Before Fiber, Light Learned to Follow a Stream
In 1854, John Tyndall made sunlight take a turn. He shone it into a basin of water with a hole in its bottom, then watched the light travel with the stream as the water poured out. The demonstration gave guided transmission of light a form anyone could see: light could enter a flowing path and remain within it as that path bent toward the floor, according to the fiber-optics history archive.
The basin had a history of its own. In 1841, Swiss physicist Daniel Colladon and French scientist Jacques Babinet had demonstrated that light could travel through a stream of water, establishing the striking physical effect that Tyndall would make famous. Tyndall then supplied the formal account in 1870, explaining the principle of total internal reflection, as TTI Fiber recounts. Light could meet a boundary at the right angle and reflect back into its path, again and again, rather than escaping through the side.
That explanation turned an elegant experiment into a usable idea. A stream of water offered the clearest picture: give light a path with the right boundaries, and the path can carry it around a curve. The future medium would trade water for glass, yet the operating logic had already appeared in that falling stream.
Inventors quickly began looking for places where guided light could do useful work. In 1880, William Wheeling used mirrored pipes to carry light from one source into many rooms. His system arrived during the rapid spread of Edison’s incandescent light bulb, which won widespread popularity and left Wheeling’s approach without a broad market, the fiber-optics history archive explains. The same year, Alexander Graham Bell invented the photophone, a device that carried voice signals through the air instead of wires.
Those two efforts pointed in different directions, one toward illumination inside buildings and the other toward communication, while sharing a confidence that light could carry something useful from one place to another. The water-stream experiments had shown how to guide it. The next challenge lay in making that guidance precise and durable enough for a signal to arrive cleanly after a meaningful journey.
Images Could Enter Glass, Then They Fell Apart
That challenge soon drew inventors toward images, which demanded that many points of light arrive in their original arrangement. In 1920, John Logie Baird in England and Clarence W. Hansell in the United States jointly filed a patent for carrying television images through transparent pipes, according to the fiber-optics history archive. The proposal gave the earlier demonstrations a new purpose: light could carry a picture if its separate rays could be kept in order from entrance to exit.
A 1933 transmission through transparent pipes showed that this purpose could be achieved in principle. The image quality was very, very low, and the transmission covered a very short distance, the history archive records. Those limits defined the engineering problem with unusual clarity. A picture depends on a signal retaining enough strength and structure to remain recognizable, and early bundles of fibers could preserve neither for long.
From the 1930s through the 1950s, fiber bundles used for imaging suffered signal loss above 1,000 dB/km, according to TTI Fiber. Light entering one end therefore thinned away rapidly as it traveled, leaving little value for any system meant to reach across a substantial distance. The work of the period centered on making the fibers themselves more useful, especially by giving light a more reliable path through glass.
In 1954, Narinder Kapany and Harold Hopkins made bundles of fibers that transmitted images, as the Fiber Optic Association timeline documents. Their bundles showed that glass could carry the visual information researchers wanted to send, bringing the concept closer to an instrument rather than a demonstration. That same year, Abraham van Heel suggested cladding fibers to reduce attenuation. His design placed a transparent coating around a bare glass fiber, using material with a lower refractive index so the light remained contained within the fiber, according to the fiber-optics history archive.
Cladding gave researchers a practical way to hold light on its course, and image bundles showed what organized light could convey. Yet the extraordinary loss still drained the signal before it could travel far. Glass had begun to guide light with control; the next task was to make that glass clean enough for communications.
Charles Kao Made the Case for Pure Glass
That task came into sharper focus in the 1960s, when the laser arrived and gave researchers a concentrated source of light to send into fiber. The fiber itself remained a severe constraint: its losses stood at 1 dB per meter in 1960, a level that kept the prospect of long-distance communications far from reach. The fiber-optics history archive records the scale of the gap, and the gap gave the decade a clear engineering question. Light could be launched into glass and held there. What, exactly, was consuming the signal along the way?
Charles Kao and George Hockham, working at Standard Telecommunication Laboratories, gave that question a practical answer in 1966. They identified impurities in the glass as the source of the high signal loss, then argued that high-purity silica could bring losses below 20 dB/km. According to TTI Fiber, their proposal changed the problem from an apparent limit of the medium into a problem of materials and manufacturing. The path forward required glass made with far greater purity, because every impurity along the route weakened the light carrying the message.
