Building a lighthouse on a rock

AThe lighthouses that are remembered are not the ones on headlands. They are the ones built on rocks that are covered by every high tide, which could be worked on for a few hours at a time in summer and not at all for most of the year. The engineering problem was not height. It was that the structure had to resist a force which arrives sideways and which was, for most of the period, not measurable.

BEarly attempts failed in a way that taught the discipline its first rule. A tower built of timber and weighted down with iron was lost entirely in a storm, with its builder inside it. A later one of stone and wood burned. What both had in common was that they resisted the sea by mass alone, and the conclusion drawn was that mass was not enough: the shape had to turn the force aside rather than stand against it.

CThe shape that worked was borrowed from a tree. A trunk is wide at the base, narrows rapidly and then tapers slowly, and that curve puts material where the bending force is greatest. A tower of that profile presents a surface which rising water runs up rather than striking flat, and the wave loses much of its energy climbing. The same outline has been used ever since and it was arrived at by looking at something that had already solved the problem.

DThe second innovation was in how the stones were joined. Blocks were cut so that each one locked into its neighbours on every side, with projections on one face fitting hollows in the next, so that the courses could not slide and no single stone could be drawn out. A tower assembled this way behaves as one piece of rock rather than as a pile, and the shaping was done on shore, since a block that did not fit had to be discovered before it was carried out to the reef.

EThe mortar was the third problem and the least expected. Ordinary lime mortar does not set under water and dissolves in it, so the early towers depended on their joints being dry, which on a rock covered twice a day they were not. The answer was a cement made by burning limestone containing clay, which sets through a chemical reaction with water instead of by drying. This material was developed for lighthouses and went on to be used for almost everything else built since.

FWorking conditions determined the schedule more than any engineering decision. Men were landed by boat when the sea allowed, worked until the tide returned, and were taken off again, so a season might yield a few hundred hours. A tower that would take months on land took years, and the greater part of the work was waiting. Several builders lived on the rock in a temporary barracks on legs to save the landing time, which was dangerous and was judged worth it.

GWhat is striking in retrospect is how much was settled by the first successful towers and how little has changed. The profile, the interlocking blocks and the hydraulic cement were all in place before any of the forces involved could be calculated, and later analysis largely confirmed choices that had been made by judgement. The lesson usually drawn is about the ingenuity of the builders. The more useful one is that a structure which has survived a century of storms is itself a measurement, and a better one than was available at the time.