The Romanesque traditions, however, prevented the Christians from adopting it in Europe till forced to do it from constructive necessities; and the mode of its introduction into the early churches in Provence renders them singularly important in enabling us to arrive at a correct solution of this much mooted question.[20] It is hardly worth while discussing whether the form was borrowed from the East, where it had been used so long before it was known—or at least before we are aware of its being known—in Europe. It may be that the Pelasgic Greeks left examples of it in Provence, or that persons trading to the Levant from Marseilles became familiar with its uses; or it may be, though very unlikely, that it was really re-invented for the purposes to which it was applied. In whatever way it was introduced, it at least seems certain that all the churches of Provence, from the age of Charlemagne to that of St. Louis, were vaulted, and have their vaults constructed on the principle of the pointed arch. It has nevertheless long been a received dogma with the antiquaries of France, as well as with those of England, that the pointed arch was first introduced in the 12th century—the first example being assumed to be the work of Abbot Suger at St. Denis (1144-52), the result of which is that all who have written on the subject of Provençal architecture have felt themselves forced to ascribe the age of the churches in question, or at least of their roofs, a date subsequent to this period. The use to which the Provençal architects applied the pointed arch will be evident from the annexed diagram, the left-hand portion of which is a section of the roof of one of the churches at Vaison. The object evidently was to lay the roof or roofing-tiles directly on the vault, as the Romans had done on their domes, and also, so far as we know, on those of their thermæ. Had they used a circular vault for this purpose, it is evident, from the right-hand side of the diagram, that to obtain a straight-lined roof externally, and the necessary watershed, it would have been requisite to load the centre of the vault to a most dangerous extent, as at A; whereas with the pointed arch it only required the small amount of filling up shown at B, and even that might have been avoided by a little contrivance if thought necessary. By adopting the pointed form the weights are so distributed as to ensure stability and to render the vault self-supporting. It has already been observed that the Gothic architects everywhere treated their vaults as mere false ceilings, covering them with a roof of wood—an expedient highly objectionable in itself, and the cause of the destruction, by fire or from neglect, of almost all the churches we now find in ruins all over Europe; whereas, had they adhered either to the Roman or Romance style of roofing, the constant upholding hand of man would not have been required to protect their buildings from decay.