How Did Anybody See After Dark for Five Thousand Years? A Wick in a Dish

A wick made of twisted fibre sits with one end in the oil and the other projecting above the vessel. Capillary action draws oil up through the fibres continuously, and heat from the flame vaporises the oil at the tip, so what actually burns is oil vapour rather than the wick or the liquid. The wick is consumed only slowly because the evaporating oil keeps it cool, which is why a lamp burns for hours from a small quantity of fuel and why the wick lasts. Brightness depends on how much surface the flame has, so a flat or broad wick gives more light than a round one of the same fuel consumption.

The fuel depended entirely on what was locally available:

• •Olive oil across the Mediterranean, which burns cleanly and was abundant
• •Animal fats and tallow where oil was scarce
• •Whale oil, which drove an entire industry from the eighteenth century
• •Sesame, castor, rapeseed and nut oils in various regions
• •Fish oil in the Arctic, burned in stone lamps with a moss wick
• •Paraffin from the 1850s, refined from coal and then from petroleum

The basic design was essentially unchanged for millennia and then improved twice in quick succession. Aime Argand patented a burner in 1780 using a tubular wick with air drawn up the inside as well as around the outside, which supplied far more oxygen to the flame and produced light several times brighter with less smoke, and adding a glass chimney above it created a draught that improved combustion further. Paraffin, refined commercially from the 1850s, supplied a cheap fuel that flowed freely and burned brightly, which made bright lamps affordable rather than a luxury. Together those changes produced the lamps that lit homes and streets until electricity replaced them.

The quantity of light available before electricity was far lower than most people imagine and it shaped daily life accordingly. A single wick lamp produces roughly the light of a modern candle, which is a tiny fraction of what one ordinary electric bulb supplies, and lighting a room adequately required many of them, which few households could afford in fuel. The cost of artificial light relative to wages fell by an enormous factor over two centuries, and economic historians have used that decline as an illustration of how conventional price measures understate improvement. The practical consequence was that fine work stopped at dusk for most people, that lamps were moved to wherever the work was rather than lighting a room, and that reading after dark was a luxury.

These objects are among the most useful finds on a site, for reasons beyond illumination. They were cheap, broke easily and were made in enormous numbers, so they occur constantly and in datable quantities. Their forms changed rapidly and regionally, which makes them precise chronological markers, and Roman examples in particular were mass-produced in moulds with makers' marks that allow production centres and trade routes to be traced. Decoration on the discus of a Roman lamp carries images ranging from gods to gladiators to explicit scenes, which supplies evidence about popular taste rather than elite taste. Residue analysis identifies the fuel burned. And their distribution within a building indicates which spaces were used after dark.

Capillary action draws oil up a wick and the heat vaporises it, so what burns is vapour and the wick is spared. Fuel was whatever was locally available, from olive oil to fish oil to paraffin. Argand's tubular wick with an internal air supply in 1780 and cheap paraffin from the 1850s were the two changes that mattered. The lamps themselves are precise chronological markers on archaeological sites.