Light in glass was an old curiosity
The idea that light can follow a curved path inside a transparent material is older than electronics. In the 1840s, the physicists Daniel Colladon and Jacques Babinet showed that light could be guided along a stream of water, and John Tyndall gave a well-known public demonstration of a similar effect in 1854. The principle is total internal reflection: when light inside a dense medium meets its boundary at a shallow angle, it reflects back instead of escaping.
For decades this stayed a demonstration. In the 1950s, researchers working on medical viewing instruments, including Harold Hopkins and Narinder Singh Kapany, bundled thin glass fibers to carry images. Abraham van Heel and others found that coating a fiber with a layer of lower-index glass, called cladding, protected the surface and improved guiding. Kapany is often credited with popularizing the term fiber optics. These early fibers were fine for looking inside the body over a few feet, but they absorbed too much light to carry messages for any distance.
The problem was loss
By the early 1960s, the invention of the laser had given researchers a bright, narrow source of light, and engineers began to ask whether light could replace copper for communication. Lasers rely on stimulated emission, and they made the idea seem practical. Alexander Graham Bell had experimented with a light-beam telephone in 1880, but fog and rain made open-air links unreliable, so some form of guide was needed.
Ordinary optical glass lost most of its light within a few meters. Measured in decibels per kilometer, the loss was often cited as on the order of 1,000 dB per kilometer. A loss of 20 dB means only 1 percent of the light remains after one kilometer, so a system needed to get to about that level to be worth considering.
The 1966 proposal
Charles Kao and George Hockham, working at Standard Telecommunications Laboratories in Harlow, England, examined the question carefully. In a paper published in 1966, they argued that the high loss in glass was caused mainly by impurities rather than by anything fundamental to the material, and that a single-mode fiber with loss below 20 dB per kilometer could be practical for communication. This was a prediction, not a demonstration, and some colleagues were doubtful. Kao traveled to promote the idea, and the British Post Office and others began research. Kao shared the 2009 Nobel Prize in Physics for this work on light transmission in fibers.
The 1970 breakthrough
In 1970, researchers at Corning Glass Works in New York State, Robert Maurer, Donald Keck and Peter Schultz, made a fiber with a loss reported at around 20 dB per kilometer. They used fused silica, an extremely pure glass, with titanium added to the core to raise its refractive index. Corning’s fibers were improved rapidly in the next few years, and other laboratories, such as Bell Labs, developed manufacturing methods that deposited glass vapor inside tubes, which could then be drawn into a thin strand.
Two other advances arrived at about the same time. Semiconductor lasers that could run continuously at room temperature were demonstrated in 1970 by the group of Zhores Alferov in the Soviet Union and by Morton Panish and Izuo Hayashi at Bell Labs. Small, long-lived lasers and light emitting diodes made practical transmitters, and semiconductor detectors served as receivers. Together, a clear fiber, a compact light source and a sensitive detector made up a working system.
From trials to service
By the mid-1970s, loss in the best fibers had fallen below 1 dB per kilometer in some designs. Field trials followed: in 1976, Bell Laboratories ran an experimental system in Atlanta, and in 1977 telephone companies in the United States, including Bell and GTE, and the British Post Office began carrying live traffic over fiber in short city routes. These links joined telephone exchanges, replacing copper cables that were congested or costly to expand.
Early systems used light with a wavelength near 850 nanometers. Later work shifted to 1300 and 1550 nanometers, where glass is clearer, and lowered attenuation to roughly 0.2 dB per kilometer at 1550. At that level, a signal loses 20 dB over 100 kilometers, since 0.2 times 100 equals 20, which is why long routes could use few repeaters.
Why commercial use took off
Fiber won out for practical reasons. It is thin and light, immune to electrical interference, and carries far more information than copper pairs. Glass is made mostly from sand, though the purification process is expensive. As manufacturing improved, costs came down, and installers began laying fiber across countries in the 1980s.
This history belongs to a wider story of how telephone companies moved from analog to digital networks, a shift that also supported the growth of the internet’s backbone.