There is no spring on the plateau of Puy-d’Issolu; but several issue from the sides of the mountain, one of which, that of Loulié, is sufficiently abundant to provide for the necessities of a numerous population. This was the spring which the Romans succeeded in turning off. At the time of the siege, it issued from the side of the mountain at S (see Plate 31), at twenty-five mètres from the wall of the oppidum, and at a distance of about 300 mètres from the Tourmente. These 300 mètres make about 200 Roman paces. We see, therefore, that in the Latin text the word pedum must be replaced by passuum. We also see that the word circuitus (VIII. 4) must be taken in the sense of the course of the river.
The “Commentaries” say (VIII. 33) that Rebilus established three camps in very elevated positions. Their sites are indicated by the nature of the localities: the first, A, was on the heights of Montbuisson; the second, B, on those of the Château de Termes; the third, C, opposite the defile of Roujon, on the Pech-Demont. It appears from the excavations that the Romans had not retrenched the two first, which is easily explained, for the heights to the west of the Puy-d’Issolu are impregnable. Moreover, the Romans at Uxellodunum were not in the same situation as at Alesia. There they had before them 80,000 men, and in their rear a very numerous army of succour; here, on the contrary, it was only a question of reducing a few thousand men. The camp C required to be protected, because it was possible for troops to descend from the tableau of the Puy-d’Issolu towards the defile of Roujon, which, being situated fifty mètres lower down, gives an easy access to the heights of Pech-Demont. The excavations have, in fact, brought to light a double line of parallel fosses, which barred the defile behind, and formed at the same time a countervallation.
The Gauls could only quit the town by this defile, and by the western slope of the mountain. It became, consequently, interesting to ascertain if the Romans made a countervallation along the Tourmente, on the slopes of the heights of the castle of Termes and of Montbuisson. Unfortunately, the railway from Perigueux to Capdenac, which passes over the very site where the countervallation might have been made, has destroyed all traces of the Roman works: the excavations made above this line have produced no result.
The most interesting discovery was that of the subterranean gallery.[554] Until the moment when the excavations were commenced, a part of the rain-water absorbed by the plateau of the Puy-d’Issolu issued near the village of Loulié by two springs, A and A´. (See Plate 32.) The spring A´ flows from a ravine, and corresponds to the thalweg of the slope; as to the source A, it is easily seen by the appearance of the ground that it has been turned from its natural course. The excavations, in fact, have proved that it is produced by the waters which run in the Roman gallery. This gallery has been opened over an extent of forty mètres. It was dug in a solid mass of tufa nearly ten mètres thick, which had been formed in the centuries anterior to Cæsar. Its form is that of a semicircular vault, supported by two perpendicular sides; its average dimensions are 1·80m. in height, by a width of 1·50 mètres. The mud carried along by the waters, and accumulated since the time of the siege of Uxellodunum, had almost filled the gallery, leaving only at the top of the intrados an empty space in the form of a segment of a circle, with 0·50m. chord by 0·15m. for the absciss of its curve. Through this empty space the water ran at the time of the excavations.
Before reaching the tufa, the first subterranean works of the Romans had been made in the pure soil, which had to be propped: fragments of blindage have been found, some of them fixed in silicious mud, corroded or reduced to the state of ligneous paste, others petrified by their long contact with the waters charged with calcareous sediments. A considerable number of these petrified blocks, and remains of wood, collected in the interior of the gallery, are deposited in the museum of Saint-Germain.
The gallery does not lead directly to the spring which existed in the time of the Gauls. The Roman miners, after having made their way straightforward for a length of six mètres, came upon a thick bed of blue marl of the lias: they turned to their left to avoid digging through it, and advanced four mètres farther, following the marl, which they left to the right. When they reached the end of the marls, a horizontal layer of hard rock, one mètre thick, obliged them to bring the gallery back to its former direction, and to raise it, in order to avoid this new obstacle without going out of the tufas, which, formed by the waters, would necessarily lead towards the spring. (See Plate 32.) From this second turning, the gallery continued close to the line of separation of the marls and the tufas. It rose rapidly, until it reached the limit of the deposits of tufa. At this point blindage had again been necessary. It was there chiefly that the blocks of petrification presented a peculiar character: some lay thrown down in the gallery, pierced by sockets with a rectangular section, which show the dimensions and the way in which it had been worked; others, with a rounded base, are veritable uprights, still standing on the rock.
Independently of the excavations made to find the Roman fosses and the subterranean gallery, others have been made on the slope of Loulié, in the soil which is near the spring. They have brought to light numerous fragments of Gaulish pottery and of amphoræ, and, which is a further confirmation of the identity of the Puy d’Issolu with Uxellodunum, remains of arms similar in all respects to those found in the fosses of Alesia.[555] Under the earthfalls which during nineteen centuries have taken place on the slope of Loulié, all the traces of the fire described in the “Commentaries” have been also found. The site of the terrace and the covered galleries which were fired were also traced on the ground. Plate 32 represents the slope which was the scene of the struggle: the terrace, the tower, and the covered galleries are represented on it, as well as the subterranean gallery, according to a very exact survey made on the spot.
Complete Submission of Gaul.
VIII. Whilst these events were taking place on the banks of the Dordogne, Labienus, in a cavalry engagement, had gained a decisive advantage over a part of the Treviri and Germans, had taken prisoner their chief, and thus subjected that people, who were always ready to support any insurrection against the Romans. The Æduan Surus fell also into his hands: he was a chief distinguished for his courage and his birth, and the only one of that nation who had not yet laid down his arms.
From that moment Cæsar considered Gaul to be completely pacified; he resolved, however, to go himself to Aquitaine, which he had not yet visited, and which Publius Crassus had partly conquered. Arriving there at the head of two legions, he obtained the complete submission of that country without difficulty: all the tribes sent him hostages. He proceeded next to Narbonne with a detachment of cavalry, and charged his lieutenants to put the army into winter quarters. Four legions, under the orders of Mark Antony, Caius Trebonius, Publius Vatinius, and Q. Tullius, were quartered in Belgium; two among the Ædui, and two among the Turones, on the frontier of the Carnutes, to hold in check all the countries bordering on the ocean. These two last legions took up their winter quarters on the territory of the Lemovices, not far from the Arverni, so that no part of Gaul should be without troops. Cæsar remained but a short time in the province, presiding hastily over the assemblies, determining cases of public contestation, and rewarding those who had served him well. He had had occasion, more than any one, to know their sentiments individually, because, during the general revolt of Gaul, the fidelity and succour of the province had aided him in triumphing over it. When these affairs were settled, he returned to his legions in Belgium, and took up his winter quarters at Nemetocenna (Arras).