In 2005 the University of Trieste began excavations in an area beside the via Gemina corresponding with one of the ancient city’s largest residential insulae, close to the forum and the river port. The entire area was investigated in the 1930s by Brusin; other sondages with oblique trenches closer to the area in question were excavated by Bertacchi. However, the zone next to the modern road, adjacent to the point where the so-called “Flowered carpet” was found, had never been excavated using modern methods. The investigation identified an important building complex, probably belonging to an imperial functionary or local person of high status, which is situated in one of the most interesting residential complexes of the late antique period in northern Italy. Although the complete plan of the house is not yet known, it may be suggested that it had more than one entrance. The main one (not yet identified) probably opened onto the via Gemina, indicated by some as the town’s decumanus maximus, situated south of the modern road, while a secondary entrance, with a paved courtyard and well, was partially excavated on the east side of the house, opening onto the cardo bordering the insula to the east. What is clear from the house plan is the creation of groups of rooms around at least three open courtyards linked by corridors, but in some ways independent and defined by different functions. The house owes its name to a polychrome mosaic floor, with eroti inside floral garlands, which decorated one of the private rooms. Worthy of note is the fact that it was possible to date the creation of one of the mosaic floors very precisely to the years after 337-340 A.D. This terminus post quem together with the evaluation of the levels and stratigraphic relationships of the floor surfaces made it possible to identify three construction phases. The original construction of the house dates to the central decades of the 4th century A.D., while the main phase of restructuring dates to the 370s A.D. The last phase did not significantly change the house plan, although some changes were made to the overall decorative scheme. In fact, the most noticeable intervention was the remaking of several mosaic floors at a level more than ten centimetres higher than the original floor. The third phase (5th century A.D.?) seems to correspond with the restoration of some floors or the creation of new ones, in addition to the building of walls to divide the large rooms into smaller ones. It is very interesting to note how the decorative scheme of the first phase domus recalls models and motifs of clearly classical derivation, while in the second phase the choice of decoration seems to favour the formal language of the period.