Glass Veins Worldwide: Kao, Corning, Ocean Thread

In 1887 a physics teacher fixed a glass rod to an arrow and loaded it into a homemade crossbow. He melted the tip into a pinhead bead and fired. What stayed behind was a glass thread less than a tenth of a human hair across. I count this experiment among the strangest lab moments ever. The teacher hoped to weigh tiny forces with such threads. Yet within decades other teams saw a different future in them: carrying messages on light.

Light staying trapped inside glass looks like sorcery. The rule is plain: moving from dense glass to thin air past a certain angle, light cannot leave and turns back. That is total internal reflection. The beaker demo shows it; a laser crossing a water vessel skims along the surface. Per the Britannica entry, a strand spans 125 micrometres outside while the light-carrying core is only about 10. To my eye it is a slide carrying water without spilling a drop.

In the 1960s phone companies faced trouble: everybody wanted to talk. Daily calls in America climbed from 285 million to 576 million from 1960 to 1973. While NASA carried people to the Moon for $26B in total, the phone giant had to bury nearly three times that sum in lines. Cables ran short, exchanges jammed. This picture explains why glass thread counted as necessity rather than luxury. More talk had to fit into a thin road.

Back then a voice rode copper as a wave. A membrane trembled, the current swung, and a far-end speaker turned it back into sound. One line meant one call; yet no firm could wire every home. Engineers found a sly path: grasping speech needs only 300 to 3400 hertz, so discard the rest. Then stack each talk on its own shelf and send them down one wire. At the far end filters sort the shelves, much like tuning a radio.

Piling on more talk meant climbing to higher frequencies. But copper showed a streak. Quick currents breed loops of current inside the wire; these loops fight shifts at the center and boost shifts near the skin. So the signal flees to the layer while the middle idles. The label for this is the skin effect, and it grows with frequency. Chasing more shelves made each shelf pricier. The bench demo of this tug signaled that copper had hit its ceiling.

Glass looked the best bet for the voyage of light. Yet strands were awful: a signal died before one kilometer. Two partners took the blame. One was scattering : flaws inside the glass throw light every way, like headlights in fog. The other was absorption: atoms swallow light and turn it to heat. The worst dirt was iron; a pinch could kill a line. The Wikipedia record gauges loss in dB/km; then figures ran into the hundreds.

While all blamed the glass, an engineer named Charles Kao flipped the question: what if the grime within was at fault? His 1966 paper showed: with pure enough glass, light could travel kilometers, and 20 decibels per kilometer would work. The meaning was stark: one kilometer eats 99 percent of the power, yet the surviving 1 percent bears the message. For bench tests Kao found hard-to-melt silica rods and gauged short pieces. That forecast earned him the 2009 NobelPrize.

Purifying glass fell to a trio at Corning: Robert Maurer, Donald Keck and Peter Schultz. Their gambit ran against instinct: lace the core on purpose with additives to tune its power. Through the summer of 1969 their furnace roared; titanium trials drank the light. On 22 July 1970, a stub went to the bench and read 17 decibels. Keck logged the figure and scribbled a note beside it. In its Kao farewell, Nature frames the race as fancy meeting craft.

The wonder hid in the making. The team fed silicon-bearing gases into a flame and settled the soot layer upon layer inside a spinning tube. Then they glazed the tube near 2000 degrees and slumped it shut into one rod. This stock is the preform ; from a thick billet the tower pulls hair-fine strand. My image: like candy floss gathering off spinning sugar, glass thread is born from gas. The road from bench tubes to factory furnaces opened that way.

The craft runs at factory scale. The Prysmian tour shows this order: first the soot rod glazes in a furnace, and the gap seals itself through surface pull. Then silica joins the outside until the core fills just 7.2 percent of the width. During the pull a plasma torch breathes vapor three times hotter than the Sun. Millions of kilometers of strand leave each month. That scale startled me; the bench glass rod had grown into an industry.

Glass strand passed its true test under the ocean. TAT-8, live from December 1988, linked America with Europe as the first glass cable across an ocean: 280 megabits over 5846 kilometers, room for 40 thousand separate talks, at a price of $335M. The ETHW milestone roll calls this line the forebear of the present web. Demand filled it in eighteen months; nobody had foreseen such hunger. To me this marks a moment when people, not tools, steered the course.

Across the decades loss sank near its floor: about 0.2 decibels per kilometer at 1550 nanometres. Dirt no longer leads; the unavoidable shivers of the glass frame itself do, under the name of Rayleigh scattering. Subsea lines now push 24 terabits a second through one strand. At the rate Nature cites in its farewell, over 95 percent of traffic rides these glass threads. In short, the 1 percent Kao hoped for has swollen into a river a million times over.

The detail that grips me most: Kao was no glass specialist but an engineer. From outside, he could pose the query none posed: instead of firing, why not wash it cleaner? The Corning crew likewise walked the odd path, fixing the core by spoiling it. My takeaway is plain: in each jammed setup the culprit sits beyond your stare, in the corner you skip. Glass needed refining, so did the gaze. The secret carrying the net is this refining.