← THE PEOPLE WHO WORKED THIS OUT

Thomas Edison

1847–1931 · 1879

A filament that took thousands of tries to find, by a lab full of people

Edison built a research lab at Menlo Park, New Jersey, in 1876 and ran it as an operation, not a one-man workshop: a staff of machinists, glassblowers and chemists, paid to grind through failure after failure until something worked.

Arc lamps already lit streets, but nobody had a small incandescent bulb that lasted. Edison’s team tested thousands of filament materials over more than a year, and on the night of 21–22 October 1879 a carbonized cotton thread burned inside a vacuum bulb for about thirteen and a half hours before it failed. He filed for a patent on the carbon-filament lamp two weeks later; it was granted in January 1880.

Thirteen hours is a laboratory result, not a product people can buy. Getting from there to a lamp worth selling took a different filament, made a different way — Lewis Latimer, working for a rival company, patented a method for manufacturing a carbon filament that lasted longer and was cheaper to produce, and his page on this site is where that half of the story belongs.

The lamp needed a power system behind it, and that is where Edison and Tesla ended up arguing for a decade, in what the histories call the war of the currents. Edison’s networks — Pearl Street, in Manhattan, the first of them, in 1882 — ran on direct current, the kind a battery makes; Westinghouse bought Tesla’s alternating-current patents in 1888, and AC could be stepped up to a high voltage for the wire and back down at the lamp, which direct current of the day could not do over distance. Edison’s side warned the voltages AC needed were dangerous; Westinghouse’s side answered that direct current could not reach past a small radius around each station. It was settled by economics: Westinghouse built the generating station at Niagara Falls on Tesla’s patents, sending its first power in August 1895 and lighting Buffalo, twenty-five miles off, by November 1896.

The thing you can build here

Burn out a lamp →

The filament problem is the burn-out problem: eighteen volts through a lamp built for nine, so you watch in a few milliseconds what a whole lab spent a year trying to avoid.

Sources

  • Edison, US Patent 223,898, "Electric Lamp", granted 27 January 1880.
  • Friedel and Israel, Edison’s Electric Light: Biography of an Invention, Rutgers University Press, 1986.
  • Jonnes, Empires of Light: Edison, Tesla, Westinghouse, and the Race to Electrify the World, Random House, 2003.