20 th century, the establishment was decorated with Neo-Arab details, such as false horseshoe and polylobate arches, skirting boards made of tiles in the style of the Nazar dynasty of Granada, and Arabesque-looking plasterwork. The restoration process that took place from 1953 to1963, together with the closing of the baths in 1959, completely concealed and destroyed these ornamental additions of the modern era, which had survived the post-war period. The only exceptions were the entrance façade and the skirting-board tiles, which were recently eliminated. The most striking of all was the destruction, by the use of a hammer, of the pink bathtubs made of solid marble.