Above is a floor plan of Bath C. This bath stood on the Island in relative proximity to the Imperial Palace. It was fully excavated by the Princeton team and represented the most complex and substantial building found in the project. The building was around 40 metres wide and sixty metres deep from the street facade. Poccardi posits that the blocks on the Island (by utilising the "temple" as a form of measuring stick) were around 107m x 35.5m. Thus the complex of the baths/palestra were a big deviation to the usual street plan. Uggeri speculates that the stadium/palestra might have been part of the Seleucid Regia (palace). As no excavations took place on the southern side of the exercise field, this remains a possibility meriting further investigation. Bath C is certainly luxurious enough to have been part of a complex for the powers that be.
While the floor plan was clear, it was however stripped down to its foundations by subsequent looters of building material leaving little idea of what its vertical dimensions might have been.
Christine Kondoleon's excellent catalogue for the exhibition Antioch: the Lost Ancient City included several computer-generated reconstructions of these baths produced by the Boston architect and graphic artist, James Stanton-Abbott. This work, and the images, were commissioned by the Worcester Museum of Art to accompany its exhibition than ran in 2000-2001. The major essay on Antioch's baths in the catalogue is by Fikrit Yegul of University of California Santa Barbara.
The first reconstruction below (click to enlarge) reproduces and enlivens the floorplan shown above. None of the other baths found in Antioch had such a strictly symmetrical layout as these baths. This is not to say that there were not other baths in this format. Not enough has been excavated of Bath F to preclude that it did not originally have a very formalised layout. This version also adds some more detailed mosaic features. The hot rooms (caldaria) are to the left in this plan. Notice that the mosaics shown here differ from those in the floor plan. above. This was done by "reusing" images of other Antioch mosaics to give an impression of how the floors really looked rather than just using pure geometrical designs as a filler.
The second image of Bath C from the catalogue is an axonometric view of the bath complex with the wall height projected up to the level of the column capitals. this brings the building alive and gives an idea of how it might have appeared to one of the bath's clientele as they progressed through the rooms.
The ruins complex was so scoured of remains by looters that no indication remains of how the roof and ceilings may have appeared (for instance whether there may have been an open space over the octagonal pool in what was probably the frigidarium). The octagonal pool measured 9.5 m x 9.5 m x 1 m . This pool had a significantly larger capacity (c. 70.8 m3) than the pools in the cistern-fed private bathhouses of earlier periods.
Though none are shown in the octagonal room (as no evidence survived to justify them), one might consider some columns may have existed there and presumably both the frigidarium and main caldarium had high rooflines than some of the smaller rooms in the complex. The main baths in Rome, while on a much vaster scale had vaulted and coffered ceilings in these types of rooms.
The other key source on this subject is the report in Volume I of Antioch-on-the-Orontes by Elderkin and Stillwell. This relates the discovery and excavation of the baths. Unfortunately, in my opinion, they did not advance much in putting the baths (or much else they excavated) into an urban context by advancing onto the street in front of the baths and investigating the adjoining buildings. Was the baths structure on one of the four radiating avenues that Libanius speaks of?