Forum Sempronii sorgeva su un vasto terrazzo del Metauro a una distanza variabile da 200 a 700 m ca. dall’attuale alveo. Sui latimeridionale e occidentale le mura correvano sul ciglio delle scarpate del terrazzo stesso, mentre sul lato settentrionale correvano allabase del versante collinare; più problematica la posizione sul lato orientale, non corrispondente ad alcun elemento dell’attuale fisiografia.Alcuni caratteri geomorfologici e depositi recenti mettono in evidenza modificazioni minori, ma importanti, e permettono un tentativodi ricostruzione dell’assetto fisiografico del locale paesaggio in epoca romana. In particolare, due erano gli elementi principali checaratterizzavano il paesaggio in questo periodo: 1. Il fiume verosimilmente scorreva sul lato sud a ridosso della cinta muraria, come ètestimoniato da un vasto meandro abbandonato, con forme freschissime e depositi contenenti frammenti di terrecotte elaborati dall’azionefluviale all’ingresso del meandro. 2. il lato orientale era con tutta probabilità bordato da un fosso oggi quasi interamente colmatoe del quale restano tuttora la sola porzione iniziale, sul versante, e terminale, sulla scarpata del paleomeandro. Tale configurazionepermetterebbe di superare alcune apparenti incongruenze fra l’attuale paesaggio e le primarie necessità sia difensive che di approvvigionamentoidrico e di viabilità dell’insediamento romano.
The Roman town of Forum Sempronii, founded between 132 and 126 b.C., is set near the modern village of San Martino del Piano, ca.2 km to the east of Fossombrone. The ancient settlement rests on the flat surface of a wide upper-Pleistocene terrace of the MetauroRiver, ca. 20 m above the modern stream. At least three sides of the Roman town walls took advantage of geomorphologic elements.Indeed, they were against the foothill to the north; to the west they ran along the edge of a steep scarp on the left of Fosso dellaConserva (a small left-tributary stream of the Metauro River), and they rose southwards on the edge of the Pleistocene terrace. Theeastern side is more problematic as regards both its true position and its relation to some specific landform: in all probability it developedalong the bank of a no longer recognisable gully. One of the most important features of the modern landscape in the vicinity of thearchaeologic site is a roughly flat-bottomed, broad semicircular depression representing a pronounced relict meander of the MetauroRiver. It is entrenched into the upper Pleistocene terrace unit and bounded to the north by the terrace scarp on which the southernwall of the Roman town was built. Hence, the scarp enclosing the abandoned meander separates the upper-Pleistocene terrace alluviumfrom the Holocene one: it is well-preserved, with a sharp upper edge and scarce colluvium/debris accumulation at the toe.Furthermore, the abandoned Metauro River channel is still well recognisable along almost all the scarp toe. The excellent preservationof such landforms hints at quite a recent cut-off of the meander neck and abandonment of the previous course. Moreover, the veryoccurrence of rounded Roman bricks and tiles in alluvial deposits at the mouth of the abandoned meander hints at a recent fluvial activityon at least a sector of the relict fluvial bend. Finally, maps dating back to 16th –18th century show that the meander at issue wasalready inactive and only minor adjustments of the fluvial course did occur at its mouth. On the basis of these considerations, andsince important modifications of fluvial-channel patterns in nearby areas can be ascribed to late Roman-early Medieval times, we canspeculate that the “roman” Metauro River flowed through the abandoned meander of today, close to the southern walls of the town.The presence of a large fluvial channel just close to the town walls would have enhanced their defensive effectiveness. Indeed, its vicinityto the river would have made the southern side of the town the more defensible one as well as a preferential way for watering.Moreover, the presence of an active channel at the foot of the scarp would have been able both to improve gully development and tosustain existing gullies free from debris-filling, thus allowing ditches similar to the one which has been hypothesized to border theeastern wall of the town to be sustained. On the contrary, the meander cut-off and the abandonment of the previous river coursewould have been able both to enhance the filling-up of such gullies and to accelerate their extinction.