The piles were sunk into the soft sea-bed by means of water-jets, which, playing about the foot of the pile, burrowed a hole into which it could move downwards. A depth of 15 feet had been considered necessary to secure the desired rigidity, and as a rule the pile could be driven to this depth in about four hours. When the pile-driving commenced, however, it was found that the sandbank had undergone a marked change since the surveys were made. Erosion had been very active owing to the currents having been checked by the obstructions which the legs of the staging offered. Under these circumstances a novel experiment was made upon the site. One of the piles was lengthened by 14½ feet, to be driven to its limits, just to ascertain how far it would go into the sand. This in itself was a somewhat daring undertaking, seeing that the tiny colony on the staging did not possess the facilities which were available on shore for the work. However, it was accomplished satisfactorily, and when the pile was sunk it was found to descend another 13½ feet, where it touched hard rock. This discovery brought about a modification in the plans. As a solid foundation could be gained at a depth of 28½ feet, and as the piles could be lengthened successfully upon the site, it was decided to extend all the piles to a complete length of 64½ feet, and to drive them down to the hard bottom. When the piles were all lowered, they were subjected to four blows from a “monkey” weighing 2½ tons, dropped from a height of 4 feet. But these four final blows only drove the piles from ¼ to 7/8 inch farther into the sea-bed, whereas, according to the specification, a margin of 1 inch was allowed for this test.