Kao made the target even more concrete by describing a communications fiber with a pure glass core and optical cladding, a design that joined the material challenge to the guiding method researchers had already established. The Fiber Optic Association timeline credits him with revealing how low-loss fiber suitable for communications could be made through that combination of cladding, purity, and impurity removal. The proposal supplied engineers with a number to pursue, less than 20 dB/km, and a route toward it, high-purity glass.
That precision mattered because it turned a promising idea into a testable demand. A laboratory could now ask whether its processes could produce glass clean enough to preserve a signal across the required distance, rather than merely demonstrate that light would enter and travel through a fiber. Kao’s work later received the 2009 Nobel Prize in Physics, as TTI Fiber notes, yet the immediate stakes in 1966 were direct and unresolved: could anyone manufacture the extraordinarily clean glass his target required?
A Laboratory Fiber Became a Working Telephone Line
The immediate stakes came into focus at Corning Glass Works in 1970, where Robert Maurer, Donald Keck, and Peter Schultz produced the first practical low-loss optical fiber. Their fiber reached attenuation of 17 dB/km, a result that placed the manufacturing challenge on the same side of the line as the communications target researchers had been pursuing. TTI Fiber records that achievement as the first practical low-loss fiber, and the word “practical” carried real weight: a promising material now had a version that engineers could build upon.
Corning then pushed that result forward with unusual speed. By 1972, the company had refined its manufacturing process enough to lower loss to 4 dB/km, making the glass far more suitable for the work ahead. According to TTI Fiber, the decline from 17 to 4 dB/km came within two years. The same year, Keck, Schultz, and Maurer developed a vapor-deposition method for making high-purity, low-loss fibers, a process that gave the breakthrough a route out of the laboratory and into repeatable production. The Fiber Optic Association timeline places that manufacturing advance alongside the sharp reduction in loss, linking the quality of the material directly to the possibility of a real network.
That possibility reached the telephone system in April 1977, when AT&T installed the first telecom link in coal tunnels beneath Chicago, Illinois. The setting mattered. Fiber had moved from controlled experiments into the physical spaces where telephone infrastructure already had to operate, amid the practical constraints of a working city. The Fiber Optic Association dates the installation to April and identifies the Chicago coal tunnels as the site. Three weeks later, GTE sent live telephone calls through fiber in Long Beach, California, turning the new medium from a single installation into a clear demonstration that voice traffic could travel on light. The same timeline records the Long Beach calls as the next step in 1977.
Those linked demonstrations changed the terms of the project. Phone companies had begun using fiber, and the work now moved toward the demands of an operating communications system: greater transmission speed, longer reach, and reliable service at scale. The glass had earned its place in the telephone network; engineers could begin asking how far that network could carry light.
TAT-8 Turned Light Into a Transatlantic Utility
That question carried engineers from city networks toward the far larger challenge of an ocean crossing. Between 1980 and 1984, AT&T built East and West Coast backbones in the United States that ran at 45 Mb/s, using 850 nm lasers in multimode fiber. Those routes gave the industry experience with fiber as a working backbone, where speed and dependable operation had to hold together day after day. The next step required the same confidence in an environment where repair and access came with entirely different stakes: the Atlantic floor. In 1983, AT&T Bell Labs tested an undersea fiber-optic cable in Atlantic water approximately 5 km deep, turning a communications ambition into a demanding engineering trial. The Fiber Optic Association timeline records both the growing domestic backbone and that deep-water test, two parts of the same preparation for a much longer route.
Bell Labs had already set out the central design in 1980, proposing TAT-8, a transatlantic cable built around single-mode fiber. The plan called for 565 Mb/second over two-pair fiber, a large jump from the earlier telephone links that had proved fiber could serve ordinary calls. The fiber-optics history archive captures the proposal at the moment when the industry could see a route across the Atlantic in technical terms: a defined fiber type, a defined speed, and a cable system designed to make the passage useful.
