Lights aboard ship served two different purposes. Some illuminated the water around the vessel so that the crew could see what they were doing. For this work, the light needed to be close to the area being lit, directed outwards and downwards, and bright enough at short range.

Others were intended to be seen from another vessel. They communicated the ship's presence, heading or status rather than illuminating its surroundings. Height and an unobstructed view were therefore more important than lighting the water nearby.

Both uses appear in ancient written sources and archaeological evidence. The written accounts associate signalling lights particularly with sailing in convoy and in fog, when vessels needed to make their presence known to others nearby.

In the tragedy Rhesus, now dated to the late fourth century BC, Hector describes torches burning all night aboard ships. Archaeological evidence includes four fragments from three plant-fibre torches, bearing traces of burning, found in first-century flood layers at the Cantiere delle Navi in Pisa. They are now in the Museo delle Navi Antiche.

Modelling a torch as a point source four metres above the water, with an output of 100 to 200 lumens, gives an improvement in visibility of about ten metres. This short-range benefit came at a cost: the flame reduced the dark adaptation of crew members nearby, making it harder for them to see at middle distance.

A Dressel 1A amphora found in the Hérault has drilled holes, binding marks and traces of burning. The triangular openings at the top were probably smoke vents, while a single large opening, about 10 cm across, directed light onto the water around the bow.

At an assumed output of 150 to 250 lumens, the illuminated area would have had a radius of roughly 50 metres, five times the reach calculated for a torch. The body of the amphora shielded the crew's eyes from the flame, while the opening concentrated the light in one direction.

In the fourteenth-century BC tomb of Kenamun at Thebes, amphorae are depicted bound to the stems of two Canaanite ships. A second-century relief on Trajan's Column in Rome shows ships leaving a harbour, with a bronze lamp hanging from the stern of one vessel. This has been interpreted as a position light.

Comparable objects have been found in the Grand Bassin B wreck, dated to 110–90 BC, and the Comacchio wreck, dated to 25–1 BC. These finds provide further evidence of the lighting devices carried aboard ancient ships.

Four Canaanite lamps with soot traces were recovered from the Uluburun wreck of the late fourteenth century BC. Similar finds have been made at Tektaş Burnu, Ma'agan Mikhael, Cala Culip IV and Comacchio.

On Caligula's first ship at Lake Nemi, dating from the first century, wall brackets were distributed along the length of the hull. Calculations for a lamp placed at a height of 1.3 to 1.6 metres give a lit radius of 2.5 to 3.2 metres. At the edge of that circle, the illumination would have been weaker than moonlight under a full moon.

Artificial light aboard ship made faint stars harder to see, complicating astronomical navigation. It also raised the ambient light threshold, making signal fires ashore more difficult to distinguish. During covert naval operations at night, a light brought the further disadvantage of making the vessel visible from the coast.

Lights were therefore used for particular tasks and circumstances rather than kept burning as a routine condition. The nineteenth-century steamship brought the idea that a vessel should carry lights whether or not its own crew needed them to see by.

Coastal lighting operated on a different scale. Wood fires ashore produced light intensities of around 5,000 candela. Fires were burned on towers from the sixth century BC, and the first purpose-built lighthouses appeared in the third century BC. An inscription on the tower at Cape Pyrgos on Thasos describes it as standing at the end of the harbour to guide and protect ships and seafarers.

Such installations were not possible aboard ship. A marine light had to be small and portable, protected from wind and spray, and safe near tarred rope and timber. These remained important constraints on lantern design, discussed further in Glass, Lenses and Burners.

Medieval and early modern sea law treated collision primarily as a question of liability. Article 15 of the Rolls of Oléron describes how the loss should be shared when one vessel strikes another:

If a ship is in her passage, at anchor or moored and another vessel comes from abroad and is not well steered, and strikes herself against the ship which is in her way… according to reason the damage of the blow ought to be valued and divided half by half of the two said ships.

The master and crew of the vessel responsible for the collision had to swear on the Gospels that they had not acted deliberately. The stated reason for dividing the loss was commercial: it discouraged the owner of an old vessel from placing it in the path of a more valuable one.

The article imposes no duty to show a light or keep a lookout, and gives no rule about which vessel should give way. It deals with responsibility after a collision rather than measures to prevent vessels from meeting in the dark.

By the seventeenth century, stern lanterns indicated rank. The French admiral's ships carried three large lanterns on the taffrail, as did the general's galleys. Squadron commanders carried a single stern lantern. These were tall, hexagonal, glazed lanterns mounted on an iron support.

Smaller lanterns were carried at mastheads and yardarms. Others lit the storerooms, hold and, later, the compass binnacle. The large stern lanterns identified who was in command rather than the vessel's heading or size, an important distinction for ships sailing together in a fleet after dark.

The French term phanal appears in a text by Rabelais dated 1552, describing a fire enclosed in a protective housing. In its later spelling, fanal, it came to describe a fire in a large marine lantern, placed high at the stern where it could be seen at night.

In 1776, Admiral Lord Richard Howe issued a signal book that included steering rules and night signal number 40. One light was shown at each cat-head, with a third placed above one of the pair to indicate whether the ship intended to hold her wind or bear away. Above the weather cat-head, the extra light meant bear away; above the lee cat-head, it meant hold the wind.

This is the earliest documented three-light arrangement traced for these pages. Unlike lights indicating position, rank or size, it communicated the vessel's intended manoeuvre. The wider use of lights to convey such information developed in the nineteenth century.

The steering rule itself was older. Royal Naval Signal Books issued before 1780 stated that when ships on different tacks had to cross close to one another, the vessel on the starboard tack should keep her wind, while the vessel on the larboard tack should pass to leeward. The steering instruction thus preceded the lights needed to communicate and follow it in darkness.

The subsequent development of lighting regulations is covered in The Road to Regulation, 1840–1977. That guide continues the account into the nineteenth century.

Primary: The Roll of Judgements of Oléron, article 15; Instructions for the Conduct of Ships of War, issued with the Royal Naval Signal Books before 1780; Trajan's Column, Rome, 107–113 AD; Reinier Nooms, The Battle of Leghorn, 1653–1664; Lieve Pietersz Verschuier, The Arrival of Charles II of England at Rotterdam.

Secondary: Mauro and Durastante, "On the Use of Lights as Night Navigation Aids During Antiquity", International Journal of Nautical Archaeology, 28 July 2025; R. Bahr, The Development of Regulations for Preventing Collisions in Inland, Inshore, and Open Waters of the U.K. During the First Half of the Nineteenth Century, MPhil, University of St Andrews, 1998; J.F. Kemp, Factors in the Prevention of Collisions at Sea, 1973, appendices II and III; B. Mahot, Les lanternes, phares et fanaux.

The paintings on this page are displayed at the Rijksmuseum.