M. de Fontenay suggests that this upper part was restored at some later and degenerate period of Roman architecture. Certainly the plain pilasters are badly designed and carelessly set, while the capitals, of a composite, semi-ionic order, appear to be too narrow for their pilasters—not too wide, as stated by M. de Fontenay, and also by M. Déchelette in his careful little guide to Autun. The Porte St. André shows no signs of having been fitted with a portcullis. Hamerton[30] states, no doubt correctly, that the door was barred by strong beams inserted into holes and grooves. These are still visible.
The Porte St. André, it should be noted, is one of the most complete Roman gates existing in France. The lower portion of one of the flanking towers, which rose originally several feet above the attic story of the gate, owes its escape from destruction to its shape, which, coinciding with that of a typical Romanesque chapel, tempted certain ecclesiastics of the middle ages, to dedicate it to St. André, as a place for Christian worship.[31] These flanking towers comprised three stories. The first communicated with the Chemin de Ronde, along the crest of the walls, the second was a vaulted chamber, and the third remained open to the sky. Access was obtained by a double staircase.
On the question of the period during which these gates were built, M. de Fontenay and Viollet le Duc are again at variance,—the latter attributing them to the fourth or fifth centuries,[32] the former to the reign of Vespasian (a.d. 69-79).[33] Strange as it may seem that another writer should contradict continually so eminent an authority as the last-named, the author of "Autun et ses Monuments" again has reason on his side, since he can refer to the orator Eumenes as describing the gates in the year 311 a.d.
Moreover, since the Roman walls were admittedly broken down in many places during the siege by Tetricus, in the year 269 a.d.,—an event of which Eumenes was a witness,—and the inhabitants were already beginning to retire within the safer precincts of the Citadel; what would be the reason for erecting elaborate gates on the line of the ruined wall? I am not an authority upon ancient architecture; but I was certainly astonished to read the date given in the "Dictionnaire Raisonné"; and I should be glad to know whether other experts support M. le Duc's theory.
Of the two other gates—the Porte St. Andoche, on the south-west, and the Porte de Rome on the south-east, in the direction of Lyons,—nothing remains except the rectangular portion of one of the lateral towers of the former. The loss of the Porte de Rome, or the "Porte des Marbres," as it was popularly called, is especially to be regretted, since that name alone suggests the truth, of which evidence exists, that it was by far the most beautiful of the four. Moreover, it has additional historic interest as being the gate by which Constantine entered Autun in 311.
The old writers agree in describing the Porte des Marbres as a thing of beauty,—a quality which was its undoing, as offering an irresistible attraction to the Mediæval and later builders. Several ancient Corinthian capitals, not otherwise easily accounted for, are to be seen to-day in the porch of the cathedral built at the close of the 12th century. It appears, indeed, that around the present site of the Fountain of the Pelican was a burial ground named Les Marbres, on account of its richness in borrowed sculpture. The site of the gate was known from the 14th century onwards as "à Marbres" or "de Marboribus." At the time of the construction of the bastion of the Jambe de Bois, the workmen unearthed many marbles, including columns, capitals, and bases of the Corinthian or some composite order.[34] The date of the final destruction of the Porte des Marbres is uncertain; but its flanking towers, then known as the Fors de Marboribus, were still standing in the middle of the 14th century.
Before beginning another chapter, let me speak one word of warning. If you ask one of the humbler inhabitants of Autun where the Porte St. Andoche stood, you will be directed, without hesitation,—as we were,—to the bank of the river, near the railway station. Having reached that spot and crossed the bridge, you will obtain a lovely view of the Roman wall and the river beside it, but will fail to find the remaining tower of the gate, for the reason that it is not there. The Porte St. Andoche stood at the foot of the Boulevard Schneider, opposite to the Couvent du Sacrement, on the road leading, by way of the enceinte, to the Tour des Ursulines.
The reader may ask why the peasants should misdirect him? They misdirect him, in this particular instance, because they believe that the gate really stood on the town side of the western bridge of the Arroux, in the corresponding position to that of the Porte d'Arroux. But my point is, that, had your informant been utterly ignorant of the supposed whereabouts of the gate, he would, very probably, have directed you with almost equal facility. The Burgundian peasant is very ignorant; but he is also very proud,—much too proud to admit that he does not know the whereabouts of a monument that a stranger has come, perhaps, a thousand miles to visit. His swift imagination, therefore, promptly creates the site; and he, or she, will tell you promptly, volubly, and with much circumstantial detail, exactly how to get there. In this snare we have been taken many times during our travels among the Burgundians. The Provençals have a different and preferable method. They do not invent a site for the monument; they deny its existence. Speaking with some experience of the peculiar ways of the French peasant in such matters, my advice to the gentle stranger is—not to trust him. Get the best guide and map that you can find in the local librairie; and rely on them, and on your own intelligence. You may then, when you are at fault, consult the passer by as to details, letting your judgment decide whether, in his particular case, he is to be trusted.