Night Watch at Point Reyes

At half past nine on the night of 4 April 1888, John C. Ryan was called to the lighthouse. The light was going out. Its lamp plunger was slipping.

Next day, the lamps were changed. The old lamp and burner were found very dirty. The fault lay in the small fuel system at the centre of a much larger optical machine.

Point Reyes stood on the coast north-west of San Francisco, warning ships approaching the busy port. Inside its lantern, a great assembly of glass turned around a stationary light. Twenty-four panels gathered that light into separate beams. One revolution took two minutes, so a beam swept past an observer at sea every five seconds. The lamp itself was supposed to burn steadily.

Ryan had taken charge in January. His first assessment was severe: the station was filthy and broken, more like an old sawmill than a lighthouse. Third assistant James Nightwire signed beneath it, agreeing. The following days filled with work on dwellings, burners and the steam siren. Ryan was inheriting a workplace with a fine lens and a long list of things to mend.

A tower partway down the cliff

The lighthouse had opened on 1 December 1870. Putting it on top of the headland would have made construction easier, but the height was a disadvantage in fog. George Davidson, an experienced surveyor of the Pacific coast, had argued for a site lower down, where lifting fog might leave the light visible.

In June 1869 Davidson and two fellow surveyors followed the ridge below the proposed upper location and found a projecting shelf. The site needed rock removed and offered little room. He also urged the provision of a powerful fog signal lower still. In thick weather, he warned, it would be more useful than the light.

The decision placed the tower hundreds of steps below the keepers' homes. Workers blasted and levelled its footing and built a tramway beside the stairs to carry materials down. Supplies arriving by sea had first to be landed in sheltered Drakes Bay and hauled over the headland. Once lowered to the site, the sections of the iron tower were bolted to a foundation anchored in the rock.

Joseph Bien, a San Francisco machinist, assembled the tower and installed the illuminating apparatus. During the preparations a case of glass prisms went missing in transit. The engineer and shipping company searched for it through July and August; it was found in time for the work to proceed.

The apparatus had been made in Paris in 1867. Its surviving clockwork case carries the name Barbier & Fenestre, the French firm which supplied lenses and other lighthouse equipment. Bien repaired damage sustained in shipping before the parts went on to Point Reyes. Seven men were hired to lower the tower sections and apparatus to the construction site in October.

By then Fresnel optics were established equipment, manufactured and exported by specialist firms. The name belonged to Augustin-Jean Fresnel, the French physicist whose first large revolving lighthouse apparatus entered service at Cordouan in 1823. Point Reyes's machine was a later development, made forty years after Fresnel's death.

Giving the light a direction

A lamp sends light upwards, downwards and sideways. Much of that is unhelpful to a ship looking towards the coast. The optical task was to collect a substantial share of it and send it out near the horizon.

A conventional large lens would require a heavy thickness of glass. Fresnel divided the useful curved surfaces into steps: a central lens surrounded by separate rings whose angles redirected the light. The rings could do their work without all the glass that would have filled the spaces between them in a solid lens.

Point Reyes's panels combined several kinds of optical surface. Around the central bull's-eye, rings bent rays by refraction as they entered and left the glass. Prisms farther above and below also used reflection inside the glass to turn more steeply travelling light outwards. The angles mattered: each element had to receive light from the lamp's position and return it in the intended direction.

The brass frame held these separate pieces in relation to one another. It was an assembly of shaped glass, with an accurately placed source at its centre. The bright colour seen in a modern photograph comes from light passing through the prisms; the historic navigational signal was white.

The description “first order” identified a standard size by its focal distance—the distance from the source to the lens. It was a large coastal apparatus, rather than a claim that this was Fresnel's first design or the first lens installed in America.

At sea, the intervals helped identify the station. A navigator could compare the light's colour and rhythm with a published light list or chart, then take a bearing on its known position. Seeing a bright light was useful; knowing which light it was made the observation navigational information.

Two different weights

Turning the glass steadily required another machine beneath it. The lens rode on wheels, while clockwork driven by a descending weight supplied the motion. A governor regulated its speed. A description published in 1880 gives the driving weight as 175 pounds and the winding interval as two hours and twenty minutes. Throughout the night someone had to raise it again.

The fuel supply had its own weight. By September 1887 Bien had returned as a Lighthouse Service machinist to replace the earlier lard-oil lamps with lamps burning mineral oil, or kerosene. A contemporary newspaper described the new apparatus with five circular wicks and a 120-pound plunger forcing oil from a reservoir towards the burner.

It was a fault in this part of the oil-feed system that brought Ryan to the tower the following April. The plunger belonged to the lamp, rather than the clock which turned the lens. The distinction explains why glass could remain perfectly in place while the signal weakened: there had to be a dependable light at its focus.

The new lamp also changed the routine. The older four-wick lard-oil system had required oil to be pumped from a lower chamber into an upper one every two hours. In the 1887 arrangement the reservoir was filled and the plunger raised during the day. The newspaper reported that the new lamp needed no trimming during the night. It still needed care. On 25 March 1888 the log noted trouble with a burner and short wicks; the burner was changed the next day. The April alarm followed little more than a week later.

Work outside the lantern

Ryan's assistants were not employed solely to watch glass turn. In February they overhauled the lens machinery, changed and cleaned burners, and cleaned the tower's clockwork. On 20 March the second assistant worked on the lens's truck wheels. The same log records repairs to gates, windows, water pipes and the machinery below the lighthouse.

Fog could remove the light from a ship's view altogether. The station therefore needed its steam-powered sound signal, with boilers to feed, water to collect and parts that repeatedly failed. A lighthouse on an exposed cliff still depended on rainfall and fuel brought to it. In March the men put down 226 sacks of coal.

The log is full of work done twice. A siren was overhauled and put right, then failed again. New pins, gaskets and governors followed. On 14 June Ryan recorded that an assistant was complaining about the amount of work. A few days later, the men were whitewashing the coal chute and steps on an inspector's orders.

Such entries also show the limits of the keeper's account. Ryan named the jobs and recorded the complaint; he did not preserve the assistant's explanation. The station's repairs were real, but the log was also the superior's record of his staff.

The historic lens continued in service through changes of lamp and power until the light was automated in 1975. Its glass and mechanism remained at Point Reyes. Caring for the preserved apparatus still involves reaching individual prisms and keeping their surfaces clean.

In Ryan's year, a day's list might contain both a lens cleaning and the arrival of a cow. Other entries record hurt hands, journeys for mail, substitute keepers and visits by a machinist. On 29 November the log allowed a short exception: Thanksgiving, no work. The next day brought whitewashing around the dwelling and cleaning in the tower.