That definition mattered because an ocean cable had to do more than demonstrate that light could survive a journey. It had to become a dependable part of the telephone system, connecting people and networks across countries through a route that would stay in service. The years between the proposal and activation were therefore years in which fiber moved from promising transmission medium to infrastructure capable of carrying international traffic. By 1988, AT&T activated TAT-8, bringing the plan into operation. The fiber-optics history archive identifies 1988 as the year of activation, while the cable’s wider significance came from where it ran and what it established.
TAT-8 connected the United States, the United Kingdom, and France as the first transatlantic fiber-optic cable, and it remained in operation for 13 years. The Fiber Optic Association timeline places those facts alongside the cable’s installation by AT&T, making the scale of the achievement clear. A principle first made visible in flowing water had become a service route beneath the Atlantic, carrying signals through glass across the distance that had once separated continents by the limits of older communications systems. With that route in place, the focus could shift from crossing the ocean to multiplying what a single fiber could carry.
Engineers Found Room for a Hundred Signals in One Fiber
That service route beneath the Atlantic carried a new question with it: how much traffic could a strand of glass carry once it had proved it could cross an ocean? The answer came through a series of refinements that made room inside the same basic route. In 1988, 1300 nm became the standard wavelength for fiber-optic systems, a practical choice that shaped the equipment built around the expanding network. Four years later, 1550 nm systems appeared, bringing fiber attenuation to 0.2 dB/km at that wavelength, according to the fiber-optics history archive. That lower loss gave engineers more useful distance between the points where a signal needed attention, and it made each installed route more valuable.
Value rose further during the 1990s with the arrival of erbium-doped fiber amplifiers, or EDFAs. The new optical-amplifier technology opened the door to what historian Jeff Hecht describes as a bandwidth revolution in his history of fiber optics. Amplification changed the practical economics of a fiber route because engineers could build on the glass already in the ground and under the sea, then push far more traffic through it. The focus of the industry began to shift from creating a viable line to filling that line efficiently.
Dense Wave Division Multiplexing, widely available by 1998, supplied the next step. DWDM systems gave a single fiber room for multiple waves, and some of those systems reached an aggregate capacity of 10 Gb/second. By 2001, the technology had advanced to systems capable of transmitting 100 waves at 10 Gb/sec, for aggregate capacity of 1 terabit per second, the fiber-optics history archive reports. The progression matters because it changed the scale of the network without requiring a separate strand for every new stream of traffic. Engineers had found a way to multiply the work performed by the route they already had.
That multiplication created a powerful commercial story around telecommunications capacity. Investment raced toward the prospect of networks built for volumes of traffic that the new equipment could technically carry, and the industry entered a speculative bubble similar to the railroad boom of the late nineteenth century, as Hecht’s account observes. The telecom bubble burst in 2001, the same year that 100-wave DWDM systems reached their 1-terabit aggregate capacity. Fiber had delivered abundance with startling speed, while the businesses built around that abundance still had to find their pace.
The Glass Kept Improving After the Network Arrived
That break in the market arrived after the network itself had already taken on a global shape. Around 1990, the global fiber-optic telecommunications network was complete, according to Jeff Hecht’s history of fiber optics, linking places through routes that carried information as pulses of light inside glass. The crash changed the commercial mood around capacity, while the engineering work inside the fiber continued with steady purpose. Every reduction in loss gave those routes greater reach and made the material beneath the network more capable.
That progress had already gathered pace during the 1980s, when fiber loss fell to 0.5 dB/km. During the 1990s, the figure reached 0.2 dB/km, and in the 2000s ultra-low-loss fiber with attenuation as low as 0.16 dB/km became standard for submarine and long-haul networks, a standard that remains in place today, according to TTI Fiber. These are small-looking measurements with a large practical consequence: signals can travel farther through the same strand before the route needs help carrying them onward.
The network therefore kept gaining strength after its broad map had been drawn. Undersea cables and long-haul links could draw on fiber refined for the distances that define them, while the physical principle at the center of the system stayed remarkably consistent. Light enters glass, the glass holds its course, and the signal reaches farther as the material grows cleaner and more precisely made.
Don Keck captured that continuing momentum in 1995 with a brief assessment: “It’s not over,” as recorded in Hecht’s account. The line fits a technology whose decisive advances came through patient refinements in the medium itself. Modern communications gained their great highway through glass engineered to turn light into a signal, carry that signal across enormous distances, and keep it on course.
Comments (0)