The architect of Valentré Bridge was wiser than Perronet, every arch in his work being an elastic bow that moves between two piers without conveying its oscillation beyond these supports. To our modern eyes, no doubt, there are too many arches across the River Lot at Cahors, but this defect seemed necessary in the Middle Ages, and for two reasons. It was regarded as a defensive precaution, because narrow arches were easier to protect from the roadway when an enemy tried to assemble boats under a bridge; and since in the frequent wars of those days a bridge had often to be cut as a final resource against defeat, it was essential that the destruction of one arch should not upset another by the withdrawal of a counterbalancing thrust from one side of a pier. Many piers of a large size were looked upon as particularly needful when the greater lateral thrust of round arches had to be considered in its relation to a bridge cut in a single place. Also, as we have seen (p. 264), bridges in the Middle Ages were built very slowly, and as war at any moment might stop the masons, piers were regarded as abutments and made very strong. This much is known, but none can say why piers were built unreasonably large. Frequent inundations from obstructed rivers were as evidently harmful as weak piers that floods overthrew; and the genius that solved so many problems in church architecture ought to have shown in bridges a riper discretion. Often piers and arches were of the same width--a waste of labour and material, as well as of space in the waterway. Even the Romans, though their piers were less bulky, impeded the current of rivers with too much stone; and to save their work from the floods which they provoked, they built relief bays for spate water above that part of their piers where adequate resistance had been obtained against the lateral thrust of heavy arches.