While Eads had promised the company to prove by careful experiment, so far as was possible, everything connected with the bridge that had not already been fully demonstrated in practice, he did not pretend that in his main outlines he was without some examples. It was in his development of known ideas and his expedients for simplification that his genius perhaps most strikingly showed itself. Again and again he contrived some device so simple that, like a great many strokes of genius, it seemed that anybody should have thought of it. The massive piers were sunk to the bed-rock by means of metal caissons. These were adapted in design from some he had seen in use in France, and had examined during a trip his doctors ordered him to make in 1868. Eads himself compared them to inverted pans. They were open at the bottom, but perfectly air-tight everywhere else. They had several important features which were entirely original. Such caissons, sunk to the bottom, have the masonry of the pier built on top of them even while they are sinking; and workmen inside them keep removing the sand from underneath, and throwing it under the mouths of pipes which suck it up to the surface of the river. Evidently the caissons must be filled with compressed air to equalize the external pressure, which is constantly increasing as ever deeper water is reached; they must also have an opening connecting with the surface; and to admit of passing from the ordinary atmosphere to the denser one, there must be an air-lock. Before this bridge was built, the air-lock had always been placed at the top of the entrance shaft, where, as the caisson sank and the shaft was lengthened, it had to be constantly moved up. Eads placed it in the air-chamber of the caisson itself, where it never had to be moved; and thus, as the shaft was not filled with compressed air, less was needed, and there was less danger of leaks. Another of his useful innovations was to build his shaft of wood, and another was to put a spiral stairway into it. Indeed, in the last pier he put an elevator into the shaft. Moreover, he was the first person to run his pipes for discharging the sand, not through the shaft, but through the masonry itself; and he invented a very simple and effectual new sand-pump, which was worked by natural forces without machinery. All these improvements and various others seem to have been thought of so easily, that we are inclined to wonder why clumsier methods had ever been in use. He described them all in his reports and his letters about the bridge in a style which is not only clear but actually fascinating even to a person who has scant scientific knowledge or taste.