Deux embouchures de trompettes métalliques ont été découvertes au XIXe s. par l’abbé Baudry parmi le mobilier enfoui dans les puits n° XVII et XX sur le site de Troussepoil en Vendée. Elles sont conservées aujourd’hui au Musée de L’Historial de la Vendée. L’analyse typologique des objets s’accompagne d’une mise en série avec des embouchures comparables découvertes dans les Gaules. L’article questionne également les problèmes liés au contexte, à la nature exacte des puits et à leur datation.
The Gallo-Roman site of Troussepoil, in the municipality of Le Bernard in Vendée, in the civitas of the Pictons, presents a major specificity related to the abundance of wells discovered (no fewer than 34 : fig. 4), making it a unicum in Gaul. Although wells have been identified on other sites in Gaul, they are never in comparable numbers, which is why the nature of this place is still a subject of discussion (secondary agglomeration or villa ?). Among the material buried in wells XVII and XX (figs 5 and 6), two separable metal trumpet mouthpieces were discovered during excavations led by Abbot Baudry in the C19th. No part of the trumpet itself (bell, pipe) was found among the material. Even today these are the only Roman trumpet mouthpieces known in the province of Aquitania. These mouthpieces, made of a copper alloy, comply with the known standards for the manufacture of objects of the type in the Roman Empire : they are composed of a cup linked to a conical tail. They were reported succinctly in Baudry’s monograph in 1873 with a rough sketch (fig. 2 ) , but have not been studied since 1873. Baudry for that matter does not provide any precise information and indulges in some rather whimsical assumptions about them. Through a fresh study of material found on the site, housed at the musée de L’Historial de la Vendée, and through growing interest in the study of Roman trumpets, conditions are now set to propose an initial presentation of these metal parts. With the re-examination of the Gallo-Roman tuba of Neuvy-en-Sullias in 2007, the publication of the PhD dissertation of C. G. Alexandrescu on trumpeters of the Roman army in 2010, which gives pride of place to archaeological material, and the recent study of the cornua of Pompeii in 2019, we now have an essential documentary basis for understanding the making of Roman trumpets. The two mouthpieces from Troussepoil were x-rayed and examined at the Arc’Antique laboratory in Nantes (cf. appendix ) to look into the technical aspects (manufacture, bore, taper, cup profile). Seriation suggests they are mouthpieces of cornua , the curved Roman trumpets used in the legions and in the amphitheatre for gladiator shows, because we are now able to tell apart mouthpieces that belonged to straight trumpets, the tubae (they are short, with a small cup and narrow throat) and mouthpieces of cornua (longer, with a wider, more flared cup and wider throat). The two artefacts from Troussepoil belong to a family of mouthpieces that are well known in the western Roman Empire, with which they share most of the essential features : dimensions and morphology. However, two particular features can be identified on each of these mouthpieces : the presence of a swelling on one of the pipes (inv. 880.1.143) while the other (inv. 880.1.144) has a ring on the tube, which might be a repair or fitting mark, identical to the one on a mouthpiece from Bavay (known only from a sketch). Examination confirms both the care taken with their manufacture and their kinship with modern mouthpieces. The difference lies, however, in that they did not fit inside the pipe, as with modern trumpets, but covered it over. The study draws up the list of mouthpieces catalogued to date in Gaul and Gallia Narbonensis , among which cornua mouthpieces are in the minority (table 1 ) . Precise dating is scarcely possible because little is known of the stratigraphy, approximately reconstructed in Baudry’s drawings, although it is known that for the mouthpiece from well XVII, the date of abandonment was probably in the second half of the C2nd, or even the early C3rd. As concerns well XX, ceramic fragments of Lezouox dated to the late C2nd have been unearthed there. The other controversial point concerns the purpose of these mouthpieces. The absence of militaria on the Troussepoil site does not argue in favour of the presence of trumpeter soldiers. Nor is there any sign of a theatre-amphitheatre that might explain the utility of a cornician in the area, like the one known in Bordeaux. The celebration of civic ceremonies is to be ruled out ; as for the funerary use of the cornu , this was not obviously the case here. There remains the hypothesis of a votary deposit in the well, because this practice is well documented in Gaul with the deposition of trumpet or mouthpiece fragments in sanctuaries often related to a local Mars (Neuvy-en-Sullias, Sceaux-du-Gâtinais, Saint-Just-sur-Dive). Even if one of the mouthpieces was discovered close to a statuette of a divinity, caution is required inasmuch as the function of the wells is still to be determined. Problems relating to the context and exact nature of the wells are in question. Although the funerary postulate advanced by Baudry has been questioned in recent years, no hypothesis is fully convincing as yet. It seems that these wells, while initially designed for drawing groundwater, were then filled with varied material (ceramics, glass, faunal remains, etc.) thereby serving as rubbish tips.
