The Keepers' Index

The Tower and the Sea

To write about lighthouses is, inevitably, to write about failure. Every tower in the index exists because some earlier arrangement — a beacon fire on a headland, a buoy chained to a reef, a lantern in a church tower — failed to keep the ships off. The light at the Lizard had to be built twice over before anyone got it right. The first Eddystone was blown away; the second burned; the third was taken apart stone by stone; the fourth is the one we have. To catalogue these towers is to catalogue the sea's opinions about architecture, and those opinions are rarely kind.

Engineering at the edge

What the Stevensons and Douglasses really invented was a way of working on a rock that the sea would not let you stand on. The method was identical from Smeaton's Eddystone to Fastnet: cut a few feet of level floor at low water, bolt a temporary barrack to it, and then — with a crew living offshore in a tender and rowing in at the turn of the tide — cut and fit each granite course with such precision that it interlocked with the one below, dovetail and joggle, so that the tower became a single stone, a mass the sea could pull on but never prise apart.

The engineering was genuinely new. Smeaton's Narrative of 1791, the book that made this kind of work imaginable, was the first account of a marine structure designed by calculation rather than tradition, and every lighthouse built after him is in some way his descendant. The Stevensons made the method their family business across four generations; they were to the sea-washed tower what the Stradivari were to the violin.

On the keepers

The automation of the last manned British lighthouse, the North Foreland, happened in 1998, which means the keeping of a light is now historical in the way that the shoeing of horses is historical. For most of its life the job was unglamorous to the point of asceticism: three men in a stone room, winding a clockwork weight, trimming a wick, polishing a lens, reading the weather, watching for ships, not speaking much. Regulations forbade gambling, fighting, and — in some years — women on the rock. Wives and children were a shore-side proposition.

What remains in the archives are logbooks. They are, as documents, almost unbearably dull — a daily record of wind, weather, oil burned, ships passed. And then, without warning, a page will tear open: a wreck sighted, a boat put out, a body washed in, a keeper's mind broken by the long watch. The logbook of the Flannan Isles keepers in December 1900 stops, mid-entry, and does not resume. The men are still missing.

On optics

Before Augustin Fresnel, a lighthouse burned whatever would burn — coal, wood, tallow candles, an argand lamp if you were lucky — and threw the light where it would go. A mirror behind the flame caught some of what leaked backward and sent it out, dimmer, in the general direction of the sea. The efficiency was roughly one part in ten. Fresnel's lens, announced in 1822, changed this to four parts in five. It was a stepped bulls-eye of prismatic glass that bent the stray horizontal rays forward, so that a single flame became a blade of light on the horizon, visible twenty miles out and sometimes more.

The character of the light — fixed, flashing, group-flashing, occulting — was a matter of rotating the optic. A single-panel lens turned once a minute produced one flash. A four-panel lens twice a minute produced a group of four. Mariners learned the signatures the way one learns the voices of neighbours, and could name a light by how it breathed in the dark. The signatures in this index are those used today, most of them unchanged from the day the tower was first commissioned.